THE MISSING ELEMENT OF FIRM DIGITALIZATION Lessons for the EU Member States Best practices for enhancing MSMEs Digitalization and Managerial Capacity Building Funded by the European Union THE MISSING ELEMENT OF FIRM DIGITALIZATION Lessons for the EU Member States Best practices for enhancing MSMEs Digitalization and Managerial Capacity Building Digitrans Project – Output 3. Note on Global Lessons and Evidence Funded by the European Union © 2023 International Bank for Reconstruc- tion and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and con- clusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning legal status of any territory or the endorsement or acceptance of such boundaries. Rights and permissions The material in this work is subject to copy- right. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, includ- ing subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@ worldbank.org. This document was produced with the financial assistance of the European Union. The views expressed herein can in no way be taken to reflect the official opinion of the European Union. Trust Fund (No. TF073759) – EC Contract No 2021CE160AT132. Cover and layout design and typesetting: Piotr Ruczynski, London, United Kingdom CONTENTS Acknowledgements   7 Abbreviations   8 Executive Summary   9 INTRODUCTION A global review applicable to catching-up regions in Europe    12 1 Summary of lessons and promising strategies    16 1.1 Typology of firms to target and their challenges in digitalization    18 1.2 Menu of promising instruments based on a review of global evidence    24 1.3 Lessons learned from global review of programs    29 1.4 Design guide for policy makers    42 2 Diagnostic tools    50 2.1 Surveys    53 2.2 Digital diagnostic tools    63 3 Programs and instruments    73 3.1 Informational constraints    75 3.2 Preferences and behavioral factors    87 3.3 Organizational misalignments    94 ANNEX 1 List of analyzed Digital Diagnostic Tools    101 ANNEX 2 List of analyzed programs and instruments    105 ANNEX 3 Evidence of impact of instruments for addressing gaps in complementary capabilities    119 Bibliography    121 BOXES 1.1 What’s new about this report    17 1.2 Examples of multi-instrument programs    27 1.3 Examples of online one-stop-shop information platforms    32 1.4 Examples of customizable public instruments supporting MSME digitalization    38 1.5 Examples of community engagement in support programs    40 1.6 Sectoral considerations in support for complementary capabilities of MSME digitalization    44 2.1 Post-diagnosis support DRAT    68 FIGURES 0.1 Structure of the diagnostic component of the Digitrans project    13 1.1 General business functions and technologies    19 1.2 Target groups of MSMEs by level of digital sophistication    20 1.3 Main challenges in complementary capabilities inhibiting digitalization of MSMEs    21 1.4 Typology of instruments for addressing gaps in complementary capabilities    25 1.5 Funnels in the typology of instruments    28 1.6 Evolution of approach to support digital and innovation policy in the EU    31 1.7 Areas covered by successful transformation plans    33 1.8 Self-assessment with respect to other firms in Poland    35 1.9 Logic model    46 1.10 Capabilities escalator for agencies supporting the digitalization of small and medium enterprises    47 2.1 Categorization of Digital Diagnostic Tools (DDTs)    63 2.2 MSMEs digital sophistication level according based on the Digital Maturity Assessment    65 2.3 Components for calculating Digital Quotient score    72 TABLES 0.1 Key recommendations    14 1.1 Lessons learned for firm digitalization and managerial capability building    30 1.2 Best-practice tools for generating evidence to inform instrument design    36 1.3 Steps in designing digitalization instruments    43 1.4 Comparison of market and system failure    43 1.5 Criteria for selecting a digital diagnostic tool    45 1.6 Instrument features for increasing the scale of intervention    48 1.7 Do’s and don’ts in monitoring and evaluation of pilot programs    49 2.1 Comparison of surveys and digital diagnostic tools for pilot design    52 2.2 Summary of Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises    54 2.3 Information covered by the Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises    55 2.4 Composition of the Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises    56 2.5 Summary of Firm-level Adoption of Technology    57 2.6 Information covered by the Firm-Level Adoption of Technology survey    58 2.7 Composition of the Firm-Level Technology Adoption survey    58 2.8 Summary of Management and Organizational Practices Survey    60 2.9 Information covered by the Management and Organizational Practices Survey    61 2.10 Composition of the Management and Organizational Practices Survey    62 2.11 Summary of Digital Maturity Assessment    64 2.12 Summary of Digital Readiness Assessment Tool    66 2.13 Summary of InnoCAPE Digital Maturity Assessment    69 2.14 List of partners in InnoCAPE consortium    70 2.15 Summary of Digital Quotient    71 A2.2 List of instruments and programs    105 A2.1 Countries included in analysis of programs and instruments     105 ACKNOWLEDGEMENTS This document was prepared by a team led by Natasha Kapil (Task Team Leader, Senior Pri- vate Sector Specialist), Francisco Campos (Task Team Leader, Senior Economist), Leonardo Iacovone (Task Team Leader, Lead Economist), and Damian Iwanowski (Extended Term Con- sultant), that comprised Raphael Martins (Consultant) and Eric A. C. Le Houerou (Consultant) from the Finance, Competitiveness, and Innovation Global Practice of the World Bank Group (WBG). The authors are grateful for useful comments from peer reviewers Ana Paula Cusolito (Senior Economist) and Anwar Aridi (Senior Private Sector Specialist), as well as inputs pro- vided by Łukasz Marć (Senior Economist), and the excellent editorial support of Aarre Laakso (Consultant). The team thanks the guidance and oversight of Ilias Skamnelos (Practice Manager, FCI), Lalita Moorty (Regional Director), Gallina Andronova (Country Director), Lasse Melgaard (Resident Representative for Bulgaria and Slovakia), Marcus Heinz (Resident Representative for Poland), and Reena Badiani-Magnusson (Program Leader for Equity, Finance, and Institutions Practice Group). The team benefited from useful discussions with European Commission DG REGIO, notably with Ramunas Linartas and Jeroen Van Oel, as well as greatly from insights provided by the authors and managers of surveys, digital diagnostic tools and public support programs reviewed for the purpose of this document, including the Agency for Science, Technology and Research Singapore; the Bundesministerium fur Wirtschaft und Klimaschutz (Germany); the Colegio des Caminos, Canales y Puertos; the Department for Business, Energy & Industrial Strategy (United Kingdom); Eurostat; Fomento y Turismo (Chile); the Infocomm Media Development Authority (Singapore); the Innocape Consortium; Invest in Denmark; McKinsey; the Ministerio de Economia (Spain); the Ministry of Business, Innovation and Employment (New Zealand); TM Forum; Red.es; and the US Census Bureau. Ultimately, the opinions expressed in this report are those of the authors. 7 ABBREVIATIONS AI artificial intelligence B2B business-to-business CATI computer-assisted telephone interview CIS Community Innovation Survey COVID-19 coronavirus disease 2019 CSI Community Survey on ICT Usage and e-Commerce in Enterprises DDT digital diagnostic tool DIH digital innovation hub DQ Digital Quotient DMA Digital Maturity Assessment DRAT Digital Readiness Assessment Tool DSP Digital Skills Partnership EU European Union FAT Firm-Level Adoption of Technology HR human resources ICT information and communication technology IDP Industrial Development Plan IT information technology KPI key performance indicator M&E monitoring and evaluation MOPS Management and Organizational Practices Survey MSME micro, small, and medium enterprise OAP Acelera Pyme Office OECD Organization for Economic Co-operation and Development R&D research and development SME small and medium enterprise 8 EXECUTIVE SUMMARY Objective As increasing the pace of productivity growth continues to be a strong challenge in Europe despite the continuous global technology developments, promoting the adoption and usage of digital technology has become more important than ever, particularly among micro, small, and medium enterprises (MSMEs). This guide aims to provide policymakers and practitioners with a comprehensive resource that outlines effective strategies for addressing gaps in com- plementary capabilities to the digitalization of MSMEs that can help enhance their productivity and resilience. Complementary capabilities are informational, managerial, and behavioral factors condition- ing successful digital technology adoption. Policymakers need to consider the importance of these capabilities when designing support instruments for firm digitalization. Lack of knowl- edge and skills, particularly among managers, is a significant barrier to technology adoption. Understanding behavioral factors such as beliefs, organizational culture, and influence of peers and networks, is crucial for policymakers to design effective interventions that address these barriers and promote digitalization. The guide offers a range of policy recommendations on how to identify challenges specific type of firms’ face, what interventions to offer, and how to structure those interventions for maximizing development impact. Policymakers and practitioners can use these recommen- dations to support the development of MSMEs in Europe, ensuring that these businesses have the necessary tools and resources to thrive in the digital economy. The strategies outlined in this guide are very useful for the design and implementation of programs to improve the com- petitiveness and productivity enhancing of MSMEs, promote economic recovery, and expand the living standards of citizens. Analytical approach This report is designed to provide practical guidelines for designing programs to help MSMEs increase the adoption and usage of digital tools and systems. It does this by first outlining the target groups of MSMEs and their challenges with digitalization, then leveraging a global literature review to provide recommendations for specific interventions linked to the identi- fied challenges, and finally distilling the most relevant operational features to consider when implementing such programs. The main challenges in complementary capabilities that deter firms from adopting new tech- nologies and scaling up the usage of their digital tools can be broken down into three groups: (1) informational constraints, (2) preferences and behavioral biases, and (3) organizational misalignment. Among the informational constraints, firms face both demand side constraints such as lack of awareness of possible technological upgrading opportunities or lack of knowl- edge on benefits of upgrading, and supply side issues including market gaps in offering. Among preferences and behavioral biases, MSMEs face present-preference bias, status-quo bias, change-holding behavior, and risk and loss aversion, as well as workers resistance to use 9 Executive Summary new technologies. Among organizational misalignment, firms are challenged by the low digital literacy of workers and gaps in managerial practices. The report suggests establishing foundations for adoption and usage of digital tools, stan- dardizing business processes and monitoring their performance, and integrating advanced technologies in multiple business functions to improve the level of digital sophistication. The report also provides actionable recommendations for designing interventions to support firm digitalization by reviewing the literature on the role of complementary non-technological aspects in firms’ adoption and use of digital technologies. This can help policy makers create targeted and tailored interventions when it comes to digitalization of MSMEs. Promising instruments The report highlights a range of promising instruments that can address digital challenges faced by MSMEs. The instruments identified in the report are based on a review of global pro- grams and policies that are applicable to European Union Member States. The report empha- sizes the importance of adopting international best practices that have proven to be effective in mitigating these challenges. Promising instruments to address these challenges include providing timely and relevant knowledge, training, technology demonstration, and implemen- tation assistance. Additionally, small, and temporary subsidies, showcasing successful peers, and offering resources for low-risk experimentation can help address behavioral constraints limiting the speed of adoption and usage of digital technologies by MSMEs. Other suggestions are provided throughout the report. The report suggests that there is no set of instruments that would be effective in every context, as each target group has a unique combination of digital capabilities. The instruments assigned to groups of challenges should be considered based on the level of digital sophistication of MSMEs. The staged approach can offer instruments of various levels of sophistication, providing more advanced support only after firms have benefited from more fundamental interventions. Lessons learned Furthermore, the report presents eight lessons on effective interventions supporting MSMEs in the adoption and usage of digital technologies. These lessons are derived from an analysis of programs and instruments implemented in several countries, a review of literature and im- pact evaluation studies. The first three lessons address implementation challenges relevant to program managers and public organizations, while the remaining five lessons focus on specific target groups or particular challenges. To support MSME digitalization, it’s important to recognize that there’s a scarcity of proven approaches, and interventions must be carefully designed to target complementary capabilities. By addressing overconfidence in firms’ current digital capabilities, especially at the foundational level, interventions can facilitate technological upgrading. Including diagnostic components in MSME digitalization programs is essential to identify areas of improvement and provide targeted solutions for different firms. Programs should prioritize market-based interventions for older and more established firms, while customized interventions can target firms at intermediate and high level of digital sophistication. Combining online and offline methods can amplify program impact, and establishing trust with the target population is crucial to securing buy-in. Finally, exploring group interventions can provide cost efficiency and adaptability. By implementing these lessons, MSMEs can benefit from digitalization, driving growth and innovation in the business community. 10 Design guide Executive Summary The report guides policymakers on designing effective interventions for digitalization gaps among MSMEs. The process involves a diagnostic component and an intervention component, with six key steps to enhance adoption and usage of digital technologies. Effective implementation requires prepared organizations and program managers, and programs need to be gradually scaled up with accumulated resources. Moreover, policymakers should consider regulatory and infrastructure bottlenecks and ensure an adequate external environment for program design. Embedding monitoring and evaluation throughout each step of the program design and im- plementation process is crucial for generating knowledge, tracking impact, and improving learning for the further improvement of interventions pursuing the digitalization of MSMEs. This helps policymakers to assess the effectiveness of the interventions and identify areas for improvement, leading to more informed decision-making and the potential for greater success in achieving the desired outcomes in digital transformation in Europe. 11 INTRODUCTION A GLOBAL REVIEW APPLICABLE TO CATCHING-UP REGIONS IN EUROPE Introduction  Digital technology is an important driver of the productivity growth of micro, small, and medium enterprises (MSMEs), but their potential is not realized due to many factors, including gaps in complementary capabilities. MSMEs are the backbone of the European Union (EU) economy, constituting over 95 percent of firms in all Member States.1 They generate more than half of the gross domestic product and employ over 100 million people. MSME digitalization can lead to higher productivity, which is critical for recovery from the COVID-19 crisis and improving the living standards of European citizens. Yet, many MSMEs are not taking full advantage of the opportunity to improve their productivity and resilience through technology adoption. One  A global review applicable to catching-up regions in Europe issue is that many attempts to increase technology adoption among MSMEs have only focused on technology. However, improving complementary non-technological capabilities (including managerial capabilities, technology preferences, level of knowledge, and behavioral attitudes toward digital transformation) is also essential for effective technology adoption. In this context, the Digitrans project aims to design and pilot a cutting-edge pilot instrument to strengthen firm productivity and resilience through promoting the adoption and usage of digital technologies by addressing challenges in complementary managerial and organiza- tional practices. The project will achieve this through a rapid diagnostic, pilot design, and field testing, followed by a pilot intervention and evaluation in selected pilot countries, with recommendations for future scalability. The pilot instrument will consist of diagnostic and intervention components to address specific barriers to adopting digital technologies. The diagnostic component will leverage on surveys and digital diagnostic tools (DDTs) to assess and reach the target population. The intervention component will provide direct support to beneficiaries through a combination of interventions being designed. This report focuses on catching-up firms but also includes information relevant for more mature enterprises. This report (Output 3) is part of the diagnostic component of the project, which is developing a comprehensive evidence base for designing a pilot instrument (Figure 0.1). Sep- arate, country-specific documents (Output 2) will provide information on needs, barriers, and institutions in the firm digitalization ecosystem for the selected pilot countries. FIGURE 0.1 Structure of the diagnostic component of the Digitrans project OUTPUT 1 OUTPUT 2 Framework Country-specific reports Note will be prepared by Insights on the World Bank team economic needs Methodology on Programs, Institutions based on learning from and feasibility of Note and Barriers global engagements addressing them Comprehensive evidence base for designing OUTPUT 3 the pilot Note on Global Report will reflect the Information on needs of respective pilot instrument Lessons instrument features countries to cover most maximizing its and Evidence pressing knowledge gaps impact Source: Digitrans team. The primary objective of this report is to inform pilot interventions aiming for addressing com- plementary capabilities for firm digitalization by emphasizing good practices for design and implementation. The good practices are based on a systematic review of the global lessons and evidence that policy makers could leverage for assessing and addressing firm-level gaps in capabilities necessary for MSMEs to leverage the full potential of digital technology adop- tion. The report codifies lessons from this systematic review of the relevant global evidence across three chapters. 1. Digitrans team based on Eurostat data. 13 Introduction  Chapter 1 provides a summary of lessons learned to inform four crucial policy decisions for successful advancing MSME digitalization (Table 0.1). It presents a typology of firms to target for digitalization support and identifies their challenges, answering the question where to intervene. It is followed by a menu of effective instruments addressing gaps in complementary capabilities, highlighting key success factors, necessary implementation capabilities, and policy features to adapt instruments to the heterogeneous needs of firms, guiding decisions on what interventions offer to firms? It then de- scribes a set of lessons learned from global programs for designing effective instruments addressing non-technological challenges hindering the adoption and usage of digital technologies, informing  A global review applicable to catching-up regions in Europe how interventions should be structured. Finally, it offers a design guide for policy makers with a series of six steps to answer how institutions could effectively design and implement interventions. TABLE 0.1 Key recommendations Target Firms are not monolithic; different parts of an enterprise usually have different technological advancement. groups and MSMEs can be grouped in three levels based on their digital sophistication, that is, on which technologies classification they most frequently use for performing general business functions. of challenges Three types of challenges in complementary capabilities prevent MSMEs from advancing digital sophistication: • Informational constraints, which include lack of awareness of possible technological upgrading, lack of knowledge on enhancing productivity of digital investments, and supplier market failures. • Preferences and behavioral biases, encompassing behavioral bias, and workers resistance to change and incentive misalignments. • Organizational misalignments, related to low digital human capital of workers and inadequate managerial practices. Various challenges impact each of the three groups of MSMEs to a different degree, and their prevalence reflect also firm characteristics other than level of digital sophistication, such as location and sector. Typology of We have identified 20 instruments with potential for addressing respective types of challenges. But while promising research has illuminated the evidence of impact for some of instruments, many of them remain unexplored, instruments especially within the context of complementary capabilities for MSME digitalization. • Instruments to tackle informational constraints include providing timely and relevant knowledge, training, technology demonstration and implementation assistance. • Policymakers could seek to address behavioral biases by demonstrating the benefits of digitalization for firms, showcasing successful peers, and offering resources for low-risk experimentation. • Promising instruments to address misalignment in adopting and using digital tools relate to various forms of managerial capacity building, targeted business advisory and workers’ education. Multiple instruments can be implemented and managed together to leverage synergies between them and form programs, as is often done in most developed digital ecosystems. Lessons We have identified 8 lessons for successful interventions that have been extracted from analysis of programs learned and instruments implemented in over a dozen of countries on four continents. Cross-cutting themes 1. Carefully design interventions targeting complementary capabilities given scarcity of proven approaches 2. Combine online and offline methods to amplify impact of programs enhancing complementary capabilities 3. Establish trust to secure buy-in from the target population Specific lessons 4. Include diagnostic components in programs supporting MSME digitalization 5. Address overconfidence preventing technological upgrading of firms at foundational level of digital sophistication 6. Customize interventions targeting firms at intermediate and high level of digital sophistication 7. Prioritize market-based interventions for older and more established firms 8. Explore group interventions to drive impact, cost efficiency, and adaptability 14 Introduction  Design guide The guidance describes institutional recommendations for integrating recommendations and lessons learned into design and implementation of instruments. It is organized in six steps to organize work of institutions managing interventions aiming for enhancing MSME digitalization: • Step 1. Determine target group and specify the challenges • Step 2. Identify firm needs and determine outreach strategy • Step 3. Create transformation plan • Step 4. Determine support activities • Step 5. Set up implementation arrangements  A global review applicable to catching-up regions in Europe Monitoring and impact evaluation should be embedded at all stages of program design. Information gathered in this chapter will also be useful for policy makers aiming to enhance MSME digitalization in other contexts. Source: Digitrans team. While Chapter 1 summarizes main findings and recommendations, Chapter 2 identifies global best practices in diagnostic tools used in interventions supporting firm-level digitalization. Chapter 2 defines tools used to generate information about digital needs of enterprises, both in aggregation and on the level of individual firms. It discusses characteristics of surveys and light-touch and high-intensity DDTs and their relevance for the design of the pilot instrument. The Chapter review in detail best practice examples of tools owned by public and private orga- nizations and highlights aspects crucial for adaptation to catching-up contexts and replicability in various social, economic, and institutional settings. List of all analyzed DDTs, out of which best practices were identified, is presented in Annex 1. Chapter 3 reviews programs and instruments, following a problem-driven approach. The chap- ter catalogs challenges related to the operational design features and evaluation practices of global programs for promoting the adoption and usage of digital technologies and complemen- tary managerial and organizational practices. For each of ten challenges often encountered in firm-level support instruments aiming to enhance digitalization, the chapter justifies public intervention and provides policy options. The challenges fall into three groups — information constraints, preferences and behavioral bias, and organizational misalignment — that are well grounded in the scientific literature. Subsections dedicated to each group of challenges contain conceptual introductions that highlight common causes of the challenges in the group and discuss related market and system failures. Annex 2 presents all instruments and programs to support firm-level complementary capabilities identified for the purpose of this report. This report could be extended to new interventions or examples of best practices to respond to the needs identified over the course of upcoming pilot instruments. The report is useful for supporting programming efforts in different settings, including interventions implemented under the 2021 – 2027 Multiannual Financial Framework. Insights collected in the report would also be useful for designing operational features of programing in multiple Member States. Finally, information in this report applies to interventions outside EU policy settings, and the report presents useful recommendations for enhancing the digitalization of catching-up firms in economies worldwide. 15 1 SUMMARY OF LESSONS AND PROMISING STRATEGIES 1 Chapter 1 summarizes lessons from global programming and evaluation evidence about what   works to address gaps in MSMEs’ adoption and usage of digital technology. This chapter seeks to catalog the operational design features and evaluation practices of global programs for promoting digitalization and complementary managerial capabilities.  Summary of lessons and promising strategies In designing programs, it is important to consider global evidence and existing practices. This chapter includes the following elements: (i) identifies the target groups of MSMEs, (ii) describes their challenges at the intersection of managerial capabilities and the digitalization process, (iii) leverages a global literature review to provide recommendations for specific interventions linked to the identified challenges, and (iv) distills the most relevant operational features to consider when implementing such programs. These elements help inform implementation for a greater impact. They also elucidate the most important aspects of the digitalization gaps in the EU context. The practical guidelines presented in this chapter were informed by a thorough review of global good practices and the literature. (See Box 1.1) Over 70 diagnostic tools were reviewed, together with almost 100 instruments and over 40 programs. They were analyzed in light of the results of literature research on digitalization of MSMEs and best practices in design and implementation of firm-level support instruments. Guidelines based on the review provide policy options for pilot instruments that accommodate the readiness of firms at various levels of digitalization and align with resource constraints. When applicable, the guidelines refer to research directly related to instruments for supporting complementary capabilities. BOX 1.1 What’s new about this report 1. The report fills in the gap in knowledge on policies for by reviewing the literature on the role of complemen- addressing challenges in complementary capabil- tary non-technological aspects in firms’ adoption and ities of MSMEs that prevent catching-up firms from use of digital technologies. advancing digital sophistication through a struc- 3. This report outlines key implementation and institu- tured analysis of global policies and instruments for tional good practices by analyzing lessons learned in addressing the most prevalent market and system the management and implementation of firm-level failures. instruments in the context of knowledge, managerial, 2. This report provides actionable recommendations for and behavioral factors determining the digital trans- designing interventions to support firm digitalization formation of MSMEs. 17 1.1 TYPOLOGY OF FIRMS TO TARGET 1   AND THEIR CHALLENGES IN DIGITALIZATION  Summary of lessons and promising strategies KEY LESSONS FROM SECTION 1.1 Firms can be grouped in three levels based on their technological sophistication, that is, on which technologies they most frequently use for performing general business functions. These are foundational, intermediate, and advanced. MSMEs have heterogeneous characteristics, capacities, and constraints, and it is crucial to differentiate and separate firms into target groups. Providing targeted interventions to enterprises of similar characteristics reduce the risks of not meeting firms’ needs. Different parts of an enterprise usually have different technological advancement, which determines the need to utilize granular data when defining target groups. A granular approach allows for identifying shallow digitalization and non- technological barriers preventing firms from leveraging the full benefits of digital technologies. Firms can be grouped in three levels based on their technological sophistication, that is, on which technologies they most frequently use for performing general business functions. Enterprises at foundational level of digital sophistication prioritize the adoption and usage of technologies that are considered easiest to use. Firms in foundational and intermediate groups exhibit higher gaps in complementary capabilities compared to firms at high level of digital sophistication, which have already addressed most challenges related to complementary capabilities. Challenges related to complementary capabilities are internal to firms that condition the digital transformation process. They are often overlooked by firms and policymakers, and include Various challenges impact each of the three groups of MSMEs to different degrees. As the characteristics, capacities, and constraints on MSMEs are heterogeneous, a crucial first step is to differentiate and separate firms into target groups. Compared to designing a single intervention for all firms, providing a targeted intervention to enterprises of similar character- istics reduces the risks of not meeting firms’ needs. Global evidence from programs focuses on targeted types of firms, which means that the challenges associated with specific groups of firms will be easier to match with a menu of alternative programs. 18 1 Firms are not monolithic; different parts of an enterprise usually have different technological   advancement, which determines the need to utilize granular data when defining target groups. All firms perform general business functions, such as administration, planning, purchasing, marketing, sales, and quality control. In all these functions, firms use different technologies, ranging from  Summary of lessons and promising strategies pen-and-paper tools to advanced artificial intelligence (AI) solutions, the digital sophistication of which may be scored from index 1 (the lowest level of technology adoption) to index 5 (the highest level of technology adoption) (Figure 1.1). To capture this heterogeneity, the common approach is to differentiate between adoption and use of technology in defining the potential target groups (Nadkarni & Pruegel 2021). Adoption of technology, like the extensive margin, measures the most advanced technology of a given type acquired by a company. Usage of technology, like the intensive margin, refers to the technology used most often. Such a granular approach allows for identifying shallow digitalization, which occurs when firms adopt digital technology in isolation, without changes in management processes, adjustments to the organization, or connection with other solutions (Borowiecki et al. 2021). The granular approach also guides analysis to identify non-technological barriers preventing firms from leveraging the full benefits of digital technolo- gies, which the most common interventions often omit (World Bank 2019). FIGURE 1.1 General business functions and technologies Functions Index Production or service Marketing and Administration operations Sourcing and Consumer Payment Quality (HR, Finances) planning Procurement Information Sales methods control Manual search Manual, visual or Handwritten Handwritten without Informal chat Direct sales written process- 1 processes processes centralized (face-to-face) at the Cash es without the database premises support of digital technologies Direct sales Bank Computers Computers Computers by phone with wire with standard with standard Online or e-mail 2 software software standard software chat Manual, visual or (e.g. Excel) (e.g. Excel) (e.g. Excel) written process- es with the sup- Sales port of digital Credit or technologies through so- debit card cial media Mobile Apps Mobile Apps Mobile Apps Structured 3 or digital or digital or digital costomer platforms platforms platforms surveys Online sales using external Online platforms banking Statistical process Customer control SRM (not inte- Relationship Specialized Specialized grated with 4 software software production Management (CRM) Online throuh planning) Online sales platfrom software (own website) (PayU) ERP (Enterprise ERP (Enterprise Resource Resource SRM Automated 5 Planning) Planning) (integrated Big Data, Integrated Virtual systems for or equivalent or equivalent with produc- AI with SRM currency inspection software software tion planning) Source: Cirera, Comin, and Cruz 2022. Note: AI = artificial intelligence; HR = human resources; SRM = Supplier Relationships Management. 19 1 Firms can be grouped in three levels based on their technological sophistication, that is, on   which technologies they most frequently use for performing general business functions. The project team analyzed statistical data on the adoption and use of digital technologies, lever- aging survey information collected by both Eurostat2 and the World Bank3 to define potential  Summary of lessons and promising strategies target groups (Figure 1.2). Enterprises at lower (foundational) levels of digital sophistication prioritize the adoption and usage of technologies that are considered easiest to use (Cirera et al. 2022). With growing awareness of the benefits of digitalization and increasing technological sophistication, firms tend to focus more on digitalizing the crucial business processes that determine firm competitiveness, leading to the highest efficiency gains (Stamoulis 2022). Pri- oritization allows these firms to achieve an intermediate level of digital sophistication. Firms in these two groups, foundational and intermediate, exhibit gaps in complementary capabili- ties, evidenced by, for example, a high share of firms reporting a lack of technological knowl- edge as a barrier to adopting information and communication technology (ICT)4 or a lack of awareness of their level of digitalization (World Bank 2022). Thus, they are the focus of the Digitrans project. Companies at a high level of digital sophistication have already addressed most challenges related to complementary capabilities. They remain outside the scope of the pilot instrument while being a target population for interventions to support research and development (R&D) or innovation. FIGURE 1.2 Target groups of MSMEs by level of digital sophistication Adoption and usage of digital technologies Complementary capabilities The firm has successfully integrated advanced The firm utilizes managerial practices technologies at index 4 or 5 in multiple business allowing for planning, executing, and functions. Its benefits are clear for key managers, evaluating medium- to long-term High and the firm aims to upgrade technologies across investments. The adequacy of various business functions. Most processes employees’ digital skills is monitored, in the firm are standardized and controlled, and digitalization is an integral part maximizing the effectiveness of digital tools. of training. Level of digital sophistication of MSME How to improve Standardize business processes and monitor their performance. Use the level of digital benchmarking information when making technological decisions and sophistication include the digital angle in strategic and operational plans. Despite adopting more advanced Knowledge and skills necessary to use technologies in some business functions, digital tools in a business context are the firm predominantly uses technologies limited to a small number of individuals Intermediate at index 2 or 3. Fragmentary processes directly operating already adopted are standardized and monitored, which technologies. Employees performing already generate data or create conditions other functions neither have the neces- for relatively easy digitalization. sary digital skills nor are they trained. How to improve Establish foundations for adoption and usage of digital tools—start collecting data in the level of digital digital formats, understand possible benefits from digital transformation for your firm, sophistication and secure basic digital skills among managers and key line workers. MSMEs at this level utilize basic technologies The company’s management does across all business functions, with no acquisi- not perceive digital transformation tion of technologies at index 3 or higher in any as applicable to the firm’s needs and Foundational part of their operations. All or most processes has no or rudimentary digital skills. in the firm are based on pen and paper, which Em-ployees exhibit gaps in digital does not allow quick adaptation and usage of literacy and are reluctant to introduce digital technologies. changes in their daily responsibilities. Source: Digitrans team. 2. Community Innovation Survey and Survey on ICT use and e-Commerce. 3. Results of the Technology Adoption Survey from 12 countries worldwide. 4. Based on the results of the 2021 ICT use and e-Commerce Survey, firms with very low or low digital intensity indexes report lack of knowledge as a barrier twice as often on average as firms with high and very high digital intensity indexes. 20 1 Various challenges impact each of the three groups of MSMEs to different degrees. Infrastructure   gaps, technology development gaps, access to finance constraints, regulations, and market conditions, among other challenges, impact technology adoption and usage. This report does not cover these types of constraints external to firms. They are broader issues for productivi-  Summary of lessons and promising strategies ty-enhancing strategies. Instead, the report focuses on the often-overlooked complementary capabilities internal to firms that condition the digital transformation process. And these constraints impact the target groups of firms differently because their stages of development and conditions for transitioning to digitalization are different. Based on an analysis of the literature, we have identified seven main challenges in comple- mentary capabilities that deter firms from adopting new technologies and scaling up the usage of known technologies. (See Figure 1.3) The list was developed by identifying in the literature evidence of non-technological aspects of MSME digital transformation that are under firm control and inhibit digital transformation (Cusolito 2021; Kohnke 2016). As a result, the list includes priority areas for public intervention that directly connect to well-grounded barriers to digitalization at the firm level. These 10 challenges fall into three groups according to the type of underlying market or system failure: (1) information constraints, (2) preferences and behavioral biases, and (3) organizational misalignment. FIGURE 1.3 Main challenges in complementary capabilities inhibiting digitalization of MSMEs By level of digital sophistication of MSMEs Informational Preferences and Organizational Challenge group constraints behavioral bias misalignment Challenges 1. 2. 3. 4. 5. 6. 7. Lack of Lack of Supplier Behavioral Worker re- Low digital Inadequate awareness knowledge market bias sistance and human managerial of possible on enhancing failures incentive capital practices technologi- productivity misalign- of workers cal upgrad- of digital in- ments ing vestments Higher prevalence of challenge by type of MSME: Foundational Level of Digital Sophistication Intermediate High Source: Digitrans team. Note: Size and Intensity of blue color circles indicates prevalence of a challenge among firms at a given level of digital sophistication. 21 1 The first group of challenges is related to informational constraints. Factors such as a lack   of knowledge about digital solutions and the resulting uncertainty about their returns under different conditions deter MSMEs from adopting and using productivity-enhancing digital technologies. Such challenges result from ineffective knowledge transmission channels and  Summary of lessons and promising strategies limitations on individual and social learning, which exacerbate adoption delays related to information (Kesley et al. 2015). The highest probability of adopting a technology occurs after many others have adopted it but before it has saturated the market (Bandiera & Rasul 2006). Such a delay can last many years and eventually lead to implementing technologies that have already become obsolete. A second group is associated with preferences and behavioral biases. Empirical evidence shows that MSMEs might not adopt technologies that bring high returns at a low cost, even when exec- utives know of their existence (Reynolds et al. 2021). This disconnect suggests that not entirely rational behavior may play a role in these decisions (Li et al. 2014). The present-preference bias, the status-quo bias, change-holding behavior, and risk and loss aversion can delay technology adoption decisions (Premkumar 2003).5 A recent study of 4,200 MSMEs in six countries revealed that behavioral biases such as resistance to change, uncertainty, present-preference bias, and avoidance of research due to a perceived abundance of information are consistently identified as the biggest inhibitors to adopt new digital technologies and increase usage of already known ones by respondents across various countries (Xero 2021). The mindsets that keep MSMEs from realizing the full benefits of digital technologies change slowly (Eggers 2020) but must be addressed to avoid shallow digitalization. Moreover, inappropriate internal incentives leading to principal-agent misalignments6 might limit technological upgrading (Atkin et al. 2017). One related problem, namely workers’ resistance to change, is also a well-known inhibitor of the adoption and usage of digital technologies by enterprises (Lazonick 1979). The third and last group of challenges relates to organizational misalignment. Inadequate man- agerial and organizational practices could inhibit investment processes related to acquiring new technologies or taking advantage of already adopted ones. Such managerial and orga- nizational practices include problems with coordination between managers responsible for different business functions in a firm, rigid organizational structure, and lack of change-man- agement skills (Rosenberg 1982; David 1990; Bresnahan and Trajtenberg 1995; Brynjolfsson and Milgrom 2013; Beaman et al. 2013; Brynjolfsson, Rock, and Syverson 2018; Juhasz, Squicciarini, and Voigtlander 2020). Lack of employee buy-in could arise either from incorrect perceptions of workers regarding the impact technology would bring or from actual threats to their salary or position in the company. In both cases, addressing gaps in human capital of workers would alleviate barriers inhibiting adoption and usage of digital technologies. Prevalence of challenges reflect also other firm characteristics than level of digital sophistica- tion. Even firms at similar levels of digital maturity have heterogenous gaps (Cirera, Comin, and Cruz 2022). The reasons for this differentiation are multifold. Firms are not equally digitalized across all business functions (World Bank 2022). Some gaps might be more pronounced among firms operating in particular sectors of the economy. For example, a non-negligible share of 5. The present-preference bias is a preference for a smaller reward now over a larger reward in the future. The status quo bias is a preference that things stay the same even if a change would bring a reward. Risk aversion is a preference for certainty over uncertainty. Loss aversion is a preference for avoiding losses over acquiring equivalent gains. Finally, change-holding behavior is refraining from changes that might bring beneficial changes at a given level of risk and resources. Psychologists and behavioral economists have shown that they all influence human decision-making. 6. Principal-agent misalignment is a conflict in priorities between the owner of a firm and a person to whom control of the asset has been delegated. An example of such misalignment is situation when a sales manager is tasked with implementing a CRM system that could enhance the monitoring of sales performance, putting more pressure on the manager to deliver results. 22 1 agriculture firms lacks access to the Internet. At the same time, numerous mechanical man-   ufacturing companies operate legacy systems and production processes incompatible with modern digital tools (Branca et al. 2020). Localization matters because firms in rural areas have less awareness and access to technology information than enterprises in urban centers, which  Summary of lessons and promising strategies tend to have higher exposure to digital solutions used by their competitors (Cirera et al. 2022). 23 1.2 MENU OF PROMISING INSTRUMENTS 1   BASED ON A REVIEW OF GLOBAL EVIDENCE  Summary of lessons and promising strategies KEY LESSONS FROM SECTION 1.2 This Section presents a catalogue of instruments for addressing gaps in complementary capabilities, organized according to groups of challenges and levels of firm digital sophistication. Design of instruments should be based on international best practices to increase the chances for achieving desired outcomes and streamline the design process. The lack of fundamental capabilities among potential beneficiaries reduces the effectiveness of more sophisticated interventions. The benefits of a staged approach are particularly high in the Europe catching-up context, where most representatives of a target group have limited digital capabilities. Policymakers need to consider the unique combination of digital capabilities of each target group to determine which set of instruments would be effective in their context. Bundling instruments can create synergies, prevent fragmentation and confusion among potential applicants, and tailor support packages to individual enterprises. Bundling instruments into programs can create synergies, prevent fragmentation and confusion among potential applicants, and tailor support packages to individual enterprises. The funneling approach can be used to gradually support firms in advanc- ing their digital capabilities, starting with general activities, and gradually becoming more focused and customized as companies progress. When selecting features for an intervention, it is important to analyze alternative instruments and consider complementarities, conflicts, or overlaps with existing programs. Linking the outcome of one instrument to another can reinforce the effects of various instruments. For each of the challenges identified, there is growing evidence of promising instruments to pur- sue.7 Chapters 2 and 3 describe them in full, with examples of specific programs implemented worldwide. While accounting for differences in local context and group of firms, the instruments to 7. In this document, instruments are referring to individual support activities which can be provided to firms to over- come challenges and achieve development objectives. A program is a specific set of instruments offered in combina- tion generate synergies in addressing interlinked challenges. 24 1 pursue should be based on international best practice — that is, interventions with proven impact   on alleviating digital challenges of MSMEs. Modeling interventions after features of instruments proven in practice increases the chances of effectiveness and streamlines the design process (Cirera et al. 2020). Figure 1.4 presents instruments identified from a review of global programs  Summary of lessons and promising strategies and policies. Each instrument provides a cost-efficient approach to alleviate market and system failures causing gaps in non-technological capabilities complementary to the digital transformation process. Given nascent stage of global programming in complementary capabilities for firm digita- lization, instruments differ in breadth of evidence supporting their impact. Some instruments are supported by evidence of rigorous impact evaluation in the context of adopting digital technologies by catching-up firms — highlighted on blue in Figue 1.4. Regarding other instruments, additional types of evidence were leveraged to identify their impact potential, including non-experimental impact evaluations, monitoring reports, self-reported information, and qualitative data. Informa- tion on evidence of impact of instruments in Figure 1.4 are presented in more detail in Annex 3. FIGURE 1.4 Typology of instruments for addressing gaps in complementary capabilities Challenge group Informational Preferences and Organizational constraints behavioral bias misalignment 1 2 7 Distance tech. Informational and Assistance in individual learning promotional access to finance campaigns 5 6 Small and Vocational training Foundational temporary for employees 3 4 8 subsidies Web-based Showcasing exam- Distance technology ples of successful managerial demonstration transformation learning MSMEs level of digital sophistication 9 13 Technology Managerial demonstration development 11 12 off-site programs Group tech. Lump-sum Intermediate learning payments for 10 14 workers Technological Marketplaces testbeds 15 18 Individual tech. Supplier learning development programs 16 17 19 Technology Experimental Individual High demonstration development managerial on-site programs learning 20 20 CTO extension CTO extension schemes schemes Type of instruments which impact was evidenced in rigorous evaluation studies Source: Digitrans team based on a review of global programs and instruments. Note: A full list of programs reviewed is presented in Annex 2. CTO = Chief Technology Officer. 25 1 The menu of instruments in Figure 1.4 is organized according to groups of challenges and levels   of firm sophistication to reflect conditions for policy decisions. Because each target group has a unique combination of digital capabilities, there is no set of instruments that would be effective in every context (Cirera et al. 2020). However, to facilitate categorization, we look at  Summary of lessons and promising strategies interventions individually by the type of firm and type of challenge they mainly address. Although firms at earlier stages of digital transformation could implement instruments intended for firms of high digital sophistication, they would be less applicable than instruments intended for firms with lower digital sophistication. A single program following a staged approach could offer in- struments of various levels of sophistication, providing more advanced support only after firms have benefited from more fundamental interventions. The benefits of the staged approach are particularly high in the catching-up context, where most representatives of a target group have limited digital capabilities (Cirera, Comin, and Cruz 2022). The lack of fundamental capabilities among potential beneficiaries reduces the effectiveness of more sophisticated interventions. Instruments to tackle informational constraints include providing timely and relevant knowl- edge, training, technology demonstration, and implementation assistance. This group of instruments involves designing effective awareness-raising strategies based on targeting specific nodes of a social network and using digital technologies to diffuse information about the advantage of using a technology (Beaman et al., 2021). Such interventions are necessary to reach firms at the very beginning of digital transformation; however, they could also target firms that have already adopted some digital technologies but use them on a limited scale. In both instances, the type of information that potential adopters look for to fill their knowledge gaps needs to be determined, together with an analysis of the most effective delivery method. Information sharing can affect adoption incentives through two opposite channels. In the positive “information-sharing” channel, more individuals adopt once others do because early adopters share their information with those interested in adopting. This channel counteracts the effects of the negative “information-free-rider” channel, which conditions firms to adopt a wait-and-see behavior to avoid the sunk cost of learning about the new technology (Bandiera and Rasul 2006). After firms’ interest in digitalization is piqued, they need knowledge of how to integrate various technologies to transform them into actual business decisions. Although on-site training might be necessary for more advanced technologies, knowledge about using more basic technologies could be delivered remotely (Shapira and Youtie 2014). Group training could improve the efficiency of support, not only by bringing costs down but also by facilitating peer-to-peer learning (ibid.). The second set of instruments seeks to address behavioral biases by demonstrating the ben- efits of digitalization for firms, showcasing successful peers, and offering resources for low- risk experimentation. The review of programs and instruments described in Chapter 3 shows that three groups of instruments can help address behavioral constraints limiting the speed of adoption and usage of MSMEs. First, small and temporary subsidies reduce procrastination and encourage MSMEs with present-biased preferences to adopt earlier (Duflo, Kremer, and Robinson 2011). Such instruments are also proven to effectively overcome employee reluctance to use technology (Atkin et al. 2017). Second, informational and promotional campaigns against technological and non-technological biases (Liu 2013) encourage early adoption for those with- out financial constraints and applications for external funding (for example, commercial loans) for those with financial constraints (Gine and Yang 2009). Third, demonstrating the usefulness of technologies for business through showcasing examples of successful transformation and providing technological testbeds — spaces for collecting, processing, and sharing high-frequency information on business performance — can induce behavioral changes in those interested in adopting (Beaman, Magruder, and Robinson 2014). Key aspects to consider during the design of the interventions in the preference and behavioral bias instruments include determining effective mechanisms for delivering the subsidy (for example, free or subsidized access to technology, 26 1 advisory, training, vouchers, or matching grants) and a cost-effective way of filling the infor-   mation gap. It is also important to leverage behavioral nudges, such as time-limited subsidies or risk-sharing plans, to implement technology adoption (Duflo, Kremer, and Robinson 2011).  Summary of lessons and promising strategies There are also promising strategies to address organizational misalignment in adopting and using digital tools. Adoption of technologies and taking advantage of them requires firms to upgrade managerial practices and improve the digital skills of managers and workers. The evidence shows that planning and monitoring capabilities, as well as incentives for workers to use digital technologies, developed as a result of managerial upgrading, help facilitate MSME digitalization (Verhoef et al. 2021). Effective interventions to upgrade managerial skills include providing free or co-financed consulting on management practices at various forms — including distance and face-to-face settings (Bloom et al. 2013). At the instrument design stage, policy makers need to explore cost-effective incentives, such as lump-sum transfers, to foster the adoption and usage of new technologies. Because most instru- BOX 1.2 Examples of multi-instrument ments for addressing organizational misalignment programs include providing managerial training or advisory, the Chile — Digitalize du Pyme | Instruments: 4, 7, and 12 crucial aspect conditioning impact is the quality of This program aims to incentivize SMEs to adopt the providers who directly interact with beneficiaries. basic digital solutions through a mix of market- As with technological training, upgrading managerial place and web-based technology demonstration. capabilities in a group setting (whether at a distance It focuses on the touristic sector and addresses or in person) could both lower cost and enhance informational constraints related to a lack of knowledge among touristic firms about suppliers development impact (Iacovone, et al. 2022). of technological services to digitalize key business processes such as customer management and But while research has illuminated the evidence of internet visibility (for example, SOE optimization or impact for some instruments, many of them remain the use of global booking platforms). unexplored, especially within the context of digitali- zation programs. Interventions such as informational Singapore — Digital Leaders Programme | Instru- ments: 15, 16, and 19 campaigns, subsidies and transfers and managerial ad- vice have been examined through randomized control Digital Leaders Programme, a Singaporean public intervention for digitally sophisticated firms, com- trials, which have yielded rigorous estimates of their bines face-to-face individual technology learning, impacts. However, these evaluations have been done a CTO extension scheme, and on-site technology in contexts very different from ours, of digital adoption demonstration. The program helps companies that in small and medium enterprises (they are often done are ready to integrate digital technology into their in the context of agricultural microenterprises). Fur- core business strategies build new digital capabil- ities so that they can develop innovative business thermore, other instruments, such as technological models and capture new growth opportunities. testbeds and on-the-job training of employees for technological adoption, have not for the most part Germany — SiemensX Accelerator | Instruments: 9, been the subject of evaluations regardless of context. 10, 11, 12, and 14 Programs operated by private institutions take The multiple instruments can be implemented and a similar approach. This accelerator provides a managed together to form programs and leverage portfolio of engineering software, services, and an application development platform for enhancing synergies between them, as is often done in most de- the adoption and usage of medium-to-advanced veloped support ecosystems. Bundling instruments digital solutions by manufacturing firms. It focuses together (Box 1.2) prevents excessive fragmentation. on enhancing electronic and mechanical design, Individual interventions may be dispersed in reach, system simulation, manufacturing, operations, and drive up administrative costs when combined, and lifecycle analytics. can create confusion with potential applicants (World Source: Digitrans team based on analysis of global Bank 2019). Combining instruments is also useful programs and instruments — details in Chapter 3. for embedding variants within a single intervention, Note: CTO = Chief Technology Officer; SME = small and medium enterprise; SOE = state-owned enterprise. which allows for more closely tailoring the support package to the needs of individual enterprises. 27 1 Furthermore, instruments supporting the upgrade of a given type of capability can use a funnel-   ing approach. At the initial stages of cooperation with a beneficiary firm, support activities may be general and aim to take a wide offering suitable to a broader audience (Digital Innovation Hubs Working Group 2018). For this reason, the policy menu may include instruments allowing  Summary of lessons and promising strategies for supporting firms in gradually advancing their capabilities (Figure 1.5). As companies move through the funnel, the instruments should be more focused and customized to stay relevant to evolving firm needs. Policy makers can also leverage the funnel approach to generate fur- ther interest in creating digital technologies (beyond adopting and using them) by connecting enterprises with institutions offering more advanced programs — including, for example, digital innovation hubs (DIHs) or research and development agencies. FIGURE 1.5 Funnels in the typology of instruments Level of sophistication Low High 1 11 15 20 Technological Distance technology Group technology Individual technology CTO extension knowledge individua learning learning learning schemes 3 9 16 20 Technology Web-based technology Technology Technology CTO extension demonstration demonstration demonstration off-site demonstration on-site schemes 4 14 17 Adaptation of technologies Showcasing examples of Technological Experimental successful transformation testbeds development programs to firm needs 8 13 15 Managerial Distance managerial Managerial development Individua technology training learning programs learning 12 5 Internal Lump-sum payments Small and temporary incentives for workers subsidies Source: Digitrans team. Note: CTO = Chief Technology Officer. Selecting the features for an intervention needs to be preceded by an analysis of alternative instruments and the intervention’s relation to other public programs and services offered on the private market. Policies do not have effects in isolation from other policies nor from many other contextual factors. Consideration should be given to complementarities, conflicts, or overlaps between the proposed instrument and existing instruments. Linking the outcome of a single instrument with an opportunity to follow-up to continue improvement in another instrument could reinforce the effects of various instruments. For instance, the 2021 – 2027 perspective explicitly considers the importance of instruments related to (European) DIHs. Additionally, the organizational culture of the agencies and the political culture may limit the range of instruments optimal for implementation in each institutional circumstance, which might require considering alternatives. 28 1.3 LESSONS LEARNED 1   FROM GLOBAL REVIEW OF PROGRAMS  Summary of lessons and promising strategies KEY LESSONS FROM SECTION 1.3 The section presents eight lessons derived from an analysis of programs and instruments implemented in over a dozen countries across four continents, as well as a review of evaluation studies. The first three cross-cutting lessons address implementation challenges that are relevant not only to individual interventions but also public institutions managing interventions: • To address the scarcity of proven approaches, interventions should be carefully designed to target complementary capabilities. • The impact of programs enhancing these capabilities can be amplified by combining online and offline methods. • To secure buy-in from the target population, it’s essential to establish trust. The following five lessons focus on specific target groups or particular challenges and provide evidence-based recommendations that can be used during the design and implementation of support instruments: • Programs supporting MSME digitalization should include diagnostic components. • Overconfidence in firms’ current digital capabilities must be addressed, especially at the foundational level of digital sophistication. • Customized interventions should target transformational and advanced enterprises. • Market-based interventions should be prioritized for older and established firms. • Group interventions can be explored to drive impact, cost efficiency, and adaptability. This section presents eight lessons that effectively bridge gaps in complementary capabilities for digitalization. These lessons have been derived from an analysis of programs and instru- ments implemented in over a dozen countries across four continents, as well as a review of literature and impact evaluation studies on effective interventions supporting MSMEs in the adoption and usage of digital technologies. The section begins with three cross-cutting lessons 29 1 that discuss general themes applicable to all public interventions aimed at enhancing firm   digitalization through managerial upgrading. The first three lessons address implementation challenges that are relevant not only to individual instruments but refer more broadly to inter- ventions aiming for advancing MSME digitalization . The following five lessons focus on specific  Summary of lessons and promising strategies target groups or particular challenges and provide evidence-based recommendations that can be used during the design and implementation of support instruments (Table 1.1). These recommendations are designed to be operationalized to ensure effective implementation. TABLE 1.1 Lessons learned for firm digitalization and managerial capability building Cross-cutting themes Specific lessons Lesson #1: Carefully design interventions targeting comple- Lesson #4: Include diagnostic components in programs mentary capabilities given scarcity of proven approaches supporting MSME digitalization Lesson #2: Combine online and offline methods to amplify Lesson #5: Address overconfidence preventing technological impact of programs enhancing complementary capabilities upgrading of firms at foundational level of digital sophistication Lesson #3: Establish trust to secure buy-in from the target Lesson #6: Customize interventions targeting firms at population intermediate and high level of digital sophistication Lesson #7: Prioritize market-based interventions for older and more established firms Lesson #8: Explore group interventions to drive impact, cost efficiency, and adaptability Source: Digitrans team. The primary source of data for the lessons was documentation of over 70 diagnostic tools and 40 programs, which included monitoring reports and research revealing their impact on the adop- tion and usage of digital technologies by MSMEs. The reviewed tools and programs were also analyzed against well-documented best practices in the design, implementation, and governance of firm-level support instruments (World Bank 2020b). The secondary source of data was scien- tific literature on interventions for addressing market and system failures conditioning non-fi- nancial aspects of technology adoption. The rigorous research was systematically reviewed to identify features proven to enhance development outcomes. Where applicable, the review was extended to literature discussing the digitalization of the private sector in a broader context. Cross-cutting themes LESSON #1 Carefully design interventions targeting complementary capabilities given scarcity of proven approaches There are significant knowledge gaps in understanding the success of instruments that alleviate firm constraints in complementary capabilities for digitalization. This can be attributed to two primary factors. Firstly, many challenges, such as worker resistance to change or thin supplier markets in the context of MSME digitalization, have been largely unaddressed in academic literature and policy circles. (World Bank 2021 and World Bank 2022b). Policymakers have pri- marily focused on streamlining access to finance for purchasing technology, leaving support for complementary capabilities relatively underdeveloped. Secondly, impact of instruments addressing most remaining challenges in complementary capabilities is only partially docu- mented, as such interventions have been mostly created in the past few years. This timeframe is insufficient to fully evaluate the outcomes of such interventions. Considering the nascent state of policies supporting firm digitalization in the EU and the gaps in existing policies, an 30 1 experimental approach is required from policymakers. This involves designing and tracking the   impact of interventions aimed at addressing non-technological challenges for firm digitalization, with a strong focus on monitoring and evaluation practices.  Summary of lessons and promising strategies The instruments supporting firm digitalization in EU are less sophisticated than policies sup- porting business innovation and display similar gaps as innovation policies in the past. However, much like the shift in perception of business innovation from secondary importance to a core component for maintaining competitiveness, digital transformation is now recognized as a fun- damental requirement across business functions (Urbach, Drews, and Ross 2017). Consequently, policy-making is evolving towards greater maturity (Figure 1.6). To create effective support instruments FIGURE 1.6 Evolution of approach to support digital despite the lack of academic and applied research, and innovation policy in the EU as well as previous policy experiments, policymakers Innovation Digitalization can draw on evidence from analogous programs in policy policy non-digital domains. For instance, to address the limited knowledge on optimizing digital technology ◦ Horizontal support ◦ Horizontal financial usage among firms, a standard business advisory for research & devel- support for purchas- Generation 1.0 opment works ing technology service could be developed that includes a specific ◦ Focused on financial ◦ Establishment of instrument for diagnosing and resolving incentive aspects clusters misalignment issues related to digital technology ◦ Differentiation be- ◦ Focus on non-finan- adoption. This program could be combined with an- tween various stages cial technological as- other offering lump-sum payments to workers who of innovation (TRLs) pects and related pro- exhibit positive engagement with the technology, by Generation 2.0 cesses (e.g., busi- ◦ Accounting for tech- nology heterogeni- ness-academia co- examining the applicability of lessons learned from laboration, testing) ty (Digital Innovation Hubs) successful interventions in non-digital technologies, such as the case of football manufacturing in Pakistan, ◦ Advanced targeting ◦ Integrated support (Smart Specializa- for non-technologi- and applying them to digitalization (Atkin et al. 2017). tion Strategies) using cal factors in digitali- Generation 3.0 frontier methodoo- zation gy (Entrepreneurial The COVID-19 pandemic has led to an increase in the Discovery Process) ◦ Coordinated support for firms at all stages supply of public interventions aimed at supporting of digital maturity firm digitalization; however, evidence of impact of these interventions remains limited. Despite the stated Source: Digitrans team. objective of supporting digital transformation, many Note: TRL = technology readiness level. government interventions have been uncoordinated and have focused on financial support (Avalos et al. 2023). The fragmented approach to interven- tions has resulted in overlooking the critical challenges in complementary capabilities, which pose significant obstacles to digitization, particularly for less digitally sophisticated firms. Therefore, there is a pressing need for more comprehensive and coordinated public interventions that build on the evolving understanding of policymaking in the digital domain. These interventions should be designed to address specific barriers to digitalization faced by different types of firms, as op- posed to the horizontal interventions that were rapidly implemented in response to the pandemic. Additionally, the interventions should aim to foster the knowledge and capabilities necessary for sustainable digital transformation, both during and beyond the pandemic. LESSON #2 Combine online and offline methods to amplify impact of programs enhancing complementary capabilities Policy design should both online and offline channels when building awareness of digitalization among target groups. Online strategies can target specific nodes in social networks (BenYishay and Mobarak 2019; Munshi 2004) and use digital technologies to promote the diffusion of in- formation about technology advantages (Beaman et al. 2021). Offline strategies can improve 31 1 knowledge by offering consultations on management practices (Bloom et al. 2013), providing   in-person training workshops, establishing mentorship programs, and facilitating networking events where entrepreneurs can learn from industry experts and peers. In both channels, policy design must identify the type of information firms seek and adopt the most effective approach  Summary of lessons and promising strategies to fill their knowledge gaps. Information sharing can influence adoption incentives through two contrasting channels: positive “information-sharing” channels, where adoption increases as early adopters share information with interested parties, and negative “information-free-rider” channels, where firms adopt a wait-and-see behavior to avoid sunk costs of learning about new technology (Bandiera and Rasul 2006). Policy design should capitalize on the positive channels while counteracting the negative ones. One-stop-shops stand as a prime approach for communicating with firms about opportunities related to digitalization, while fostering coordination among various public authorities involved in digital ecosystem. Engaging entrepreneurs and firm managers hinges on the quality and rele- vance of content provided by these platforms. It is essential to keep them updated, technically accurate, and easily accessible to users with diverse levels of prior knowledge. The content should cater to different learning preferences, incorporating a range of formats such as case studies, how-to guides, training materials, webinars, and interactive tools. To encourage firms to take action, the knowledge shared should be augmented with practical services, including matchmaking with technology providers, connecting with experts within the Digital Innova- tion Hub (DIH) network, integrating self-assessment tools, or facilitating access to financial resources and grants. Examples of successful online one-stop-shops that promote MSME digitalization, showcasing these best practices, are available in Box 1.3. BOX 1.3 Examples of online one-stop-shop information platforms New Zealand’s Digital Boost website offers free-to-ac- cal offices specializing in providing specific advisory cess video tutorials on various aspects of transferring services, with costs cofinanced from the state budget. business operations online, with supplemental learning The Accelera Pyme is part of the SME Digitization Plan materials and tutorial one-on-one sessions with profes- 2021 – 2025 managed by the Ministry of Economic Affairs sional consultants. Service provision is centralized with and Digital Transformation. one organization — The Mind Lab — a top graduate edu- France’s France Num offers an educational platform for cation institute in the country. The Ministry of Business, SMEs with digital learning courses that provide practical Innovation, and Employment operates the platform. knowledge on adopting and scaling the use of digital In Spain, Accelera Pyme is an Internet platform with technologies and connects with a scheme of preferen- effective tools for assessing firms’ technological tial credits for acquiring the necessary equipment. The maturity and selecting training courses to advance the financial product addresses basic technological needs digitalization of enterprises at various levels. Based with loans up to €50,000 from six commercial banks. on the results of assessment and training, the platform It aims to support technological upgrades for 50,000 forwards enterprises to a network of over 20 physi- enterprises by 2023. Source: Digitrans team analysis of program documentation. LESSON #3 Establish trust to secure buy-in from the target population Firms are more likely to be receptive to messages from trusted parties and to act upon con- sistent messages from multiple sources within their network. In the context of firms at lower levels of digital sophistication, confirmation from various sources is crucial before commit- ting to technology upgrading (Beaman et al. 2018). Repeated messaging is also essential for motivating firms to implement business changes that enable deeper integration and scaling 32 1 of digital technologies (Kane et al. 2015). Effective public support instruments should not take   trust for granted and need to incorporate features that build trust among business owners and managers. Successful government-funded programs should act primarily as independent brokers between MSMEs and experts, rather than competing with private-sector providers. This  Summary of lessons and promising strategies brokerage role lowers barriers to entry for firms by creating a safe space for interaction with external entities, protecting less sophisticated enterprises from the perceived risk of abuse by external advisors (Schizas, Jarvis, and Daskalakis 2012). Impactful programs should integrate trust-building elements into their intervention design. Program managers can enhance trust in interventions by establishing meaningful partnerships, cultivating a strong brand image, and supervising quality of providers that directly interact with beneficiaries. Collaborating closely with leading industry associations is often an effective starting point for collecting and curating detailed provider information, which can then be assembled and disseminated in a user-friendly manner. A strong brand that conveys reliability and clear benefits to firms is essential for any program aiming to assist companies with their digitalization processes. Thus, program officers must invest in building a solid reputation and robust name recognition. Furthermore, many firms are hesitant to trust external technology vendors or advisors (Williamson 1985), and MSMEs may have had negative past experiences or lack confidence in the value of support provided by external parties (Robson and Bennett 2000). By embedding certification mechanisms in public interventions, such concerns can be effectively addressed. Specific lessons LESSON #4 Include diagnostic components in programs supporting MSME digitalization Evidence Digital transformation plans are crucial for avoiding shallow digitalization and facilitating effec- tive digitalization of businesses. These plans provide a comprehensive roadmap, supported by owners and managers who can manage change, focusing on the successful integration of ICT in business operations. Research by Santos-Pereira et al. (2022) shows that isolated actions, without a well-thought-out plan, do not have a significant impact and result in shallow digitali- zation. The most common type of public intervention in EU Member States, which concentrates on financing the acquisition of IT software or equipment with minimal support for integrating digital solutions into managerial processes (Strilets et al. 2022), leads to lower productivity enhancements compared to policies that focus on upgrading technical and managerial skills (Sorbe et FIGURE 1.7 Areas covered by successful al. 2019). Combining training or advisory with manda- transformation plans tory implementation, while demonstrating practical Technological Complementary aspects of complementary capabilities, allows for factors capabilities lasting change in managers’ and workers’ attitudes Smart Integrated Value Value solution communi- generation appro- in firms (Azevedo and Almeida 2021). cations priation Hardware Networks Application Revenue Successful digital transformation plans integrate model Software Connection Use case both technological factors and complementary Competitive capabilities. These plans simplify the process by barriers linking internal and external activities and resources to adopt or enhance the successful usage of digital Source: Adapted from Lichtenhaler 2019. 33 1 technologies. They allow firms to map their ambitions, assess their current positions, and   navigate the digitalization process (Figure 1.7). Firms can greatly benefit from a dedicated digital transformation plan tailored to their needs, guiding necessary actions for effective digitalization. Digital transformation plans help avoid one-off changes that do not lead to pro-  Summary of lessons and promising strategies ductivity improvements. For example, most voucher schemes are small in scale and focus on incentivizing particular firm behavior (Cirera et al. 2020). Although effective in attracting small firms to a specific policy area, vouchers rarely lead to the materialization of intended changes without follow-up (Valbonesi and Biagi 2016). Digital transformation plans mitigate the risk that collaboration will not be sustained. The importance of a digital transformation plan is even higher among firms that have already adopted some digital technologies. Firms report more impediments while increasing the usage of known technologies than at the initial stages of adoption (Baldwin 2002). Digital transformation plans prepare firms to address these barriers in advance, minimizing the risk of shallow digitalization where adopters do not harness the productivity potential of technologies. Recommendation Impactful digital transformation plans share several common operational characteristics. They include a comprehensive set of necessary actions, list resources, clearly indicate the division of responsibilities between various parties involved in the process and allow for monitoring progress over time (Zaoui and Souissi 2020). These plans should be created top-down, as digital transformation involves numerous business functions and cannot effectively originate as a grassroots activity (MIT Sloan 2017). Digital transformation plans should schedule the im- plementation and usage of multiple technologies, even if the public intervention under which the plan was delivered covers only a single technology and show public intervention as part of a broader transformation. Adopting multiple technologies at once can lead to up to three times higher productivity increases (Abdrazakova and Salihudin 2022). The approach used for diagnosing firm needs and guiding implementation should allow for tracking the impact of support activities on the firm and the impact and relevance of the diagnosis (OECD 2019). Moreover, updating benchmarking data used for identifying needs that plan aims to address is essential, especially after the COVID-19 pandemic, as enterprises accelerated ICT trans- formation during this period. To create an effective digital transformation plan, policymakers should prioritize approaches that encompass the four phases outlined below. The first phase involves preparing a decision matrix for firms to visualize technology options and prioritize those that best align with their objectives. The decision matrix should be flexible to adapt to business priorities and the expe- rience of decision-makers with various levels of digital skills. The second phase is to perform an ex-ante assessment using the matrix to clarify existing risks. Change often feels risky, but business owners and managers frequently overlook the inherent risk of maintaining the status quo. Evaluating the current state can reveal hidden threats and help avoid the risks of intro- ducing new technology. The third phase includes conducting a rational cost-benefit analysis. Decision-makers in firms tend to overestimate risks related to adopting new technologies (Foster and Rosenzweig 2010). An objective approach to costs and benefits helps overcome resistance to change by detailing precisely what is at stake. Lastly, the fourth phase involves setting aside time for learning. Presenting relevant knowledge, especially the introduction of benchmarking data to discover how other businesses are using technology and benefiting from the process, is crucial to secure commitment (Lee 2014). 34 LESSON #5 1 Address overconfidence preventing technological upgrad-   ing of firms at foundational level of digital sophistication Evidence  Summary of lessons and promising strategies Overconfidence is a prevalent issue among catching-up firms, which can hinder their progress towards digital transformation. Firms most in need of technological upgrading may be among the most reluctant to pursue it due to overconfidence and self-assessment bias. The vast ma- jority of firms at the foundational or intermediate level of digital sophistication tend to position themselves above the median compared to the level of technology adoption among their competitors, regardless of their actual technologi- cal advancement. This gap between perceived and FIGURE 1.8 Self-assessment with respect to other actual levels of advancement is the largest for the firms in Poland least technologically sophisticated companies (Fig- 5 ure 1.8). Many firms are not aware of their own needs Self-assesment of technological or the technological and managerial opportunities available for improving competitiveness, meaning 4 sophistication they “do not know what they do not know” (Cirera et al. 2022). Conversely, firms at higher levels of digital 3 maturity tend to underestimate their technological sophistication, which could result in a reluctance 2 to adopt and use more advanced solutions due to a perceived lack of capabilities. 1 Given the high prevalence of unawareness among 1 2 3 4 5 firms, support programs should include external and Firm position in the distribution of intensive GBF index in Poland objective needs assessments. Asking firms to prepare applications with consultants does not resolve the lack Source: World Bank 2022. of awareness about needs among firms because con- Note: On a scale from 1 to 5, where 5 means that the establishment is sultants may have vested interests that do not always using the most advanced production processes available in its sector, where do you think this establishment stands with respect to other align with the best interests of the MSMEs (Lambrecht firms in Poland? GBF = General Business Functions. and Pirnay 2015). An independent third party with up-to- date technical knowledge and familiarity with the digital solutions market is necessary to assess a firm’s needs independently. Almost all of the reviewed best practice programs account for the awareness gaps described above and provide independent assistance in assessing needs. The methodology for needs assessment must also be technology-neutral and screened for explicit or implicit preferences toward particular types of digital solutions or providers. Although private enterprises have created some of the most impactful needs assessment methodologies, they need to be screened for technological neutrality to ensure fairness and objectivity. Recommendation To effectively address the digitalization needs of firms, it is crucial to utilize a variety of diagnostic tools to identify relevant needs of target groups. Surveys and Digital Diagnostic Tools (DDTs) serve different functions but complement each other in identifying these needs. Surveys are descriptive, providing representative data on the aggregate level to inform policymakers looking to understand drivers and barriers to the digitalization of firms, or identify specific target groups among MSMEs in need of support. DDTs, on the other hand, offer a more granular approach, allowing for the identification of digitalization barriers and drivers at the individual firm level. The interplay between these two types of instruments is valuable for accelerating digitalization efforts at scale. Survey results can predominantly be used to decide on the area of intervention 35 1 and characteristics of the target group, while insights from DDTs can guide the selection of bene-   ficiaries from the target population and tailor support delivered to them. Although surveys require fewer resources than DDTs, both from policymakers and firms, they generate fewer insights. The results of surveys and DDTs provide the basis for an improvement plan with specific steps  Summary of lessons and promising strategies and priorities, such as identifying and prioritizing technology or managerial practices to adopt. However, no diagnostic methodology, whether aggregate or firm level, provides combined in- sights on technological and complementary capabilities at a highly granular level. While tools for generating granular technology and managerial insights exist and have been well-tested in practice, they have not been deployed in tandem. Furthermore, most of the identified best practice tools were implemented outside the European Union, which may limit their relevance in that context. To address existing gaps in evidence for policymaking, a dedicated diagnostic tool leveraging the best methodologies for generating insights on technological and comple- mentary factors is necessary (Table 1.2). In each context, a decision must be made to optimize the use of resources for addressing the most crucial knowledge gaps. Support programs should include external and objective needs assessments to address the lack of awareness about needs among firms. An independent third party, equipped with up-to-date technical knowledge and familiarity with the digital solutions market for firms, is necessary to assess a firm’s needs independently. Utilizing a variety of diagnostic tools, such as surveys and DDTs, can provide a comprehensive understanding of the digitalization needs of firms, enabling well-informed decisions on interventions and tailored support for the target groups. TABLE 1.2 Best-practice tools for generating evidence to inform instrument design Scale of analysis Information type Aggregate level Firm level Firm-level Adoption of Technology Survey (FAT) Digital Maturity Assessment (DMA) Technological Community Survey on ICT Usage and e-Commerce in Enterprises (CSI) Digital Readiness Assessment Tool (DRAT) Management and Organizational Practices Managerial Survey (MOPS) InnoCape DMA In-depth interviews following Technology Behavioral Acceptance Model Source: Digitrans team LESSON #6 Customize interventions targeting firms at intermediate and high level of digital sophistication Evidence Young and more isolated firms encounter substantial challenges in developing the complemen- tary capabilities required for successful digital transformation. Firms with less than 3 years of business operations typically struggle to access information about the digital technology possibilities available in the market. This constraint becomes even more pronounced for young MSMEs situated in less densely populated regions (Cirera et al. 2022). The process of accumulating the necessary knowledge to build complementary capabilities for digitalization is time-consuming and complex. It relies heavily on accessing formal information channels such as informational campaigns, on-the-job training, school, and university education (Kane et al. 2017). Additionally, exposure to the managerial or technological practices employed by 36 1 industry peers plays a crucial role in shaping a firm’s understanding of digital transformation.   Inexperience and isolation often result in limited opportunities for collaboration, network- ing, and knowledge exchange, further exacerbating the gaps in complementary capabilities. These firms may struggle to establish connections with industry partners, suppliers, and other  Summary of lessons and promising strategies stakeholders, hindering their ability to identify best practices and adopt innovative solutions. Furthermore, the workforce in these young and isolated firms may have lower digital skills, limiting their ability to embrace and adapt to new technologies. These constraints lead less-mature digital businesses to focus on solving discrete business problems with individual technologies, rather than perceiving digitalization as a more complex, cross-functional process. Moreover, given their lower experience in dealing with suppliers and business deals in general, newer firms tend to be more reluctant to engage with digital suppliers and have fewer opportunities and resources to familiarize themselves with the digital supplier landscape (Jay and Schaper 2003). Similarly, employees in younger and smaller firms generally have lower digital skills and may be less accommodating to changes (World Bank 2020a). Targeted interventions can help small and isolated firms overcome hurdles in digital transformation, facilitating adoption and simplifying decision-making. By addressing the spe- cific challenges faced by these firms in developing the complementary capabilities required for digital transformation, such interventions can shift their focus from solving individual busi- ness problems using specific technologies to embracing digitalization as a comprehensive, cross-functional process. Recommendation Policymakers should design targeted interventions focusing on addressing the unique chal- lenges faced by firms to drive significant impact. Striking a balance between customization and cost-effectiveness is crucial when offering standardized yet adaptable programs that cater to the needs of individual businesses. These programs should simplify decision-making, providing clear and well-informed options that inspire confidence and facilitate adoption. Customizing interventions, even at the cost of scale, leads to a more significant impact. Early adopters and laggards exhibit different technical, organizational, and knowledge capabilities, which means that support for firms at the fundamental level of digital sophistication should differ from instruments stimulating the diffusion of frontier solutions. As firms at the same level of digital sophistication also have varying needs, adaptability at the program level is essential to main- tain intervention relevance in light of firm-level transformation plans. Successful instruments that drive the adoption of less advanced technologies can overcome constraints by offering adaptable support, leveraging widespread and popular technologies. Policymakers must determine the right balance between providing customization possibilities while maintaining cost-effectiveness and manageability. Evaluation studies of programs supporting technology adoption among MSMEs reveal that moderately customized services yield more substantial benefits, while routine services focus- ing on quality and process improvement achieve less significant results (Cirera et al. 2020). Policymakers should remain aware of the ongoing tension between expanding coverage and delivering tailored services (Shapira, Youtie, and Kay 2011). Given the limited capabilities of catching-up firms, streamlining, and simplifying choices can facilitate the adoption process and minimize the need for customization. Overloading MSME managers with excessive infor- mation may impede their decision-making processes (Kilimis, Zou, Lehmann, and Berger 2019). Policymakers should instead present well-defined, well-informed options that instill confidence in small businesses to act decisively, rather than creating confusion. 37 1   BOX 1.4 Examples of customizable public instruments supporting MSME digitalization The United Kingdom’s Digital Manufacturing on a Shoe- related to building e-commerce capabilities, including string provides a technology extension program that developing websites, applications, or digital payment  Summary of lessons and promising strategies allows SMEs to test affordable, easy-to-implement tech- infrastructure. The support is conditional on participation nologies as small-scale digital solutions that will not dis- in centrally managed informational sessions that aim to rupt their core business operations but will provide imme- equip beneficiaries with the capabilities necessary to diate benefits. It focuses on manufacturing enterprises adopt e-commerce technologies successfully. Success- and follows a three-step process with dedicated experts ful applicants can apply for a second voucher to further delivering ongoing support. First, the digital needs of develop the solutions financed from the initial grant. SMEs are analyzed to determine their three most critical Singapore’s Technical Advisory Programme provides business areas for potential improvement. Second, an firms with advisory services from a certified group of appropriate technical solution is designed to address technology experts to prepare a technology upgrading those needs using only easily approachable building strategy, including the knowledge and competence blocks; a precisely defined purchase specification is then necessary to adopt and use new solutions successfully. suggested. Third, appropriate providers are identified Beneficiaries of the program could receive four hours through the Shoestring portal, which aggregates quality of support per month for up to two years from various technology suppliers. The program is operated by the consultants selected based on needs identified on an Institute for Manufacturing at Cambridge University and ongoing basis. While in the program, SMEs can test has supported over 300 SMEs as of Q2 2022. different technological solutions and discuss R&D In Ireland, the Trading Online Voucher program uses opportunities with researchers. After completing the vouchers to incentivize SMEs to start offering products upgrade, enterprises are offered preferential access to and services online. It is a competitive process in which financial instruments, including equipment and comple- micro firms can receive up to €2,500 to cover expenditures mentary costs. Source: World Bank staff analysis of program documentation. LESSON #7 Prioritize market-based interventions for older and more established firms Evidence Older and more established firms often exhibit a higher resistance to external interventions and a greater degree of inertia in their work practices. This resistance can make it challenging for these firms to modify existing complementary practices, which are essential for advancing their level of digitalization. This reluctance to change is frequently rooted in a shift in growth orientation over time (Gray 2002). Initially, entrepreneurs and key managers in MSMEs are more open to change shortly after establishing a business, as they actively seek to develop a profitable business model and adapt to the market demands. However, as their businesses mature, stabilize, and achieve a certain level of success, their focus often shifts towards maintaining the status quo and preserving the established business model. This shift in priorities can lead to a decreased appetite for risk and innovation, potentially hindering their ability to embrace new digital technologies and opportunities for growth. Compared to the average MSME, firms with positive performance may exhibit a heightened focus on maintaining the status quo in how their business processes are organized and managed. As these businesses achieve greater success and profitability, they often develop increased confidence in their established business model, leading to a diminished sense of urgency for change and innovation. This reluctance to embrace change might be even more pronounced in firms with high profits, due to the so-called “fat syndrome” (Osiyevskyy and Dewald 2014), which can foster complacency and hinder growth. Conversely, when firms are underperforming or facing existential challenges, their managers are more likely to adopt a risk-taking mindset and explore new opportunities (March and Shapira 1987). As a result, programs aiming to stimulate more active and innovative behavior from firms should prioritize those with relatively positive performance, as they may be more receptive to support and guidance. 38 1 Recommendation   Experienced and established firms tend to favor interventions of moderate scale. To effectively influence their digitization strategy, interventions must be designed and communicated in a  Summary of lessons and promising strategies way that appeals to their expertise and experience. Therefore, moderate interventions that are thoughtfully designed, strategically communicated, and appropriately scaled can increase their implementation feasibility and development impact. Small, incremental changes that bring quick and measurable benefits encourage firms to participate in public support programs, as it resembles the investment processes undertaken in typical business practice (Eller et al., 2020). By framing technology adoption and usage in terms of small, manageable steps rather than high- cost, high-risk large investments, entrepreneurs are more likely to digitalize. Furthermore, direct demonstration of the relevant and tangible impact on business operations can help to dispel uncertainty and correct misconceptions about what it means to go digital among MSMEs (Cirera et al., 2020). For catching-up firms and simple technologies, small, early subsidies have been shown to induce a larger adoption response than large subsidies for more complicated processes, which is consistent with present biasedness (Duflo, Kremer, and Robinson, 2011). Therefore, the appropriate intervention, given a proper diagnosis of change-holding behavior as a constraint on adoption, is much less costly and entails much lower distortions than traditional heavy subsidies. Moderate interventions are well-suited to leverage market mechanisms and minimize the risk of distorting market signals, which can lead to economic inefficiencies. Such interventions are less likely to disrupt market dynamics and have a lower likelihood of yielding unintended outcomes. By designing instruments that leverage market mechanisms, policymakers can effectively address challenges in firm digitalization in a way that does not distort market signals (OECD, 2018). Howev- er, even a light intervention, such as the creation of an e-portal to connect MSMEs and suppliers, requires constant follow-up and updating to avoid becoming deserted and obsolete. Directly engaging in matchmaking efforts can overcome a lack of results but requires strong public sector capabilities or a budget to hire them in the market and entails the possibility of government failures. This approach also places the government in the position of a supplier of a service traded in a free market (Office of Fair Trading, 2009). Therefore, it is essential to avoid the risk of government failure, which can lead to wasted resources and market distortions. Program design should focus on supporting private sector initiatives and calibrating the degree of intervention in the market, weighing the importance of market failures against the possibilities of government failure. LESSON #8 Explore group interventions to drive impact, cost efficiency, and adaptability Evidence Research indicates that incorporating features that foster collaboration between firms in in- strument design can enhance their effectiveness and cost-efficiency. Such instrument design that promotes cooperation among firms can be utilized in designing subsidy and education interventions. In particular, group-based consulting have been found to be effective in increasing the adoption of new technologies among small and medium-sized firms (Iacovone, Maloney, and McKenzie 2022). Moreover, community engagement components can serve as a means of attracting more firms to participate in support programs and facilitate scaling-up interventions. Managerial training delivered through group-based consulting has been found to yield higher effects with limited costs, especially for catching-up firms. Another effective approach is signaling effects, where firms that engage in the creation of prototypes and pilots for IT adoption, and demonstrate steady progress, can incentivize their peers to take similar chances in adopting new technologies (Srinivasan, Lillen, and Rangaswamy 2002). 39 1 Recommendation   The utilization of role models and group-based approaches has been found to accelerate adoption decisions among non-adopters, but this approach necessitates robust community  Summary of lessons and promising strategies management capabilities. Experiences from multiple countries have demonstrated that dis- seminating knowledge about successful organizations or individuals who have implemented a particular digital technology can accelerate adoption decisions among non-adopters (Friederici 2022). Policymakers have explored various forms of group-based approaches in delivering support, such as the exchange of knowledge and experiences and community feedback on providers (Box 1.5). The project team’s review identified that these approaches are usually sup- plementary to the primary support activities since they do not influence the modality of support delivery but can provide additional impact. However, some programs, particularly training and advisory programs, integrate group-based approaches as a core activity. The integration of a group approach in core activities requires robust community management capabilities, which are often provided to public institutions by an external partner. BOX 1.5 Examples of community engagement in support programs AUXILIARY ACTIVITIES in the system, and the value of its offering is publicly val- idated by beneficiaries. MSMEs that interact with a given Exchange of experiences | Mittlestand 4.0 Compe- provider are incentivized to submit a quantitative rating tence Centers/Germany. All centres organize a variety and a qualitative review through a formal requirement of modalities for exchanging experiences between to provide feedback in the agreement between MSMEs MSMEs, including, but not limited to, conferences, and A*STAR, which co-finances the cost of cooperation learning meetings, newsletters, and discussion plat- with technology providers and advisors in the ACCM forms. A variety of communication channels are utilized, database. both online and offline, to reach respective segments of the target population. Discussions between peers allow CORE ACTIVITIES for the dissemination of first-hand information and Group training | Innovation Manager’s Academy/ informal advice on overcoming barriers to digitalization Poland. A dedicated program of up to 120 hours of lec- and the potential resulting from using ICT technologies. tures and workshops delivered to a group of managers Showcasing role models | Early Warning Europe/ from several dozen MSMEs. As of Q3 2022, five editions Denmark. Propagation of information about firms or of the Academy had been implemented, with partici- individuals that have successfully adopted certain pants from over 150 firms consistently assessing this digital practices builds confidence among MSMEs that program as one of the most beneficial for the adoption of productivity benefits could be replicated in their organi- novel technologies and business practices in firms. The zations. The development potential is especially strong public institution that manages the Academy partnered when success stories are shared directly by business with a leading national business university and a startup owners or managers that went through the process accelerator to design and implement training, as well themselves. Following these assumptions, the Early as manage the community during the project and in the Warning Europe program directly connects role mod- follow-up stage, after its implementation. els — experienced managers that underwent restructur- Peer learning | Activa Industria 4.0/Spain. Program ing processes — with firms in financial distress to build supporting collaborative technology demonstration confidence in measures needed to regain solvency. and experimentation with features facilitating peer Community feedback on providers | The A*STAR Col- signaling. Firms engage with programs in the creation laborative Commerce Marketplace (ACCM)/Singapore. of prototypes and pilot programs for the adoption of IT ACCM is a government-driven e-portal where busi- technologies. Thanks to the dedicated time set aside for nesses collaborate to find business and technology syn- observation of the efforts and steady progress achieved ergies and look up providers of digital technologies and by their peers, firms may become motivated to take advisory services. Each provider has a separate profile chances on the adoption of new technologies. Source: Digitrans team based on the review of programs and instruments in Chapter 3. The review of programs and instruments has revealed that community engagement can be incorporated into three types of instruments to expedite the technology adoption process. Firstly, small and temporary subsidies such as time-limited discounts have been found to 40 1 reduce procrastination and encourage adopters with present-biased preferences to adopt   earlier, as well as effectively overcome employee reluctance to use technology (Duflo, Kremer, and Robinson 2011). Secondly, education schemes aimed at addressing technological and non-technological biases encourage firms interested in adopting to do so early and encourage  Summary of lessons and promising strategies those facing financial constraints to apply for external sources of funding to cover the fixed cost of adopting and speed up the process (Gine and Yang 2009). Thirdly, demonstrating useful technologies through offering spaces for collecting, processing, and sharing high-frequency information on business performance can induce behavioral changes in those interested in adopting Beaman, Magruder, and Robinson 2011). Leveraging behavioral nudges such as time-lim- ited subsidies or risk-sharing plans can also be effective in encouraging technology adoption (Duflo, Kremer, and Robinson 2011). Community engagement approaches such as exchanging experiences, showcasing role models, and providing community feedback on providers can be used in the design of these interventions, to enhance their effectiveness and promote greater participation from stakeholders. Group-based approaches can also be utilized, and effective mechanisms for delivering the subsidy and filling the information gap should be considered during the intervention design. 41 1.4 DESIGN GUIDE FOR POLICY MAKERS 1   KEY LESSONS FROM SECTION 1.4  Summary of lessons and promising strategies This section presents a five-step guide to determine essential components of effective firm-level interventions for addressing gaps in complementary capabilities necessary for digitalization among MSMEs. Interventions should be tailored to the specific challenges and characteristics of the target groups as unique technological challenges and requirements are faced by different industries. The successful implementation of a digital agenda depends on the capabilities of the organization or program manager. Quality assurance and certification mechanisms are necessary for maintaining quality in external providers since different programs require varying resources for effective implementation. Institutions’ capabilities must align with the requirements of support activities, and corrective actions are needed if there is a mismatch between agency capabilities and instrument requirements. When scaling up pilot instruments, it is important to consider potential risks and mitigation measures. An analysis of the external environment should be conducted to identify regulatory and infrastructure bottlenecks. To ensure progress monitoring and impact evaluation, establish monitoring and evaluation capabilities and commission evaluations to external, independent institutions. In this part of the report, we present brief guidance to secure the successful design and imple- mentation of instruments for addressing gaps in complementary capabilities for digitalization among MSMEs. It is a complementary set of institutional recommendations to facilitate the incorporation of target groups, challenges, and lessons learned. The project team prepared the design guide by adapting well-documented best practices regarding effective policymak- ing in other areas related to private sector development — digitalization (World Bank 2020a), entrepreneurship (Olafsen and Cook 2016), research & development and innovation (Correa 2017), competitiveness (Cirera et al. 2020) — through the lens of complementary capabilities for digitalization. Particular attention was paid to factors crucial for pilot interventions, as the nascent stage of global programming in the area of complementary capabilities justifies testing instrument designs before proceeding with full-scale rollouts. The guidance describes six key steps for determining the crucial features of an effective firm-level intervention to enhance the adoption and usage of digital technologies for enhanced productivity. The process is organized around two parts of successful instruments: the diagnostic component and the intervention component (Table 1.3). Particular attention is paid to identifying gaps regarding the complementary capabilities of firms in a catching-up regional context. Information gathered in this chapter will also be useful for policy makers aiming to enhance MSME digitalization in other contexts. 42 1 TABLE 1.3 Steps in designing digitalization instruments   Step 1. Determine target group and specify the challenges Step 2. Identify firm needs and plan outreach strategy Embed progress Diagnostic  Summary of lessons and promising strategies monitoring and component Step 3. Leverage digital diagnostic tools and market providers of transformation plans impact evaluation across design Step 4. Determine support activities process. Intervention component Step 5. Set up implementation arrangements Source: Digitrans team. STEP 1 Determine the target group and specify the challenges Design quality is the most important factor driving the effectiveness of a support instrument (Cirera et al 2020). Articulate the challenge and the relevant market or system failures through a transparent and legitimate process should guide the design process and allow public author- ities to avoid inferring or competing with markets and private entities (Table 1.4). Putting market or system failures at core of design process is necessary, as public interventions in areas with no market or system failure would likely crowd out private investments by disincentivizing firms to be active in this area (Rupeika-Apoga, Bule, and Petrovska 2022). Sole reliance on informa- tion from beneficiaries is not sufficient for justifying public intervention. For example, many enterprises constantly report lack of finance as one of the main barriers for adopting digital technologies, regardless of the availability of financial opportunities on the market and short payback periods of many ICT solutions (European Investment Bank 2019). TABLE 1.4 Comparison of market and system failure Market failure System failure Cause Inefficient allocation of goods, services, and Lack or insufficient supply of a necessary social capital by a free market leading to a social welfare good without government action or without a loss (Martin and Scott 2000). change in its current form (Edquist 2011). Examples • Information asymmetry • Unequal access to broadband • Incomplete information • Prohibitive regulations • Externalities • Inadequate capabilities Source: Digitrans team. Preparing a high-quality justification requires appropriate analytical capabilities on the side of managing authorities. External support to conduct analyses could be leveraged, if appro- priate internal capabilities do not yet exist within public institutions. Assessment of available human and financial resources would inform decision on the approach to the design process. In situations with constrained resources, policymakers could consider limiting consideration to more straightforward instruments, where all aspects of their justification are well understood (Grindle and Hilderbrand 1995). Unequivocal definition of challenges addressed by interven- tion is particularly important in pilot instruments, to generate relevant learning. Policymakers need also to have appropriate capabilities to understand sectoral considerations and design targeted policies that address the specific challenges and opportunities faced by firms per- forming particular type of economic activity (Box 1.6). 43 1   BOX 1.6 Sectoral considerations in support for complementary capabilities of MSME digitalization Different industries face unique technological challenges Additionally, market competition and government and requirements, which can impact their willingness to policies play a crucial role in shaping technology  Summary of lessons and promising strategies adopt new technologies. For example, labor-intensive sec- adoption in different sectors. Companies operating in tors like textiles and garments may be hesitant to automate highly competitive industries may prioritize investing production processes due to high investment costs and the in new technologies to maintain a competitive edge, potential for job losses. These industries may also heavily while those in less competitive sectors may have less rely on low-skilled labor, perceiving new technologies as a motivation to do so. Moreover, government policies such threat to job security. Conversely, capital-intensive indus- as subsidies and tax incentives can either encourage or tries like automotive manufacturing face challenges such discourage technology adoption in various sectors. By as high capital costs and complex technological require- adopting a sectoral approach to technology adoption, ments but may benefit from increased efficiency and pro- policymakers can facilitate inclusive and sustainable ductivity gains. Therefore, they may be more likely to invest economic growth, while maximizing the development in new technologies that can enhance their operations. impact of public investments. Source: Digitrans team based on Cirera 2022. STEP 2 Identify firm needs and determine outreach strategy Firms at different levels of technological and managerial capabilities need different support activ- ities. Policymakers must understand the current needs and barriers faced by a target population of enterprises in real-life regional or sectoral context. Approaches based on directly copying other economies — where firms could face different challenges — might not be effective (Intarakumnerd and Goto 2016). Each intervention needs to be tailored to challenges and characteristics of target groups identified in Step 1, by leveraging the most granular data on the use of technology, contex- tual factors, and barriers. Analysis of firm needs should also provide insights for implementation. For addressing gaps in complementary capabilities, many of which are dependent on lack of awareness or biases, firms must have trust in participating in the program and confidence that intervention will yield benefits for their business operations (Schizas, Jarvis, and Daskalakis 2012). Each group of firms needs to be contacted differently. Identifying the right population of enter- prises is necessary but not sufficient for success, and communication failure is a potential risk. The risk is particularly high in catching-up companies, which exhibit gaps in awareness of both offerings of public support and their own needs (World Bank 2022). Many firms at the foundational level of digital sophistication have not benefited from public support programs. Outreach to them requires employing mass-scale communication methods utilizing offline channels. Examples of cost-efficient approaches to such campaigns include automated tele- phone calls or actionable flyers attached to letters from the public administration (Huizingh and Mulder 2015). The unit cost of such methods is lower than 1 USD per firm, and leveraging behavioral design allows for achieving conversion rates higher than 10% (ibid). STEP 3 Leverage digital diagnostic tools and market providers of transformation plans Select a DDT aligned with characteristics of target group, defined challenges, and available resources. Digital diagnostic tools have by design various characteristics. Many DDTs are light on resources and respond to needs of enterprises in most economic sectors. Such tools usu- ally relate to firms at earlier stages of digital transformation, as challenges in more advanced firms require customized approach (Rush, Bessant, and Hobday 2007). On the other end of the 44 1 spectrum, there are DDTs with narrow sectoral and technological focus, catering to the needs   of enterprises that already frequently use advanced technologies and integrate digital angle in strategic plans. Policy makers need to choose the DDT that best suits their project objectives (Table 1.5). It is also worth considering mixing features from multiple tools. The approach used  Summary of lessons and promising strategies for diagnosing firm needs should allow for tracking not only the impact of support activities on the firm but also the impact and relevance of the diagnosis itself. TABLE 1.5 Criteria for selecting a digital diagnostic tool Area Options Focus on technology Mixed approach – combining Thematic scope Focus on managerial practices aspects tech and management Sectoral focus Sector-specific Manufacturing / Service firms Horizontal Size groups Micro/small firms MSMEs Medium/large firms Required initial level of Foundational Transformational Mature technology sophistication Delivery mode Online Mixed – online and in-person In-person Time – firm 20 minutes 2 x 30 minutes 5 – 10 hours Time - provider Automated 1 – 5 hours 1 – 4 weeks Cost per firm <10 USD 50 – 100 USD 100 – 1,000+ USD Evaluation features Initial only diagnosis Follow-up track of firm-impact Tracking adequacy of a tool itself Source: Digitrans team based on review of digital diagnostic tools listed in Annex 1. To assure the quality of treatment received by firms, facilitate partnerships between local providers delivering transformation plans to firms and the creators of DDTs or providers with track records in their successful implementation. All diagnostic plans must be delivered by a professional and experienced team. Some approaches require particularly high resources from providers. However, a shortage of providers capable of delivering quality assistance to MSMEs is a common factor limiting impact of firm-level support instruments (OECD 2017). Partnership between junior providers and senior experts are the quickest and most efficient way to alle- viate supply constraints (Bennett 2012). Such collaboration could include a train-the-trainers scheme, where the staff of local providers would learn how to prepare transformation plans from experienced users, or direct involvement of an external party in service delivery (for ex- ample, by validating collected information or providing feedback to draft transformation plans). STEP 4 Determine support activities Firms with different technological and managerial capabilities need different support activities. Each intervention needs to be tailored to the particular objectives and characteristics of the expected beneficiaries. However, we have identified five common characteristics of programs and instruments successful in enhancing digitalization among catching-up firms: 1. Encourage small steps, not giant leaps. Framing technology adoption and usage in terms of small, incremental changes rather than all-in bets with high costs and risks can help dis- pel uncertainty and correct misconceptions about what it means to go digital. 45 1 2. Celebrate small business peers. Highlighting examples of small businesses that have adopted   digital technology — and sharing both their successes and their struggles — helps to normal- ize a process that can appear much more daunting than the status quo.  Summary of lessons and promising strategies 3. Integrate technology and managerial training. The provision of technology is a necessary first step but not enough for its further use — even in the case of simple solutions, both man- agers and workers could benefit more after exploring their functionalities in a safe, non-judg- mental learning space. 4. Prioritize solutions that bring quick, even small, but measurable benefits. Entrepreneurs are much more likely to digitalize if they see a relevant and tangible impact on small busi- ness operations based on their first-hand experience. 5. Narrow and simplify choices. More information is not always better. Often it causes small business owners and managers to freeze. Replace information and option overload with clear-cut, well-informed technology adoption plans that give MSMEs confidence to act, not confusion. Create a clear logic model8 that establishes explicit and plausible connections between the challenge, target group, the chosen instruments, and the desired outcomes. The logic model documents how an instrument is supposed to work and can help minimize the risk inherent to any public intervention (Figure 1.9). Without a logic model, many components of the intervention may go unexamined, leading to numerous unintended and usually undesirable consequences, most of which could be anticipated and prevented with a better design (Gugerty and Karlan 2018). FIGURE 1.9 Logic model Explicitly documented and Explicitly documented and analyzed assumptions analyzed assumptions regarding regarding translation of impact of products on higher-level planned inputs into products economic goals Problem Problem Inputs Activities Outputs Immediate Long-term causes symptoms outcomes outcome Source: Adapted from Gugerty and Karlan 2018. Given the inevitable engagement of external providers in almost all instruments, an appropri- ate certification mechanism is necessary for maintaining quality. In a certification program, the certifying institution can be either a government body or a third party. While determining support activities, policymakers need to decide on requirements that providers will need to fulfill to earn the certification. A results framework for evaluating providers should support experimentation and provide some degree of flexibility, at the same time providing relevant quantitative key performance indicators (KPIs) for quality supervision (OECD 2018). 8.  Logic models are also sometimes called ‘logic frameworks’ or ‘logframes,’ ‘theory of change models’ or simply ‘theories of change,’ or ‘program matrixes.’ 46 STEP 5 1 Set up implementation   arrangements Effective implementation of policy instruments and programs is not solely dependent on good  Summary of lessons and promising strategies policy design, but also on the capabilities of the organizations and program managers respon- sible for their implementation. Therefore, implementation arrangements cannot be discussed in isolation from their mother organizations/agencies. To effectively implement a digital agenda, many countries have established dedicated agencies for its promotion and dissemination. Different programs require different resources on the implementation side. Instruments ded- icated to firms at the foundational level of digital sophistication could be relatively quickly deployed at scale even with limited resources; however, raising interest and building trust among the target group is challenging, which, due to widespread gaps in awareness among firms at the foundational level (World Bank 2022), are usually a necessary first step (Cirera et al. 2020). The information presented needs to be of high quality, related to the everyday business context of catching-up companies and delivered in language understandable by non-experts. On the other hand, programs addressing the needs of more digitally advanced firms usually require more scarce resources, such as technical experts for design, specialistic providers to interact with beneficiaries, or expensive, highly customizable technologies, and their aggre- gation adds a layer of complexity for policymakers (Miorner et al. 2019). Capabilities of implementation institutions need to align with the requirements of support ac- tivities. Ensure that the identified market and system failures can be addressed in the context of scarce government capabilities. Public institutions need to gradually accumulate know-how and human and financial resources along the so-called capabilities escalator to provide relevant support for upgrading firms (Figure 1.10) (Cirera and Maloney 2017). When there is a mismatch between existing agencies’ capabilities and the instrument’s requirements, corrective actions should be planned. Policy makers could increase the capabilities of existing agencies, bring in capabilities from external institutions, or both. However, it is capable internal staff and strong managerial practices that condition institutional capabilities in the long term (Aridi and Kapil 2019). FIGURE 1.10 Capabilities escalator for agencies supporting the digitalization of small and medium enterprises Programs facilitating products/services innovations Mature based on usage of frontier digital technologies. Stimulate digital support creation of digital technologies through creating favorable ecosystem conditions for digital-specific business R&D. Level of sophistication Support integration of digital technology across business functions Maturing by addressing within-firm barriers for using medium and highly digital support sophisticated ICT solutions, incl. alignment of managerial practices ecosystem with technology needs (Roberts and Shaw 2022). Incentivize digitalization of individual business functions Incipient through assistance in creating foundational digital and digital support managerial capabilities, preceded by elimination of awareness, ecosystem informational and behavioral barriers to MSME digitalization. Source: Adapted from Cirera and Maloney 2017. With pilot instruments intended for future scale-up, several features for increasing the scale of intervention could be tested. Increasing the scale of intervention inevitably increases the risk of an instrument missing its stated objectives (Hudson, Hunter, and Peckham 2019) because 47 1 each novel feature could turn out negatively (Table 1.6). Identify, avoid, and prepare measures   for mitigating potential risks of four broad causes of policy failure. The first is exceedingly optimistic expectations — including regarding maintaining quality constant despite changes in the scale of an intervention. The others are overly decentralized governance, inadequate  Summary of lessons and promising strategies collaborative policymaking, and political economy factors (ibid.). Specifically for scaling-up pilot instruments, consider the availability of an appropriate number of providers and delivery method. Policymakers need to determine whether the targeted firms should receive funds to develop the project directly or whether a third party should control the funding according to milestones reached. When training, knowledge and advisory services are subsidized, the government usually provides the MSMEs a voucher or grant to purchase services from a third party. When the government provides the service directly, it generally has centers with local and regional advisers that deliver the services to MSMEs. TABLE 1.6 Instrument features for increasing the scale of intervention Feature Potential benefits Potential risks Mitigation measure Increased • Deliver support to more • Difficulties in maintaining • Train the trainers coordinated number of firms consistent quality of services by a leading partner with high providers (ibid.) capabilities • Limited learning due to • Explicit incentives for unwillingness to share best collaboration between practices providers in intervention mechanics. High-intensity • Access to firms that have • Low cost-benefit efficiency • Staged approach with outreach not benefited from public automated mass communication (e.g., in-person support in the past followed by more resource- visits to firms) consuming actions for target • Facilitates data collection groups. for ex-post evaluation Broader • Cover firms in various • Shortage of providers with • Community engagement — e.g., geographical regions appropriate quality through leveraging successful coverage digitalization managers as role models instead of paid experts. Adapting • Increasing the size of the • A one-off adoption, with the firm • Projects should be based intervention to target group settling back into inertia in the on the gradual build-up of firms with lower future if firm management does technological complexity. capabilities not learn from the process Source: World Bank Group based on analysis of global programs and instruments — details in Chapter 3. Ensure adequacy of the external environment by analyzing potential regulatory and infrastruc- ture bottlenecks. Use regulatory impact assessment to identify whether regulations enable all actors envisaged in the intervention to perform their activities and allow time for the expected outcomes to materialize (OECD 2021). The prospective program also needs to be checked against other external factors that could increase or inhibit its effectiveness. Some common regulatory and infrastructure bottlenecks that could impede the success of policies to upgrade digital technologies include white spots in access to broadband and a lack of data sharing and interoperability standards (ibid.). At the same time, outcomes of MSME digitalization instruments are known to be amplified when deployed along with strengthening regulatory framework (e.g., administrative acceptance of e-signature). 48 1 Embed progress monitoring   and impact evaluation across design process All instruments must include monitoring and evaluation features by design, allowing policy mak-  Summary of lessons and promising strategies ers to determine whether an intervention was successful and which areas could be improved. A robust monitoring and evaluation (M&E) framework is essential for experimentation through pilot programs to generate trustworthy evidence on the effectiveness of an intervention within a specific context. Information for M&E must be gathered in a timely and systematic manner, allowing for the creation of a baseline for future impact assessment. Including M&E approaches and systems at the design stage allows for integrating realistic methods of data collection that do not distort the instrument logic or generate burdensome bureaucratic requirements. An important decision for program designers is what criteria should be included in the pro- gram’s ex-post evaluation. It is a starting point for designing any functional M&E system (Table 1.7). For example, should adoption of a novel technology, or increase usage of technologies that firm has already at disposal, be a priority? Or is just getting the firm to take the first steps towards adoption the metric of success? As an example, the Tech Access program in Singapore focuses mostly on getting firms to experiment with new technologies, seeing that as an end in itself. TABLE 1.7 Do’s and don’ts in monitoring and evaluation of pilot programs DOs DON’Ts a • Follow the CART framework when designing an M&E system. • Don’t evaluate the impact of programs based only on beneficiaries’ surveys or fulfillment of administrative • Design the main principles of an evaluation strategy related performance (e.g., the number of supported firms to a program before starting implementation (Olejniczak, or value of disbursed funding) or generic economic Raimondo, and Kupiec 2016). indicators (such as value added or employment). • Track indicators resulting from the logic of the program. • Don’t make data collection overly burdensome for • Use appropriate statistical methods for evaluation beneficiaries; it will reduce program efficiency and that respond to the mechanics of the program (proper could disincentivize potential applicants. specifications of quasi-/experimental designs). • Don’t reduce M&E system to fulfill regulatory • Conduct systematic data collection to enable policy requirements. evaluation, learning, and improvement. • Don’t rely only on surveys of beneficiaries’ self- • Introduce learning from the M&E system of the pilot reported information. Use administrative data as intervention to improve the design and roll-out. much as possible (e.g., information from tax returns, • Make evaluation results and documentation public. statistical data, etc.). Source: Digitrans team adapted from Public Expenditure Review in Science, Technology, and Innovation in Croatia. Note: M&E = monitoring and evaluation. a. The credible, actionable, responsible, transportable (CART) framework stipulates the following principles for data collection and monitoring: (i) Credible: collect high-quality data and analyze the data accurately; (ii) Actionable: commit to act on the data you collect; (iii) Responsible: ensure the benefits of data collection outweigh the costs; (iv) Transportable: collect data that are comparable across evaluations. Because establishing M&E capabilities is resource-intensive, institutions that manage inter- ventions can commission evaluations to well-respected organizations and external experts. Impartiality of results requires commissioning the conduct of evaluations to external, competent, and independent institutions. However, gaps in internal M&E capabilities limit organizations’ abilities to generate sustained insights on instruments, which jeopardizes their long-term sustainability. Pilot instruments are a valuable opportunity to enhance M&E capabilities by incorporating the high-quality practices used to track pilots’ impact into the everyday opera- tions of an institution. 49 2 DIAGNOSTIC TOOLS 2 KEY LESSONS FROM CHAPTER 2   The two main types of diagnostic tools are surveys and Digital Diagnostic Tools (DDTs), which complement each other in identifying digitalization  Diagnostic tools needs of firms. Surveys provide descriptive data at the aggregate level, while DDTs are more granular and allow for identifying digitalization barriers and drivers at the specific level of the firm analyzed. Digital diagnostic tools are subdivided into stand-alone and integrated tools, with the former being more scalable and the latter providing more tailored feedback and guidance. The chapter compares surveys and DDTs in terms of usefulness for pilot design, delivery partners, sample size, data acquisition methods, and resources required by both institutions and firms. The effectiveness of diagnostic tools depends on their quality, and the usefulness of DDTs depends on whether they diagnose the most critical aspects of a business for digital transformation and provide post- diagnostic support to managers. The chapter analyzes three surveys and six DDTs that cover both managerial practices and complementary capabilities, with solid methodology and perceived usefulness for pilot design. It provides insights for designing the Digitrans pilot instrument. This section will review two main types of instruments for measuring digital technology adop- tion and use and complementary managerial practices at the firm level, with a primary focus on providing practical insights for designing programs to promote firms’ digitalization. The two main kinds of instruments are (i) those that are typically used to generate statistical data (i.e., surveys) and (ii) those that assess in a more granular manner needs of individual firms to prepare a digital transformation plan (i.e., DDTs). Surveys and DDTs have different functions but complement each other in identifying the dig- italization needs of firms. The former is descriptive and provides representative data at the aggregate status of digitalization level to inform policymakers who wish to understand drivers and barriers to the digitalization of firms, as well as to identify specific target groups among the MSMEs in need of support. The latter is more granular and allows for identifying digitalization barriers and drivers at the specific level of the firm that is analyzed. This diagnostic provides the basis for an improvement plan with specific steps and priorities — for instance, identifying and prioritizing technology or managerial practices to adopt. (Table 2.1). The interplay between these two kinds of instruments is useful for accelerating digitalization efforts at scale. Survey results could be leveraged for deciding on the area of intervention and characteristics of the target group, whereas insights from DDTs could guide the selection of beneficiaries from the target population and tailor support delivered to them. However, the usefulness of the diagnostic tools depends on their quality. Surveys cannot inform effective policies if the sample is unrepresentative, the data collected is of low quality, or if the wrong questions are asked. Similarly, DDTs cannot lead to successful digital transfor- 51 2 TABLE 2.1 Comparison of surveys and digital diagnostic tools for pilot design   Surveys Digital Diagnostic Tools (DDTs) Usefulness for Step 2. Identify firm needs pilot designa  Diagnostic tools Provide data for benchmarking level of digitalization against competitors in a given sector/region Step 1. Determine target group and specify Step 3. Create transformation plan challenges • Determining treatment provided to individual firms • Informing design of the Step 5. Set up implementation activities instrument — target group and challenges to be addressed • Deliver data to monitor progress in intervention delivery Step 6. Embed progress monitoring and impact evaluation • Create information for evaluating impact of the instrument Usual delivery Statistical office or business intelligence Technology advisory firm, business support institution, partner firm university, or consortium of the above institutions Sample size Representative sample — min. 100 – 150 for Potential beneficiaries of intervention, upon expressing representativeness on a national scale, willingness to participate in the program. up to 3,000 for full representativeness on multiple firm characteristics (e.g., by size groups, sectors, and regions). Data acquisition Self-administered web survey, computer- Individual in-depth interview (IDI), individual or group methods assisted telephone interview (CATI). diagnostic workshop. Resources — insti- Financial resources: pprox.. 30 – 60 USD Financial resources: from 0 USD per user (open access tutional side per interviewee in Europe and Central online self-assessment tools) up to pprox.. tens of Asia. thousands USD per firm — depending on licensing costs and the co-payment arrangements (Bellantuono et al. 2021). Resources — firm Time: 20 – 60 minutes per firm Time: from 30 minutes (online self-assessment tools) to side 15 – 20 hours (extended programs with hands-on expert consultations) Financial resources: from 0 (online self-assessment tools) to several hundreds USD — depending on the co-payment arrangements. Source: Digitrans team analysis based on a review of digital diagnostic tools listed in Annex 1. Note: a. The steps in the leftmost column reflect the model process for instrument design introduced in Section 1.4 (A guide for policymakers). mation if they don’t diagnose the aspects of a business that are most strongly correlated with successful digital transformation or do not provide post-diagnostic support to managers who want to embark on a digital transformation journey. The functionality of both types of tools also depends on the adequacy of resources dedicated to the diagnosis, both by the institution implementing the tool and by the participating firm. This section provides an in-depth analysis of diagnostic tools identified as global best prac- tices to identify components that could be utilized in the Digitrans pilot design. Out of over 20 surveys and 70 DDTs mapped globally (described in Annex 2), the design and implementation details of three surveys and six DDTs were analyzed. The selection criteria for the tools analyzed in-depth were (i) covering managerial practices and other complementary capabilities, (ii) solid methodological grounding, and (iii) perceived utility for the pilot design. The analysis utilized publicly available documentation and was supplemented by information sourced directly from the creators of the analyzed surveys and DDTs. 52 2.1 SURVEYS 2   Cross-country and country-specific sectors vary in terms of digital maturity and in terms of the drivers and barriers to digital adoption and use. As such, there is no one-size-fits-all pol-  Diagnostic tools icy approach that enables digitalization at scale. Successful support for the advancement of digitalization requires identifying policies relevant to the challenges of firms with various characteristics, including, but not limited to, firm size, age, sector of operation, geography, or export status. Collecting firm-level data on technology adoption and use and complementary managerial practice — and on drivers of and barriers to digitalization — is useful because such data can inform the implementation of targeted policies and interventions (Bloom et al. 2021). However, only a limited number of existing firm-level surveys combine technological and managerial aspects. For instance, Eurostat’s Community Innovation Survey (CIS) combines the components of managerial practices and barriers/drivers of adoption but focuses strictly on innovation rather than technology adoption and use.9 Eurostat’s Community Survey on ICT Usage and e-Commerce in Enterprises (CSI) surveys the use of technology but does not include components on manage- ment practices and questions related to barriers and drivers specific to digital transformation.10 The US Census Bureau’s Management and Organizational Practices Survey (MOPS), which was developed to capture management and organizational levels, includes very few questions on technology use and does so strictly with a focus on manufacturing activities.11 The World Bank’s Firm-Level Adoption of Technology (FAT) survey is the closest to a survey that provides information at a granular level on digital technologies adoption and combines both technology and managerial aspects. However, the technological component is not focused on digital solutions (although it includes them), and the set of questions on managerial practices is limited.12 No firm-level survey implemented regularly and on a representative sample gathers information on complementary capabilities for digitalization other than managerial capabilities, such as tech- nological preferences, level of knowledge, or behavioral attitudes toward digital transformation. Behavioral factors heavily influence technological decisions made by firms (Lee 2004). Methodologies for generating organizational and behavioral insights related to digital technologies among firms usually rely on qualitative research through interviews with individual companies (Yousafzai, Foxall, and Pallister 2008), whereas surveys tend to be limited to questions on perceived barriers to tech- nology adoption (Hameed, Counsell, and Swift 2012). Despite this lack of comprehensive surveys, to design a digital upgrading project, it is possible to rely on existing surveys for gathering evidence regarding targeting at the instrument level and assessing the eligibility of applicants. The following part of Chapter 2.1 analyzes the four surveys referenced above in light of their usefulness for the design of a digital upgrading pilot program and key implementation and design factors, including its objectives, target population, statistical characteristics, implementation history, and costs. Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises Eurostat’s CSI provides the most comprehensive and up-to-date snapshot of digital technol- ogies used by European firms. Its usefulness is limited to the adoption at the extensive margin of digital tools. The CSI survey (Table 2.2) provides inputs for high-level determination of target 9. Eurostat CIS 10. Eurostat CSI 11. MOPS 12. FAT 53 2 group and challenges by evidencing gaps in the adoption of technologies to enterprises relative   to other countries and sectors. Its results are particularly relevant for targeting firms of partic- ular size and sector because they are representative across four size groups in their respective NACE-1 groups of activities.13 Because the results of the CSI surveys are easily available, they  Diagnostic tools can support the design process. The CSI survey, as long as it tracks the same firms, can also be used as a monitoring system, especially if statistical offices and the implementing agency can share information on program beneficiaries and include this as an additional variable in the survey. However, due to the low level of detail, the questionnaire is not useful for analyzing needs at the level of individual firms or for eligibility screening.14 TABLE 2.2 Summary of Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises Objective Collect and disseminate harmonized and comparable information on the use of ICT and e-commerce in enterprises at the European level Target Users Firms of all sizes in most business sectors (public sector excluded). Managing Organization(s) Eurostat (in collaboration with OECD) Data Collected • Access and use of the internet • e-Commerce sales • Sharing of information electronically within the enterprise • Use of cloud computing services • Internet of things (IoT) • AI • Background characteristics Managerial Data Collected A. Data and knowledge management B. IoT-related practices C. AI-related practices Statistical Unit The enterprisea Sample Frame National Business Registers or Structural Business Survey (SBS) frame population Sample Size Pseudo-panel of nearly 135,000 firms across Europe Costs Time burden: An average of 40 minutes, ranging from less than 20 minutes to about 80 minutes; Financial burden (for the firm): none Source: Digitrans team based on survey documentation. Note: AI = artificial intelligence; ICT = information and communication technology; IoT = Internet of Things; OECD = Organisation for Economic Co- operation and Development. a. The enterprise is defined as “the smallest combination of legal units that is an organizational unit producing goods or services, which benefits from a certain degree of autonomy in decision-making, especially for the allocation of its current resources. An enterprise caries out one or more activities at one or more locations. An enterprise may be a sole legal unit.” The purpose of the CSI, which Eurostat developed in collaboration with EU member countries and the OECD, is twofold. First, it aims to help understand the implications of ICT-driven trans- formation of European societies and economies. Second, it aims to contribute to benchmarking for the Digital Economy and Society Index (DESI) and the Digital Intensity Index (DII). The CSI is also widely used by decisionmakers across the EU to prioritize public policies and interventions 13. NACE stands for Statistical Classification of Economic Activities in the European Community. This taxonomy includes 21 level-1 groups (e.g., manufacturing, construction, wholesale and retail trade, or professional, scientific & technical activities. 14. Another limitation is that the micro-level data are not easily available and need to be requested on an “ad hoc” basis from the National Statistical Offices. 54 2 by leveraging information on the extensive margin of technology use in firm groups of various   characteristics (Labelle 2019). The survey’s coverage is broad in terms of geography, economic activity, and technology. The  Diagnostic tools survey (Table 2.3) covers all 27 member countries as well as Bosnia and Herzegovina, Iceland, Montenegro, North Macedonia, Norway, Serbia, Türkiye, and the United Kingdom, with data col- lection and publication occurring annually. Concerning economic activity, it covers enterprises in 13 NACE Rev. 2 categories, both industrial and service firms, with 10 or more employees, and self-employed persons with microenterprises (up to 9 employees) are covered optionally. The CSI collects data on the adoption of 13 types of technologies (Table 2.3).15 TABLE 2.3 Information covered by the Community Survey on Information and Communication Technology Usage and e-Commerce in Enterprises Group of factors Indicator Framework conditions • ICT systems and their usage • Use of the Internet and other electronic networks Managerial practices • Data and knowledge management • IoT- and AI-related practices Technology group - Transactional a • E-commerce (selling goods or products online) • E-business processes (e.g., digital marketing) Technology group - Informational • Sharing of information electronically within an enterprise • Use of cloud computing services • AI • Machine learning Technology group - Operational • IoT • Use of industrial robots Background characteristics • Basic financial information Source: Digitrans team based on survey documentation Note: Groups of factors follow groups of challenges described in Chapter 3 of this report, and the typology of technologies is from World Bank 2020a. a. Tables 2.3, 2.6 and 2.9 refer to three groups of technologies — transactional, informational, and operational — following World Bank (2020). This categorization depends on their underlying source of productivity gains. Transactional technologies, such as platforms, better match supply and demand by lowering information asymmetries. Informational technologies, such as cloud computing or big data, reduce costs of data analysis and computing. Operational technologies, which include robots or drones, allow reducing labor costs by automatizing tasks. Most countries use a combination of sample and census methods when conducting the sur- vey. Typically, the survey (Table 2.4) is a census (i.e., it covers the universe of firms) for larger enterprises (250+ employees), but for smaller firms (<250 employees), firms to be surveyed are selected by stratified random sampling. The samples were predominantly stratified along economic activity (NACE-2 level) and size class (based on the number of employees), and oc- casionally geography (using the nomenclature of territorial units for statistics or NUTS), with the Neyman Allocation method as the primary mechanism for determining strata size (Eurostat 2018). Participation is mandatory in most countries as an extension to typical national statistical forms. Statistical offices employ various methods to conduct the survey, with self-administered web surveys being the most frequent, followed by computer-assisted telephone interviews (CATIs). Face-to-face interview methods are rarely used. Thus, the survey methodology does not allow for accommodating open-ended questions. 15. Full questionnaire and results of the survey is available at https://ec.europa.eu/eurostat/web/microdata/commu- nity-innovation-survey 55 2 TABLE 2.4 Composition of the Community Survey on Information and Communication Technology Usage   and e-Commerce in Enterprises Module Description Mandatory Questions Optional Questions A Access to and use of the internet 7 9  Diagnostic tools Access to the internet 1 0 Use of a fixed line connection to the internet for business purposes 2 0 Use of a mobile connection to the internet for business purposes 0 2 Use of a website 0 7 Use of social media 4 0 B e-Commerce sales 24 2 Web sales of goods or services 19 2 EDI-type sales 5 0 C Sharing of information electronically within the enterprise 3 0 D Use of cloud computing services 11 0 E Internet of things 8 0 F Artificial intelligence 14 12 X Background characteristics 3 0 Total 70 25 Source: Digitrans team based on survey documentation The CSI allows for accurate comparisons between countries, firms of various sizes, and economic sectors thanks to the collection of high-quality and harmonized data. This is accomplished using Eurostat’s model questionnaire, the clarification of concepts and questions within the methodological manual, and the set of recommendations laid out in the manual for national statistics institutes to follow when conducting the survey. The questionnaire is also granular in the type of technologies analyzed, which allows for gathering information on familiarity with a dozen or so technologies at the firm level. It also includes a few questions on technological preferences — for example, willingness to purchase software on a software-as-a-service (SaaS) model compared to on-premises investments. The survey does not provide information on the intensity of use of respective technologies and does not differentiate between business functions. Questions assessing how intensively a specific technology is used are limited to online sales (i.e., the share of products sold online). For all the other technologies, the firms are only asked whether the company uses them or not (i.e., the extensive margin). Moreover, barriers to adopting technologies are assessed only concerning AI and only as an optional question. It would be a more useful tool for policy makers and researchers if the barriers to the adoption of all technologies involved were surveyed. The questionnaire only marginally collects information on managerial practices and organization. A broader and more explicit assessment of managerial practices (e.g., the inclusion of 7 – 8 questions from MOPS) would complement the benchmarks on technology adoption and use. Nonetheless, this survey is very useful for any digital upgrading project for two reasons. First, during the rapid diagnostic of pilot countries, it can be used to gather high-quality data on techno- logical infrastructure, firm capabilities, and the adoption and use of transactional, informational, 56 2 and operation technologies. Second, this data can be used as part of the benchmarking data for   the diagnostic component of any pilot intervention to quickly assess an applicant’s level of tech- nological advancement against sectoral or regional peers, albeit only at the extensive margin.  Diagnostic tools Firm-level Adoption of Technology The FAT survey is a firm-level survey developed by World Bank in cooperation with academics and external sectoral technology experts and consultants. It was created to reflect a new approach for measuring technology adoption at the firm level by focusing on the technology associated with each business function (administration, sales, payment, production, planning, etc.) (Cirera, Comin, and Cruz 2022). The FAT identifies the array of technologies available for a firm to perform a task and reports the technology frontiers in the firm’s sector across firm sizes. Moreover, the survey allows for collecting data on proven drivers of and barriers to digitalization (Table 2.5). TABLE 2.5 Summary of Firm-level Adoption of Technology Objective Identify the technology frontier and the array of technologies available for a firm to perform a task, as well as the drivers, barriers, and impacts of tech adoption Target Users Small, medium, and large firms in all business sectors, with sector-specific questionnaires dedicated to 12 types of economic activities (e.g., agriculture, food processing, automotive, or mechanical engineering). Managing Organization(s) The World Bank Data Collected • General information and firm characteristics • Technologies used in general business functions • Technologies used in sector-specific business functions • Drivers and barriers to adoption • Balance sheet and economic performance Managerial Data Collected • Manager education level and years of experience • Information sources for technology purchases • CEO or top manager work experience • Whether the business offers incentives to improve performance • Number of KPIs monitored • R&D status Statistical Unit Enterprise disaggregated into individual business functions Sample Frame Heterogenous sampling based on business registers Sample Size Approx. 1,000–1,500 per country (for statistical significance across sectors and firm size groups) Costs Time burden: approx. 40–60 minutes; financial burden: none Source: Digitrans team. Note: CEO = Chief Executive Officer; KPI = key performance indicator; R&D = research and development. As of October 2022, the FAT has been implemented in a dozen countries16 across three conti- nents and covers a wide range of technologies used by firms of all sizes. In the EU, by October 2022, it had only been implemented in Poland. The FAT covers the use of technology across 16. Bangladesh, Brazil, Burkina Faso, Ghana, India (Tamil Nadu and Uttar Pradesh), Kenya, the Republic of Korea, Malawi, Poland, Senegal, and Vietnam. 57 2 general business functions (business functions car- TABLE 2.6 Information covered by the Firm-Level   ried out by all firms) and sector-specific business Adoption of Technology survey functions (business functions carried out by firms in Group of factors Indicator a specific sector) (Table 2.6). In addition, it asks two Framework • Access to electricity  Diagnostic tools types of questions about the technology used for a conditions • Access to Internet business function: (i) whether a particular technology is used for a given function (the extensive margin) and Managerial • Educational attainment of managers practices • Preferences in purchasing technology (ii) which technology is most intensively used for a • Green managerial practices a given business function (the intensive margin). The combination of these two features creates uniquely Knowledge and • Self-assessment of own level of digital awareness sophistication granular data on technologies used by enterprises, • Access to governmental support unmatched by any other survey implemented on a representative sample. Technology • e-Commerce (website, marketplaces) group — Trans- • Payment methods actional The FAT questionnaire captures rich information • Marketing and digital visibility on technologies used by firms; however, it collects • Customer Relationship Management limited data for assessing managerial practices. A Technology • Standard business software unique feature of FAT is generating knowledge on group — Infor- • Advanced business software mational technologies used at the level of individual business • Customer Relationship Management functions (Table 2.7). It distinguishes between the ex- • Enterprise Resource Management tensive and intensive margins (i.e., using a technology • Quality control technologies versus using it intensively). Such granular evidence • Artificial intelligence allows for more detailed targeting by fine-tuning fea- • Machine learning tures of an intervention to precisely identified gaps. Technology • Industrial machinery FAT’s questionnaire also provides a clear path for group — Opera- • HVAC technologies tional studying increasing digital sophistication — for each • Energy efficiency technologies general and sector-specific business function, it asks • Industry 4.0 about sets of technologies at increasing levels of • Internet of things technological advancement. The usefulness of FAT Source: Digitrans team data for instrument design, however, is currently Note: HVAC = heating, ventilation, and air conditioning. still limited by slim benchmarking data. Data are a. Sustainable technologies and green managerial practices were especially sparse for high-income and upper-mid- added to FAT in 2022. dle-income countries. As of end-2022, FAT produces cross-sectional data as it was implemented a single time per country. However, it could be leveraged to monitor changes over time by conducting it repeatedly over time. TABLE 2.7 Composition of the Firm-Level Technology Adoption survey Module Description Mandatory Questions Optional Questions A General information and firm characteristics 11 1 Screener 8 1 Ownership 3 0 B Technologies for general business functions 20 11 Technology enablers 7 0 Technology adoption 10 8 Green business function 3 3 C Technologies for sector-specific business functions 15–20 3–5 (12 different modules for individual sectors) 58 Module Description Mandatory Questions Optional Questions 2   D Drivers of technology adoption 16 2 Acquisition of machines, equipment, and software 2 0  Diagnostic tools Awareness and information 2 0 Access to finance 2 0 Management 2 0 Trade barriers and uncertainty of demand 2 1 Government support 2 1 Obstacles and drivers - ranking 2 0 Green management practices 2 0 E Labor, balance sheet and performance 9 1 Labor 4 1 Balance sheet 5 0 Total 71–76 18–20 Source: Digitrans team. The FAT survey is implemented on a sample representative of sector and size. The survey sample is stratified across three separate dimensions: up to 12 sectors, 6 firm size groups, and 5 – 10 regions. The predominant vehicle used to conduct FAT is a CATI, supplemented by in-person interviews in regions with low telephone penetration or a high share of informal firms (Cirera et al. 2022). The survey is conducted either by market research firms or by national statistical offices. In the latter case, statistical office data is also leveraged to prepare the frame of respondents. The FAT’s business-function-focused framework for measuring technology adoption and use permits the measurement of multiple dimensions (extensive and intensive margins) of tech- nology adoption in a granular manner. This scope and granularity allow policy makers to under- stand: (i) the process of technology adoption, (ii) the differences between sectors, (iii) the impact technology has on firms, and (iv) the main drivers of and barriers to technology adoption and use. Such a detailed understanding gives policy makers the necessary knowledge to design more targeted and effective policies that aim to increase technology adoption and use. FAT’s implementation require significant resources. Deployment of the questionnaire lasts up to 60 minutes per firm and requires a knowledgeable respondent, which limits the response rate and requires more implementation resources. A full-scale FAT survey with the participation of approximately 1,500 respondents requires 3 – 6 months, beginning with the initiation of contact with the first enterprise. The granular FAT questionnaire could be leveraged for conducting a rapid survey during the design stage of instruments supporting MSME digitalization. Although the distinction between extensive and intensive margin would enable identifying digitalization opportunities associ- ated with increased usage of technology already adopted by firms, specification of individ- ual business functions would provide evidence for specifying technologies for support in the pilot instrument. Digital technology adoption projects could utilize evidence generated from a streamlined version of the questionnaire. By reflecting assumptions of the pilot instrument, a smaller survey modeled after the FAT could focus on general business functions and collect limited information on business performance. 59 2 Management and Organizational Practices Survey   MOPS is a well-renowned tool for assessing managerial practices at level of individual firms and establishments. (See Table 2.8) It has the most developed managerial questionnaire and  Diagnostic tools has collected long panel data from over 20 countries worldwide, allowing for benchmarking across sectors, firm sizes, and geographies. However, little benchmarking data comes from European countries, which limits its current usefulness for analyzing the context of the EU Member States. MOPS is well-grounded in a scientific framework, and its results provide the basis for advancing economic research (Scur et al. 2021). MOPS has little coverage of techno- logical aspects, with questions limited to the utilization of AI for decision making, and numerous questions are applicable only to manufacturing companies. TABLE 2.8 Summary of Management and Organizational Practices Survey Objective Develop a better understanding of current and evolving management and organizational practices and to assist in identifying determinants of establishment and productivity growth. Target Users Manufacturing firms Managing Organization(s) United States Census Bureau Data Collected • Management • Organization • Costs and purchased services • Data, decision making and use of AI • Expectations • Background characteristics Managerial Data Collected • Problem solving • Number and type of KPIs tracked • Frequency of KPI review by managers and non-managers • Production targets timeframe, achievement, and awareness • Promotion/bonus practices for managers and non-managers • Reassignment or dismissal timeframes for under-performing managers and non-managers Statistical Unit Establishments a Sample Frame Annual Survey of Manufacturers (ASM) mailout sample: approx. 102,500 manufacturing establishments Sample Size Pseudo-panel of approximately 50,000 manufacturing establishments Costs Time burden: An average of 45 minutes; financial burden (on the firm side): noneb Source: Digitrans team. a. As per the United States Census Bureau’s definitions, an establishment is a “single physical location at which business is conducted or services or industrial operations are performed,” whereas a firm is a “business organization consisting of one more domestic establishment in the same geographic area and industry that we specified under common ownership or control.” b. Information gathered from managers from the United States Census Bureau responsible for managing MOPS. The purpose of MOPS is to understand managerial and organizational practices, measure them over time, and link them with productivity and firm growth. It is a large-scale survey on managerial and organizational practices, implemented since 2010 at five-year intervals, to better understand current and evolving management and organizational practices and to assist in identifying the determinants of firm-level productivity growth. The MOPS data is used to inform government policies and allow enterprises to benchmark themselves against better-performing peers. It was created in a partnership between the US Census Bureau and leading global researchers on firm-level productivity and has a firm grounding in the scientific 60 2 literature (Buffington and Jarmin 2017). The questionnaire includes information on managerial,   organizational, and decision-making practices, together with questions on business performance, costs, and expectations (Table 2.9). The MOPS questionnaire is predominantly tailored to man- ufacturing companies (but it can be easily extended to services and commerce, as was done  Diagnostic tools by the Mexican Institute of Statistics),17 and some of its questions relate to technologies use both by industrial and service firms which include predictive analytics and artificial intelligence. TABLE 2.9 Information covered by the Management and Organizational Practices Survey Group of factors Indicator Framework conditions • Data governance Technology group —  • Use of artificial intelligence in production processes (at the intensive and extensive Informational margin) • Barriers to adopting and using artificial intelligence • Use of predictive maintenance in production processes (at the intensive and extensive margin) Technology group — Operational • Use of selected Industry 4.0 technologies in production processes (at the intensive and extensive margin) Knowledge and awareness • Education level of managers • Education level of workers Managerial practices • Production maintenance • Progress monitoring • Performance remuneration • Talent management (promotion/downgrading) Organizational capital • Decision-making process • Procurement practices • The activity of labor unions • Audit and controlling practices Business performance • Employment size • Balance sheet figures • Business expectations (scenarios for next year, with probabilities) Source: Digitrans team. MOPS and surveys modeled after it have been conducted at least once in almost a dozen coun- tries worldwide, with three waves of the survey in the US. As of 2022, MOPS was implemented three times among 50,000 industrial companies in the United States, and smaller-scale face- to-face surveys modeled after the MOPS were implemented in Pakistan (Lemos et al. 2016) and Mexico (Bloom et al. 2022). Numerous statistical offices, including those in Australia, Canada, Colombia, Germany, Japan, Pakistan, and the United Kingdom, are introducing questions based on the MOPS as part of national censuses (Haltom and Bloom 2014). Thanks to the compre- hensiveness of its questions and the fact that its design allows generating comparable results, MOPS has become a global reference point for analyzing the managerial and organizational practices of enterprises. MOPS is a mandatory sample survey conducted using paper and electronic instruments. It follows a specific mailing strategy for multi-unit firms. Because practices may vary across plants within a multi-unit firm, and information about technology adoption practices may not 17. https://en.www.inegi.org.mx/programas/enaproce/2015/ 61 2 be known at the firm level, the MOPS is mailed to the physical address of the establishment’s   plant rather than to the firm’s business address. If the physical address for the establishment is not available, the firm’s business address is retrieved from the business register, and the survey is mailed there. If the form is returned by the United States Postal Service as “undeliv-  Diagnostic tools erable as addressed,” it is remailed to the firm’s business address (Buffington and Jarmin 2017). Institutions implementing surveys modelled after MOPS in other countries leveraged various forms of survey delivery, including, but not limited to, CATI and computer-assisted web inter- view (CAWI). Almost all questions are mandatory, and although most of them relate to factual data (Table 2.10), the questionnaire also includes inquiries about perception and forecasts. TABLE 2.10 Composition of the Management and Organizational Practices Survey Module Description Mandatory Questions Optional Questions A Management Practices 16 0 Performance monitoring 8 0 Performance-based remuneration 4 0 Promotion and dismission 4 0 B Organization 1 3 C Costs and Purchased Services 5 0 D Data, Decision Making, and Artificial Intelligence 8 2 E Expectations 4 0 F Background Characteristics 9 3 Total 43 8 Source: Digitrans team based on information from United States Census Bureau. A key strength of the MOPS is that, in the United States, it can be linked to Census Bureau data sets on plant-level outcomes. Because the MOPS and Annual Survey of Manufacturers (ASM) samples are identical, MOPS can be linked with near certainty to annual performance data at the plant level, including outcomes on sales, shipments, payroll, employment, and more. Also, matching MOPS to the Longitudinal Business Dataset (LBD) allows longitudinal research on establishment-level management practices (Buffington and Jarmin 2017). The questionnaire is relatively short yet collects high-quality information. Its implementation is straightforward and does not require a respondent with deep knowledge. Although the MOPS survey has been typically applied to manufacturing firms and includes very few questions on technology use, a digital upgrading project can use it in two ways. First, some of its questions about management practices could be leveraged for generating more informa- tion on firms to select target groups. The questions could also be reframed within the context of digitalization, for instance, by asking how many KPIs are used to monitor the progress of a digitalization strategy and how frequently KPIs are reviewed by managers. Second, given its focus on industrial firms, the MOPS questionnaire could be utilized to develop a manufactur- ing-sector-specific assessment. Including some MOPS questions in any survey implemented during the diagnostic phase could provide valuable evidence to inform the design of the pilot instrument. The most pertinent questions include the timeframe for planning, the responsibility for investment decisions, and the share of technological expenses spent on cloud services. 62 2.2 DIGITAL DIAGNOSTIC TOOLS 2   Digital diagnostic tools (DDTs) measure the adoption and use of digital technologies and/or complementary managerial practices by firms in more detail than population surveys do, making  Diagnostic tools it possible to tailor digital transformation plans to individual companies. DDT measurements are typically benchmarked relative to industry, national, or regional peers, with benchmarking indicators being quantitative (such as financial data) or qualitative (such as managerial prac- tices). DDTs are subdivided into ‘light-touch’ and ‘intensive’ (OECD 2020). Light-touch tools are associated with lower cost and limited feedback and guidance, whereas the intensive DDTs are more costly but usually provide more tailored and nuanced feedback and guidance. DDTs are further subdivided into stand-alone and integrated, addressing various policy ob- jectives. Stand-alone tools are typically designed as an entry point to business advisory support and can reach many users because their deployment is low-cost, given that they do not require the participation of external advisors (Figure 2.1) (ibid.). Policy makers can benefit from the scalability of these tools and often combine features of light-touch and intensive stand-alone tools by designing a free and low-time-burden tool as an entry point for advisory services (United Kingdom Department for Business Innovation & Skills 2013). However, firms that undergo standalone DDTs do not receive plans for digitalization and, in most cases, require additional advisory assistance to implement digitalization activities (Ghobakhloo et al. 2012). Standalone DDTs could be thus offered in contexts where existing advisory or technical assis- tance schemes are available to enhance its impact. Integrated DDTs are more comprehensive and require external consultants to use them. They offer more in-depth engagement than either light-touch or stand-alone tools at the cost of higher financial and time burdens and limited reach due to limited advisory staff. Integrated DDTs usually conclude with the creation of a dedicated firm transformation plan that guides future investments. The costs and scalability issues associated with these tools disincentivize policymakers from using them, and they are typically deployed by private sector consulting agencies (OECD 2020). FIGURE 2.1 Categorization of Digital Diagnostic Tools (DDTs) Digital Diagnostic Tools (DDTs) Light-touch Intensive Stand-alone Integrated Stand-alone Integrated predominantly online (e.g. Digital Readiness (e.g. InnoCAPE DMA) connected with methodology self-assessment tools Assessment Tool) for creating transformation (e.g. Digital Maturity plans using assessment data Assessment) (e.g. Digital Quotient) Source: Digitrans team The following part of the chapter analyzes one example of a best practice tool from each of the four types of DDTs. The analysis includes an overview of the tool, including its objectives, target group, assessment details, and costs to inform policy contexts in which it could be deployed. Additionally, the methodology and impacts of each DDT are discussed with insights for inform- ing the design of the Digitrans pilot instrument, specifically regarding the questionnaire and support system available to MSMEs post-diagnosis. DDT impact will be considered in terms of the number of firms that used the tool, the number of firms that have adopted new technol- ogy or managerial practices because of the tool, and performance improvement because of technology or practice adoption. Particular attention is paid to features allowing for periodic reviews of digitalization progress achieved by firms using a particular DDT. 63 2 Digital Maturity Assessment   The Digital Maturity Assessment (DMA) is an example of very quick and streamlined tool, yet effective in structuring thinking about digital transformation among MSMEs. (See Table 2.11)  Diagnostic tools The diagnostic output of this tool is most relevant for needs of firms at (pre)foundational level of digital sophistication but has also potential for filling in gaps in planning digital investments in more developed enterprises. The DMA is effective in capturing attention of enterprises and connecting them to public support instruments offered by an institution managing the tool. The tool can be quickly deployed on the technical side, but it needs strong benchmarking data. Hence, it is mostly applicable to countries where extensive survey on usage of digital technologies have been already implemented. TABLE 2.11 Summary of Digital Maturity Assessment Managing Organization(s) Business Development Bank of Canada (BDC) DDT type Light-touch — Stand-alone Target Users SMEs of any industry at pre-foundational and foundational level of digital sophistication Assessment Style On-line self-assessment questionnaire Assessment Pillars (1) Digital Intensity. This measures the use of digital technologies in the company’s operations along four axes: 1. Customer Experience 2. Processes and Collaboration 3. Data and Analytics 4. Digital Technologies (2) Digital Culture This measures the ability to implement change in the company along two axes: 1. Leadership and Strategy 2. Staff and Expertise Technological Focus Technology agnostic, limited to broad group of widely used digital solutions. Managerial Component 1. Data-driven decision making 2. Business goals strategy 3. Change management strategy 4. Understanding the impact of digital on accomplishing business goals 5. Training/education strategy to cope with change Benchmarking Data The benchmarking data comes from a survey of 1,500 Canadian businesses conducted between November and December 2021. Output Assessment report evidencing needs of a firm in the following dimensions: • Digital intensity – actual use of digital technologies in firm operations. • Digital culture – ability to implement change in firm. Costs Time burden: ~6 minutes; financial burden: none Source: Digitrans team. Note: The tool is available under the following link https://www.bdc.ca/en/articles-tools/entrepreneur-toolkit/business-assessments/digital-maturity. DMA follows the digital maturity model developed by the MIT Center for Digital Business in collaboration with Capgemini Consulting (CC). The MIT and CC understand digital maturity as the combination of the what and the how of an organization, or the digital intensity and trans- formation management intensity respectively (MIT Sloan 2017). Digital intensity, or the what, is 64 2 investment in technology enabled initiatives to change how the company operates its customer   engagements, internal operations, and business models. In other words, it is the set of digital transformations implemented by the organization and the resources used to do so. Transfor- mation management intensity, the how, involves the leadership capabilities necessary to drive  Diagnostic tools digital transformation. In other words, it is the way executives drive digital change throughout the organization. This includes creating and communication visions, establishing governance and measurement mechanisms, and building a digital ready culture. The model resulted from qualitative engagement with executives in digitally leading companies and connection between its implementation and firm productivity was tested empirically.18 For the purpose of the DMA tool, the original model was extended by cloud computing and cybersecurity technologies. Based on a rapid questionnaire, the DMA tool generates report with recommendations follow- ing identified needs and information on best practices. With a total of thirteen questions and an average completion time of six minutes, the time burden is negligible, and the information is easily retrieved by MSME managers. The questions are simple and mostly easily understood, although some would benefit from making definitions of terms such as ‘systems integration’ and ‘digital strategy’ explicit. Covering informational, transactional and operational technolo- gies, it ranges in sophistication from basic collaboration tools, such as video conferencing, to advanced digital technologies, such as AI and 3D printing. Its focus is on technology associ- ated with general business functions, and it covers all general business functions except for sourcing and procurement and quality control.19 By the end of 2021, almost 20,000 firms had used the DMA, with one third of them benefiting from support programs afterwards. The DMA tool automatically generates post-diagnosis re- port which benchmarks a firm’s digital maturity score against national and sectoral averages. Distribution of entities at various levels of digital sophistication among firms performing self-assess- FIGURE 2.2 MSMEs digital sophistication level ment with the DMA tool is similar as results of survey according based on the Digital Maturity Assessment providing benchmarking data (Business Development Techno-shy Advanced High Bank of Canada 2022). Based on answers provided by Canadian firms, six of 10 MSMEs record low scores 14% of firms 19% of firms in both digital culture and intensity (Figure 2.2). The Digital culture ‘Conservative’ firms are the main target group of ad- Medium visory services offered by the Business Development Bank of Canada. Services offered by the Business Conservative Technocentric Development Bank of Canada provide firms with expert support on i) defining a digital strategy that 57% of firms 10% of firms prioritizes projects and investment with the highest Low returns, ii) selection of technologies, and iii) choos- ing the best systems and vendors. The assessment Low Medium High report generated by the tool is used as a source of information for accessing support offered by the Digital intensity Business Development Bank of Canada. Source: Business Development Bank of Canada 2022. 18. Authors of the methodology surveyed senior executives in 184 publicly traded companies and determine their digital transformation level. Based on calculations of financial data and controlling for industry and geography, (i) companies with stronger digital intensity derive more revenue from their physical assets, (ii) companies with stronger transformation management intensity are more profitable, and (iii) companies with stronger transformation manage- ment intensity achieve higher market valuation (MIT Sloan 2011). 19. Following Cirera, Comin, and Cruz (2022), we identify seven general business functions: (i) administration (HR, finances); (ii) production or service operations planning; (iii) sourcing and procurement; (iv) marketing and consumer information; (v) sales; (vi) payment methods; (vii) quality control. 65 2 The main advantage of the DMA is its ease of use and strong grounding in evidence, with its   limited managerial component being the shortcoming. The uptake of a diagnostic tool among MSMEs depends on how user-friendly it is, particularly with respect to the time required to input the information and how easily the information required is retrieved. The DMA questionnaire  Diagnostic tools succeeds on this front. With a total of thirteen questions and an average completion time of six minutes, the time burden is negligible, and the information is easily retrieved by manag- ers. However, benchmarking data is limited only to Canadian firms and comes from a survey implemented in 2017, while during COVID-19 pandemic many firms accelerated their digital transformation (Kuriakose 2022). The tool also does not sufficiently diagnose barriers to adop- tion. The DMA only assesses internal resistance to change and lack of technology awareness. It would be useful for advisors and policymakers to know what is preventing digitalization in a broader context so that they could make more targeted efforts. This DDT is also lacking in assessing change management. The DMA, or a tool with similar features, could be leveraged to raise interest of firms un- aware about public support programs, using existing data on technology adoption by local firms. An inherent trade-off exists between ease of use and depth of assessment. Tools with very low time burdens, such as DMA, assess fewer aspects of a business than those with higher time burdens. As such, for light-touch self-assessment tools, the benefit of a high completion rate comes at the cost of not assessing some aspects of digital maturity. But this can be compensated for post-diagnosis via advisory services that can run more in-depth analyses. The DMA questionnaire is concise, but together with its diagnostic re- port, it provides users at lower levels of digital sophistication a tangible understanding of key concepts of digital transformation. The survey could be quickly deployed in countries where FAT surveys or similar projects were already implemented, as it results could con- stitute benchmarking input for the tool. It is also possible to leverage the CSI as base for benchmarking, implemented annually across the EU. Such approach would guarantee that information is up-to-date and comparable across countries, although it would provide less granular insight than FAT. Digital Readiness Assessment Tool The Digital Readiness Assessment Tool (DRAT) provides close integration between the DDT and follow-up support services. It focuses on the digitalization of business processes, not on technologies per se, assessing managerial practices along use of technologies (See Table 2.12). It is rapid, with streamlined questions accompanied by a standardized set of multiple-choice answers to all of them. Such construction of questionnaire limits adequacy of most imple- mented surveys as a benchmarking data, as they focus predominantly on assessing firms’ familiarity of technologies. Output of this diagnostic tool is directly aligned with dedicated support instruments, as providers of services for MSMEs are trained to integrate DRAT’s output into technology or managerial support services they offer. TABLE 2.12 Summary of Digital Readiness Assessment Tool Managing Organization(s) Australian Government — Business Entry Point Initiative DDT type Light-touch—Integrated Target Users MSMEs of any sector at foundational digital maturity Assessment Style On-line self-assessment questionnaire 66 2 Assessment Pillars 1. Customer   2. Operations 3. Data and Analytics 4. Technology  Diagnostic tools 5. Risk, Privacy, and Cybersecurity 6. Digital Capability and Culture 7. Digital Strategy and Innovation Technological Focus Technology agnostic. Assesses state of digitalization among respective business processes, not type of technologies used. Managerial Component In the data and analytics section: 1. Data-driven decision making In the digital capability section: 1. Hiring practices (value of digital skillset during hiring) 2. Technology and digital priorities among business leaders 3. Collaboration practices 4. Staff technology education In the digital strategy and innovation section: 1. Digital strategy practices 2. Awareness of impactful technology 3. Investment in technology Benchmarking Data Before launching, a sample of over 800 SMEs representing a variety of business characteristics and industry sectors used the tool, and this data was used as the baseline for the benchmarking information. This data is updated as more businesses use the tool. Output Dedicated digital transformation plan constituting the basis for applying for a grant for investments in digital technologies. Costs Time burden: ~10 minutes; financial burden: none Source: Digitrans team. Note: The tool is available under the following link https://business.gov.au/online/what-is-your-business-digital-maturity DRAT is a non-sector-specific DDT assessing the level of digitalization of general business processes instead of cataloguing usage of particular technologies. DRAT was created on the basis of a collaboration with Ernst & Young Sweeney, an Australian market research firm, by adopting a preexisting digital maturity framework for the DRAT. The framework includes seven key pillars of digital transformation on the following dimensions: customer, operations, data and analytics, technology, risk/privacy and cybersecurity, digital capability and culture, digital strategy, and innovation. The tool is free to use on the website of the Australian government and generates customized reports benchmarking a firm’s digital readiness against other Australian businesses and a list of recommended actions, with links to support services administered by the government-funded nongovernmental organization Navii. Navii offers several additional services ranging from 10 to 600 USD each, such as the Digital Engagement Health Check and One Hour Private Consultation (Box 2.1). The DRAT questionnaire asks whether firms have digitalized certain functions or processes, are up-to-date on the latest technologies in their industry, or have a proven track record in effectively deploying new technologies. Although these are important considerations and add value to a DDT questionnaire, benchmarking digital technology adoption and use without assessing how their current technology compares to a technology frontier is incomplete. De- spite a similar time burden, DRAT has significantly more questions, 35 compared to the DMA’s 13. This suggests that the trade-off between ease of use and assessment depth is not very pronounced, and significant gains in assessment depth can be made for little loss of ease of use. Unlike the DMA, DRAT uses the same Likert-type multiple choice answer options for every 67 2   BOX 2.1 Post-diagnosis support DRAT The Next Best Steps Program. Next Best Steps is a fully the coach can control the client’s mouse and fix issues subsidized, self-paced online course over a 5-day period on the spot. The cost of the session is USD 175, mostly  Diagnostic tools that provides firms with fundamentals for building vis- covered by the Navii. ibility on the internet, managing the process of website Small Business Advisory program—Round 1. The Small development, leveraging their presence in social media, Business Advisory program is a two round digitalization and utilizing basic software for business process optimi- effort that focuses on both SMEs and advisory services. zation and automatization of repetitive tasks. Round 1 is a 7-hour packaged service that offers one-on- The Digital Engagement Health Check (USD 225) is a one tailored support as well as group workshops and very detailed assessment of a firm’s online presence. webinars for small businesses with fewer than 20 full- This assessment is conducted by Navii’s digital strate- time employees. It costs 44 USD for the 7 hours, and the gists, who review each of the firm’s assets manually. It first interaction is free. The Digital Solutions advisers assesses the firm’s website, Facebook, Instagram, and hold formal qualifications in business or information Google My Business listings. Firms will receive a report technology related disciplines and have at least 2 years’ of about 16 pages with expert recommendations for experience providing digital advice to SMEs. every point they did not pass, so they can improve their Small Business Advisory program—Round 2. Round 2 is digital presence immediately. The cost of the service is a grant opportunity for digitalization advisory service USD 225, and the majority of it is covered by Navii. providers. The aim is to help advisory services offer low- The One Hour Private Consultation sessions give firms cost, high-quality services to SMEs across five priority the opportunity to meet one-on-one with a practicing capabilities: (i) business digitalization, (ii) websites and small business strategist with experience in entre- e-commerce, (iii) social media and digital marketing, preneurship, digital, and coaching. The consultation (iv) use of small business software, and (v) cybersecu- is tailored to achieving a particular objective of the rity and data privacy. Grant funding consists of USD 15 firm. The session is very hands on, to the point where million across three coverage areas over four years. Source: Digitrans team based on documentation of Navii support programs. question: five options ranging from strongly disagree to strongly agree (agreement range). The DRAT questionnaire includes questions on business location, making it possible to direct a firm to business support institutions active in its area of operation. DRAT includes useful features for integrating assessment of technological and managerial fac- tors; it also is a best practice in integrating automated diagnostic report with post-diagnostic support instruments. DRAT assesses more management practices than the DMA. Although the two instruments are similar in assessing digital strategy, DRAT provides more insights for assessing digital capabilities and culture. For instance, DRAT asks whether digital skillsets are a key consideration in recruitment, and whether technology and digital innovation are among the top priorities of leaders. However, like the DMA, the DRAT insufficiently addresses barriers to adoption. Barriers to adoption are key considerations when assessing change management but also have the benefit of implicitly assessing barriers to digitalization, namely, internal resis- tance to change from business leaders. The DMA’s method of varying the answer options with the question is a more efficient system. For instance, the DMA can assess cybersecurity best practices in a single question because it can list the practices as answer options, whereas DRAT must use several questions, each referring to a particular practice. InnoCAPE Digital Maturity Assessment The InnoCAPE Digital Maturity Assessment (DMA) caters to the needs of firms at the pre-DIH stage. It assesses both behaviors and current processes as well as usage of particular technological solu- tions (See Table 2.13). The InnoCape DMA puts strong focus on ensuring common understanding of key digitalization concepts by MSMEs, as ambiguity in this area was identified as main shortcoming 68 2 of existing DDTs and inspired establishment of the InnoCape consortium. Based on the result of   self-assessment, the InnoCAPE DMA connects firms to the marketplace of curated providers of support services and implementation partners; however, without links to individual public support programs. Downside of this tool is that it does not include features for tracking progress of firm  Diagnostic tools digitalization over time. Similarly, its output presents limited usefulness for post-diagnostic support. TABLE 2.13 Summary of InnoCAPE Digital Maturity Assessment Managing Organization(s) InnoCape consortium (8 Digital Innovation Hubs and 3 public authorities from North European countries). DDT type Intensive—Stand-alone Target Users MSMEs of any sector at foundational digital sophistication Assessment Style Mix of off-line self-assessment questionnaire and consultant-led interview Assessment Pillars • Digital Technology Integration • Organization and Leadership • Strategy • Digital Innovation • Digital Analytics Technological Focus Technology agnostic, limited to broad groups of technologies following typology from the ICT use and e-Commerce survey. Managerial Component Managerial Component in Organization and Leadership: A. Support and encouragement of digital innovation B. Collaboration practices for digital innovation C. Employee involvement in achievement digital strategy Managerial component in Strategy: 1. Strategic value of digital technology 2. Digital vision 3. Strategic Investments in digital technologies and digital skills Managerial Component in Data Analytics: Strategic importance of data analytics Benchmarking Data Approximately 250 firms from countries in the Baltic Sea region. Output Tailored recommendation plan with list of providers addressing identified gaps. Costs Time burden: 20–40 minutes; financial burden: none Source: Digitrans team. Note: The tool is available under the following link http://dma.innocape.eu/en Several insights on questionnaire design can be derived from InnoCAPE’s methodology and experience in piloting the DMA. First, it is important for conceptual grounding to be supple- mented by empirical insights — authors of the tool modelled questions not only on review of literature but integrated also what firms themselves say they need to transform digitally. The final form of the questionnaire was determined after testing proof of concept in practice through surveys and interviews with MSME managers. Second, the reluctance of managers to fill out lengthy questionnaires entails the need to prioritize certain questions. InnoCAPE’s DMA has been prioritized based on research and input from stakeholders. Third, there are perils in creating an assessment tool by using questions from several existing tools. Different tools may have different conceptual groundings and, as such, will measure different phenomena. It is important to make sure different tools measure the same thing before cherry-picking ques- tions from them. It is essential to make definitions clear to the users. Fourth, involve software developers early. InnoCAPE failed to create a maximally user-friendly interface because they 69 2 used two different developers for the DMA and Competence Map, and the respective instru-   ments were incompatible and could not be hosted on the same interface. The impetus for developing the concept came from identification of gaps in other DDTs and  Diagnostic tools addressing what InnoCAPE developers saw as their shortcomings. The InnoCAPE consortium, a group of public institutions and digital innovation hubs, found that many DDTs were not operating with a clear logic and were assuming that central concepts, such as digital maturity, were being TABLE 2.14 List of partners in InnoCAPE consortium understood the same way by both tool developers and users. (See Table 2.14 regarding the consortium.) Latvian Information Technologies Cluster Latvia Clearly defining concepts, developing assessment Nozaru politikas departaments pillars in relation to those concepts, and having their Sunrise Valley Science and Technology Park definitions made explicit on the questionnaire is how Lithuania InnoCAPE overcame this problem. The InnoCAPE Agency for Science, Innovation and Technology consortium’s method for developing the question- University of Tartu naire was a three-step process. First, developing a Estonia Tartu City Government conceptual grounding as the basis of the tool. This Seinajoki University of Applied Sciences grounding defines explicitly the notion of digital matu- Finland rity and other digitalization concepts such as industry University of Oulu 4.0, which allows for a clear idea of what to measure RISE Research Institutes of Sweden AB in MSMEs. Second, a literature review based on the Sweden Umea University conceptual grounding. Third, compromising with stakeholders (DIHs, policy makers and companies) Norway DIGITALNORWAY on what to include in the questionnaire. The whole design process was organized in participatory manner, Source: Digitrans team. facilitating collaboration between institutions in the consortium through clear division of responsibilities, multilateral validation, and pilot projects. It could be leveraged to organize design processes of instruments managed by multiple institutions. Post-diagnosis, a benchmarking report is generated comparing results to industry peers across the country and the Baltic Sea Region. Companies can use these results to initiate and manage internal discussion regarding the status of digitalization and the general feedback provided by the report on how to improve along specific dimensions. The post-diagnosis report could also be used to start a dialogue with a local DIH. This is useful to firms because DIHs can provide more tailored and nuanced support by connecting companies with digital expertise, facilitating networking, offering test-before-invest infrastructure, finding financing opportunities, and other support services. Moreover, DMA results can be used to find expertise through the InnoCAPE Competence Map. The Competence Map is a platform for finding digital service providers, dedicated to easing the everyday work of DIHs, although it can be used by MSMEs directly. InnoCAPE DMA is an example of a tool that was designed to be used before accessing DIHs, to cater the needs of firms at pre-foundational level. After the creation of a digital strategy, it allows advisors to easily find providers to meet the needs of their MSME clients and deepen the diagnos- tics. Building a platform like the Competence Map is a useful addition to DIHs in the pilot countries. However, it is important to have the developers of the Digitrans DDT and the provider platform collaborate to have an optimized user experience. The broad and diverse consortium and good work organization could guide implementation of other digitalization projects as well. InnoCAPE’s DMA does not track impact data. This is a common shortcoming of DDTs, and one that the DIH network is in the process of correcting by creating a new DDT that will assess the impact DIHs have on firm digitalization by re-assessing firm digital maturity after the intervention of a DIH.20 70 20. Digital Maturity Assessment for EDIH Customers. 2 Digital Quotient   Operated by a private consulting firm, Digital Quotient (DQ) focuses almost exclusively on complementary capabilities for digitalization, without explicit analysis of technological factors.  Diagnostic tools The Digital Quotient assesses strategic, cultural, and organizational capabilities for advancing in digital transformation of firms (See Table 2.15.) Its questionnaire is applicable mostly to medium-sized and large firms, although it could be adapted to increase relevancy for smaller organizations. After the initial engagement with the enterprise, a diagnostic report is produced by a technical expert, which creates a dedicated transformation plan that leverage best practices in addressing needs identified during the diagnosis. Despite its high unit cost, its streamlined version is co-financed under public support interventions in Singapore21 and South Korea.22 TABLE 2.15 Summary of Digital Quotient Managing Organization(s) McKinsey Digital DDT type Intensive—Integrated Target Users Medium-sized and large firms at intermediate level of digital sophistication Assessment Style Series of fully consultant-led on-site workshops 1. Organization and Talent 2. Agile Delivery and Culture 3. Capabilities 4. Strategy Assessment Pillars 5. Adoption and Scaling Technological Focus Technology agnostic, limited predominantly to medium and highly advanced solutions. • Management Practices in Organization and Talent: • Management practices in Agile Delivery and Culture • Management practices in Capabilities: a. Technology b. Data • Management Practices in Strategy Managerial Component • Management Practices in Adoption and Scaling Benchmarking Data Over 750 organizations globally operating in various business sectors. Benchmarking report, gaps analysis, and dedicated transformation plan with an outline of Output implementation arrangements. Time burden: ~40 minutes to fill out the assessment (30–40 senior leaders complete the self- assessment) followed by 45–60-minute interviews; actionable results delivered within 3–4 Costs weeks; financial costs vary. Source: Digitrans team based on information from McKinsey. Note: More information about the tool is available in Catlin T., Scanlan J. and Willmott P. (2015). Raising your Digital Quotient. McKinsey Quarterly June 2015. 21. Operation & Technology Roadmap operated by Agency for Science, Technology and Research Singapore (A*STAR) provides grants up to SGD 70,000 (USD 35,000) for small and medium-sized enterprises to create digital transforma- tion plan with support of one of pre-selected providers. McKinsey is one of them, along with, e.g., Accenture, Bain & Company, EY or PwC. Implementation costs fall fully on the side of beneficiary. 22. Creation of digital transformation roadmaps by the biggest global management consulting companies in advancing the digital maturity of SMEs is also co-financed by the Korean SME and Startups Agency under the Business Health Checkup & Technical Assistance program. 71 2 The DQ diagnostic measures digital maturity across management practices that correlate with   performance. It emerged from a two-year study of hundreds of companies around the world. It began as an assessment of four key areas and 17 management practices, and now it is an assessment of five key areas and 32 management practices. The DQ has evolved in accordance  Diagnostic tools with the latest research from McKinsey and other institutions investigating digital maturity and how it correlates with performance (Dietz, Vinayak, and Lee 2017). The DQ assesses managerial practices most correlated with successful digital transformation (Figure 2.3). Although some of these practices might not be relevant to the needs of MSMEs, there are some interesting practices assessed that could be applied to an MSME context, like linking the digital and busi- ness strategies and measuring the impact of digital adoption. FIGURE 2.3 Components for calculating Digital Quotient score Culture Strategy Capabilities Org & Talent The mindsets The vision, goals and The systems, tools, The systems, tools, and behaviors strategic tenents in place and technology in and technology in critical to to meet short-, mid- and place to achieve place to achieve capture digital long-term digital strategic digital strategic digital opportunities business aspirations goals goals Source: McKinsey Digital. The DQ helps ensure the accuracy of the data it collects on digital maturity by having 30 – 40 senior leaders complete the assessment. The quality of the data gathered can be further im- proved by (i) increasing the number of test takers, (ii) including an interview with the manager’s component after self-assessment, or (iii) making the Digital Quotient (DQ) an expert-led tool instead of a self-assessment to allow the interview to happen at the time of the assessment. Additionally, relevance of Digital Quotient as a diagnostic tool is evaluated against firm busi- ness objectives — DQ maturity scores directly correlate with digital and financial performance (Catlin et al 2015). The DQ tends to be part of a greater digitalization engagement where McKinsey Digital lever- ages a suite of digital maturity assessment tools, and the client is continually guided by an expert to capture value from digitalization. Although McKinsey’s clients are larger than the intended target users of Digitrans, two insights can be gained here. First is the importance of having expert guidance post-diagnosis: even large firms—who tend to be more digitally savvy and motivated—require expert assistance to embark on a successful digital journey. Second, this tool emphasizes the importance of having services or tools available post-diagnosis to further assess a firm’s digital maturity. 72 3 PROGRAMS AND INSTRUMENTS 3 KEY LESSONS FROM CHAPTER 3   The chapter describes comprehensive examination of the difficulties faced by firms in digital transformation, and the potential solutions that  Programs and instruments can help them bridge the gap and successfully adopt new technologies. The challenges that firms face in adopting digital technologies are cate- gorized into three distinct groups: informational constraints, preferences and behavioral biases, and organizational misalignments. By classifying these issues into separate groups, the chapter aims to provide a struc- tured approach to understanding and addressing the myriad of obstacles that may hinder a company’s digital transformation journey. Informational constraints refer to the limited awareness that firms might have about the benefits of upgrading their technology, the ways in which they can enhance productivity, and the potential failures that can occur in the supplier market. These factors can impede the adoption of digital technologies, as they prevent companies from making informed decisions about their investments in new technology and innovation. Preferences and behavioral factors encompass the economic motivations and biases that influence decision-making, as well as the resistance that workers may exhibit towards the adoption of new technologies. These aspects can hinder a company’s ability to integrate digital solutions into their operations, as they impact the willingness of both management and employees to embrace change and adapt to new ways of working. Organizational misalignments include the low digital human capital of workers and the inadequacy of managerial practices in supporting the implementation of digital technologies. These factors can inhibit a firm’s capacity to fully benefit from new technologies, as they contribute to the skill gap and organizational challenges that may arise during the digital transformation process. The chapter also presents a range of policy interventions that can be implemented to address each of these challenges, with an emphasis on the strengths, potential drawbacks, and risks associated with each type of intervention. Additionally, the chapter highlights key design elements that should be taken into consideration when devising policies to facilitate the adoption of digital technologies by firms. In this chapter, we move to the analysis of the main challenges firms face when adopting dig- ital technologies, as well as a discussion of programs and instruments to be used to tackle these challenges. We divide these challenges into three groups: informational constraints; preferences and behavioral bias; and lack of capabilities. For each challenge, aside from an introductory description, we discuss: what types of firms are more likely to face this challenge; what is a potential sequence of interventions that could be implemented to tackle it; examples of existing programs that could implement such interventions; evidence of impact of inter- ventions; strengths of suggested interventions; their potential drawbacks and risks; and key intervention design elements that should be taken into account. 74 3.1 INFORMATIONAL CONSTRAINTS 3   Information asymmetries are relevant factor behind the failure of firms and individuals to adopt technologies. The lack of information could be on the existence of the technologies, their features,  Programs and instruments or on the markets surrounding them. The discussion below breaks down these informational con- straints into three specific challenges: lack of awareness of possible technological upgrading; lack of knowledge on enhancing the productivity of digital investments; and supplier market failures. CHALLENGE 1 Lack of awareness of possible technological upgrading For any public policy aiming at improving the adoption of digital technologies, the first barrier that needs to be overcome is the informational one. Although the potential returns of adoption may be high, firms may lack the necessary information to make the most efficient adoption decisions. This lack of information can occur at five levels, each of which would make a firm wary of proceeding with adoption: 1. Firms may be unaware of their own weaknesses (Bloom and Van Reenen 2010; Cirera, Comin and Cruz 2022); 2. They may be unaware of the existence of certain technologies (Cole and Fernando 2020; Gupta, Ponticelli, and Tesei 2020); 3. Firms may be unaware of the returns to the adoption of those technologies (Foster and Rosenzweig 2010); 4. Firms may lack information on how costly it is to adopt certain technologies (Foster and Rosenzweig 2010); 5. Firms may lack information about who can help them identify the appropriate technologi- cal solutions to address their needs and be concerned about the validity of external advice because of asymmetric information problems (Anderson and McKenzie 2021). H. Primary target groups Younger and smaller firms: due to being in operation for less time than their peers, younger firms are more prone to have less information about digital technology possibilities. And smaller firms (a group that overlaps with younger firms) often do not have the resources — especially the bandwidth — to be attuned to technological developments (Anderson and McKenzie 2021). These firms may lack more fundamental information about the existence of technologies and their costs and returns (levels 1, 2, and 3). Firms in less dynamic regions: firms located in larger cities or in clusters of firms performing similar activities will tend to have increased access to information regarding technologies that could be profitably adopted (Saxenian 1994; Audretsch and Feldman 1996). Firms in regions with lower economic activity might be especially prone to lacking information about who can help them identify technological solutions (ibid). Thus, a program that seeks to ameliorate firms’ lack of awareness of digital technology should pay special attention to companies in more remote areas and those that do not have as much access to peer firms or to a thick la- bor market because such companies are more at risk of being cut off from information flows. 75 3 I. Potential sequence of instruments23   1 3 4 Distance technology Web-based technology Showcasing examples of individua learning demonstration successful transformation  Programs and instruments Tackling this challenge mostly requires informational interventions. They can become more intensive as the firm improves its awareness of potential avenues for digital upgrading. The first measure would be a light intervention at a distance to acquaint the firm with the digital possibilities. If firm management requires more information, this could be followed by more extensive informational engagements, which could include an in-person element that illus- trates successful examples of more specific digital adoption routes the firm might be consid- ering after the initial intervention. J.  Examples of interventions Acelera Pyme Offices and Portal Acelera Pyme/Spain: Acelera Pyme Offices (OAPs) are physi- cal and virtual points for the advice and digital push of SMEs, freelancers, and entrepreneurs. OAPs provide general information consultancy and organize awareness campaigns about the importance of digitalization. The program was created in the context of the COVID-19 pan- demic in 2020 to reinforce the previously existing Digital Transformation Offices (Oficianas de Transformacion Digital) program, with a total budget of 6.5 million euros. These funds are distributed to professional associations, technology centers, and extension centers, who can apply to establish OAPs, with the maximum and minimum number of projects accepted per region varying to maximize firm and territory coverage. Each accepted project receives between 200,000 and 250,000 euros, and each project budget will receive up to 80% in funds, with the beneficiary funding the remaining 20%. While such a program can, in principle, help with all four levels of informational issues alluded to above, it is most effective for level 1 (lack of awareness of technologies) and 3 (lack of information on cost of adoption). Industrial Development Plans (IDPs)/Singapore: The IDPs provide SMEs of various sectors with an easy-to-use, step-by-step guide on the digital solutions to adopt at each stage of growth and training programs to enhance employees’ digital skills. The sector-specific IDPs, which were created by sector lead agencies, are documents available online that lay out the digital possibilities at each step of the sector’s value chain. It describes the digital solutions available and the associated benefits, as well as the digital skills needed both for any employee in the industry and for those that would use advanced technology more directly. K. Evidence of Impact Informational interventions often have mixed or no significant impact on firms in develop- ing countries contexts, whether the objective is increasing formalization (De Mel, McKenzie and Woodruff 2013; de Andrade, Bruhn and McKenzie 2014; Benhassine et al. 2018) or usage of professional business services (Anderson and McKenzie 2021), for example. Conversely, positive effects on the adoption of agricultural practices have been found (Casaburi, Kremer, Mullainathan and Ramrattan 2014; Cole and Fernando 2016), especially when information is timely and personalized — in this sense, the provision of information through mobile phones 23. Numbers of instruments refer to Figure 1.4 in section 1.2. 76 3 has been found to be a good solution (Gupta, Ponticelli and Tesei 2020). It is unclear how these   results translate to the context of the adoption of digital technologies by MSMEs, though one can reason that positive effects hinge on at least two factors (McKenzie 2013).  Programs and instruments First, it is necessary for information to be costly to acquire. Successful informational interven- tions in development contexts were for problems for which the information being provided was much harder for individuals to acquire on their own, even if they wanted to. One example is work in informing students in the Dominican Republic about the returns to education where information acquisition is difficult because the successful individuals left the rural communities so that their information was lost to those who remained (Jensen 2010). Another is work on informing girls in Kenya about the relative risks of having sex with older vs. younger men where the difficulty is the sensitive topic with no obvious source for students to get this information (Dupas 2011). It remains to be seen whether information on the adoption of IT technologies for MSMEs falls under this category. It seems safe to assume that the difficulty in information acquisition increases with the complexity of the technology. Secondly, information needs to be directly useful. Informational interventions to increase busi- ness formalization in Sri Lanka (de Mel, McKenzie and Woodruff 2013), Bangladesh (de Giorgi and Rahman 2013), and Brazil (de Andrade, Bruhn, and McKenzie 2014) showed zero effects, despite the severe lack of information about the process and costs of formalization across firms in the three countries. Firms felt no need to pursue more information on formalization — if they felt that need, they probably would have been able to easily do so. L. Strengths Relatively easy to implement: programs like these implemented in Spain are relatively easy to setup. For the (low) level of technologies targeted for adoption, most informational content can be standardized and, thus, easily assembled and displayed on an online platform. Lower administrative and compliance costs than direct support instruments: relatedly, after an initial setup cost, overhead costs should be low going forward, especially because such programs can rely on providing basic information online  —  and such basic information often will be enough, given the lower technological levels of target firms. M. Potential drawbacks and risks Difficulty of reaching the intended target group: the flip side of these programs’ lower fixed and variable costs is that they may not reach the intended targets if they take a passive stance and wait for firms to find them. Firms may face the same challenges they have in gathering in- formation about digitalization when gathering information about such programs. Furthermore, taking an active stance and proactively seeking out firms to close the information gap could significantly increase the program’s operational costs. Limited effectiveness for firms facing multiple challenges: such programs strongly rely on lack of awareness being an accurate diagnostic of the main impediment to digitalization. That is, it assumes that after providing information, the adoption of digital technologies will follow in a straightforward manner. However, if the company is facing other challenges (e.g., related to capabilities or internal incentives), these programs will stop short of providing further assistance. 77 3 N. Key design elements   Low intensity versus high intensity: information can be provided via informational leaflets, websites, or text messages (Cole and Fernando 2021). Alternatively, it can be provided via  Programs and instruments higher-intensity procedures, which can include personal visits to firms (Benhassine et al. 2018). To the extent that the information to be passed on is of low complexity, the low-intensity pro- cedures often yield the best cost-benefit ratio. Yet, high-intensity channels facilitate future engagement with target firms, either for follow-up evaluations or for additional interventions. Another choice is the issue of whether information should be presented in a qualitative or quantitative format or in a mixture of both (McKenzie 2020). Standardized versus personalized: another form of low-intensity information provision is standardizing the information so that all target firms receive the same content. This obviously has cost advantages, especially when considering that the information can be sent in bulk to groups of firms. Again, this option is feasible in contexts of lower complexity of the informational content. Personalized advice is preferable when the context combines a higher complexity of information with target firms being at very incipient levels of digital technology adoption, with managers and employees possessing low levels of digital literacy. In the context of extension services for farmers, research has found that personalized information tends to be more ef- fective (Anderson and Feder 2004). Virtual portals tend to have standardized information, as is the case with the IDPs in Singapore. In person versus online: program designers have to decide whether they will establish a physical presence with the creation of local offices or if the program will remain largely online. Aside from being more costly, physical offices also require policymakers to decide where to locate, a decision that is not straightforward given the need to balance reaching out to more isolated firms with being at a location with sufficient scale. On the other hand, physical offices allow for organizing seminars and other in-person events, which can go a long way toward making connections with firms. The IDP program in Singapore offers only an online version, while the Acelera program in Spain contains both in-person and online options. Experts versus role models: in a higher-intensity program design that involves information being provided in person, this can be done either by an expert (i.e., a business consultant) or by a role model (a manager at a firm that has successfully adopted certain digital practices). One important consideration is which choice is associated with greater credibility for partici- pating firms (McKenzie 2020). The first option may be preferable if there is a low availability of role models, either because there are few peers that qualify or because they are not willing to participate. However, it is probably more costly, and role models may generate a greater response from target firms (Oster and Thornton 2012). Passive versus active outreach: an important design choice is whether the program will adopt a passive strategy when it comes to ameliorating information asymmetries related to digital adoption, waiting for firms that have an incipient interest to reach out, or whether it will try to identify and reach out to firms that are prone to be facing such informational challenges. Information versus implementation: program design should also consider whether it will stick primarily to tackling informational issues and trusting that firms will take the necessary steps to adopt technology after this challenge has been addressed or whether the informational inter- vention is the first step in a more intensive process of support. Work in the business formalization literature has found that the provision of information is insufficient, and only more extensive engagement in the form of advisory, training, and support has significant effects (Benhassine et al. 2018). Because these informational interventions are usually part of a broader ecosystem 78 3 of programs, they will usually be connected to implementation efforts, either directly within   the program (Acelera) or by referral to companion programs (IDPs).  Programs and instruments DO DON’T • Even if the informational intervention • Avoid overly relying on informational is of a passive nature, where it does interventions when information can not try to identify specific firms to be easily acquired and there is a low provide information for, take a proac- demand for it by firms. tive stance in publicizing the informa- tional content of the intervention to as many firms as possible. Partner- ships with industry associations can be helpful in this regard. • Consider how funding schemes can simultaneously be financially sustain- able and inclusive of firms with fewer resources. CHALLENGE 2 Lack of knowledge on enhancing the productivity of digital investments After an initial round of adoption of digital technologies, firms may get stuck when taking the next step (Zolas et al. 2020). They may not have the knowledge required to deploy these new technologies to their full potential. Alternatively, they may be unaware of the potential synergies in the joint deployment of different technologies (Brynjolfsson and Milgrom 2013). Therefore, the initial informational intervention may not be enough: programs may need constant follow-ups with adopting firms to provide more information. This would potentially require more robust advisory services, often called technology extension services. These extension centers are business advisory services oriented toward implementing techno- logical solutions such as digital technologies. They entail direct on-site assistance to MSMEs through extension staff, field offices, or dispersed technology centers to foster technological and knowledge-based modernization to improve the competitiveness of firms. Technology extension services thus can involve longer-term and more systematic engagements with MSMEs, for instance, through formal continuous improvement programs (Cirera et al. 2020). A. Target groups Previous digital adopters: by definition, firms that have already made some progress in their digitalization trajectory should be the ones facing the challenge of enhancing the productivity of digital investments. Still, while one would expect that non-users of technology would be more likely than users to report impediments to adopting technology, research has found the opposite (Baldwin and Lin 2002). While users are engaged in learning by doing, they face various problems and must overcome them, with impediments being discovered during the adoption process (Baldwin 2002). 79 3 Smaller and older firms: recent research has shown that when it comes to the adoption of   advanced digital technologies, (1) the smallest firms have the lowest use rates, and the use rates tend to increase with size, even when controlling for firm age, and (2) for small firms, use rates tend to decline with age, with the oldest small firms having the lowest adoption rates in  Programs and instruments general (Zolas et al. 2020). Still, firms that rely on technology extension services to tackle this challenge tend to be larger than firms in more incipient stages that rely more on traditional business advisory services (Cirera et al. 2020). B. Potential sequence of instruments 9 14 15 Technology demonstration Technological Individual technology off-site testbeds learning This sequence follows a similar logic as the initial informational intervention but for firms that are further along in the process. These firms are aware of initial possibilities for digital upgrading but less informed about optimizing the first steps they have taken in this direction. As such, this sequence includes interventions (such as testbeds and experimental development programs) targeted at firms at more advanced technological stages. C. Examples of interventions Mittelstand 4.0 Competence Centres/Germany: This program offers a network of digital hubs, named competence centers, that work as competent and provider-neutral contact points for information, awareness-raising, and qualification about adopting digital technologies. In these centers, MSMEs can benefit from digitalization through practical examples, demonstrations, information events, and the exchange of experiences. Each Mittelstand 4.0 Competence Centre has its own focus area. The spectrum ranges from smart production to new requirements relating to the workplace and IT security organization to legal matters. Support provided by the Mittelstand 4.0 Competence Centres to MSMEs includes: 1. Information, qualification, and training about digital transformation. 2. Testing new solutions in practice. Centers are founded and run by individual consortia. Each consortium includes research institutions (such as universities and Fraunhofer institutions) and transfer partners (such as chambers of commerce and industry as well as chambers of trade) who are familiar with the special needs of MSMEs. ACTIVA Financiacion/Spain: This component aims to support the incorporation of knowledge, technologies, and innovations targeting the digitalization of processes and the creation of tech- nologically advanced products and services with greater added value (complex digital solutions). The support is categorized as: • Industrial research, • Experimental development, or • Organizational innovation and innovation processes in MSMEs. 80 3 D. Evidence of Impact   Most evaluations of technology extension services were of manufacturing services from OECD countries. Overall, studies find that moderate but highly customized services in product devel-  Programs and instruments opment and marketing led to greater benefits. Routine services focused on quality and process improvement achieved less significant benefits (Cirera et al. 2020). Furthermore, there tends to be a constant tension between pressure to increase coverage (serve more firms) and the demand to provide customized services (Shapira, Youtie, and Kay 2011). However, unlike for business advisory services, evidence of the impact of this type of service is scarce, and more impact evaluations in this area are needed (Cirera et al. 2020). E. Strengths One-stop focal point: although large digital hubs, which cluster technology creators, integra- tors and advisors, can help companies along different stages of their digitalization processes, they can be particularly useful for those that are looking to take a bigger step forward, given that they serve as a reference point for the digitalization experiences of many companies, and thus are in a good position to share knowledge. Geographical specialization: within the network of digital hubs, each unit can specialize in the sec- tors (and associated technologies) that are relevant to the regions in which they are located. Unlike general-purpose technologies, specific technologies tend to be associated with specific industries. Thus, such geographical specialization will closely match the incidence of industries across regions. Synergies with other stakeholders: by incorporating other actors, such as universities and technology transfer agencies, these programs are better able to incorporate lessons from different areas of the economy, as well as serve as a hub for spillovers (Koch and Simmler 2020). Complementarity with skill-based programs: often, centers can provide or refer to appropri- ate skills training for specific technologies, thereby narrowing the skills gap that many firms experience when trying to adopt more sophisticated technologies F. Potential drawbacks and risks Complexity of intervention: creating such digital hubs can be expensive. They also require extensive organizing efforts, which can take a long time to coalesce into a coherent institution. Focus on larger firms: if hubs and technology extension centers are required to fund themselves to a significant extent, they may rely on providing services mostly to larger firms because they will be the ones that will be able and willing to pay the largest fees (either ex-ante or ex-post in the form of royalties) (Shapira and Youtie 2016). This defeats the purpose of targeting mostly MSMEs, for which significant subsidies are likely to be required. Uneven performance across centers: because centers differ in terms of the types of firms they work with and the technologies they work with, the quality of their services may also vary. This issue is ameliorated when the more advanced technology in question is more general-purpose, as in the Cloud Computing program in Spain. Reluctance to pay: centers that have models that rely on MSMEs paying upfront for services may struggle financially. 81 3 G. Elements for good design   Range of expertise and geography: in deciding how complex these centers should be, program managers should consider the extent of their geographical coverage (which implies deciding  Programs and instruments between having fewer and bigger hubs vs. having more hubs that are more spread out), as well as how generalist they will be. Funding scheme: Program managers must decide whether the program will be fully funded by sources other than participating firms or whether the latter would contribute. For example, the Kohsetsushi centers (Japan) are mostly publicly funded, whereas participating firms heavily finance the Manufacturing Extension Partnerships (USA) (Shapira, Youtie and Kay 2011). Given the increased cost and complexity of such programs, and given that projects that represent more advanced progress in firms’ digitalization journeys could more easily be tied to a specific revenue-generating prototype, this is a context where charging firms royalties from revenues generated by successful projects may be more feasible. Building a strong identity: in order to work well, these hubs and advisory centers must become top of mind for companies looking for help with their digitalization processes. Whenever a manager thinks about how her company can advance its technological capabilities, the hub’s name should come to mind as a potential ally. Thus, program officers need to invest in building a solid reputation and name recognition. The best-known example of an e-technology extension service that has a robust identity is Fraunhofer (Germany) (Comin, Trumbull and Yang 2016). DO DON’T • Ensure mechanisms to assess ser- • Although decentralization is the vice provider quality and consider right approach to ensure regional how to help reduce MSMEs transac- coverage, don’t rely on ad hoc, local tion costs in finding good quality pro- initiatives; a systematic, coordinated viders, for example, through accredi- approach is still needed to ensure tation and rating schemes. consistent quality, effectiveness, and cost efficiency. • Consider how funding schemes can simultaneously be financially sustain- able and inclusive of firms with fewer resources. • Work with partner organizations in the private sector to crowd in demand. CHALLENGE 3 Supplier market failures A necessary condition for the successful digitalization of MSMEs is adequate matching with suppliers from the ICT sector. However, proper supplier engagement may be hampered by supplier market failures of two types. Informational issues are not limited to firms’ lack of information on digitalization technologies. They can also arise when the firm seeks to find a supplier to provide digitalization solutions. For example, firms may simply not be aware of the 82 3 existence of suppliers that provide the technological services required, especially if suppliers   are not present in their region. Additionally, even if they are aware of their existence, firms may believe that it would be too time-consuming to find them. Furthermore, they may think that it is difficult to judge the quality of outside providers (Anderson and McKenzie 2021). And finally,  Programs and instruments they may not be confident that they can trust an outside firm (Williamson 1985). In addition, although MSMEs may be aware of the existence of suppliers that could be partners in their digitalization efforts, the market for suppliers may not be dense or dynamic enough to meet the needs of these firms. In this case, the number of suppliers is small enough that the market clearing prices for their goods and services are too high, pricing out many MSMEs. This can happen due to the ICT sector being focused on large projects, with suppliers’ perception that engagement with MSMEs is unappealing from a cost-benefit perspective, leading to lim- ited offerings for MSMEs. Alternatively, some markets may be underserved from a technology access perspective — although this is unlikely to be the case for more basic technologies, it can be a problem for more advanced technologies or for customized products. More broadly, this challenge can be linked to two well-known market failures: information asym- metry and missing markets. Options of policy interventions to tackle this issue include creating a space for information gathering and interaction between MSMEs and suppliers, providing certifications to suppliers to attest to their capabilities, and providing incentives for creating new suppliers. Given that the digitalization needs of MSMEs tend to be basic, it is likely that most countries and regions in the project will have enough suppliers to fulfill those needs — it is when it comes to more advanced stages of digitalization (cloud, Industry 4.0) that a shallow pool of suppliers may become a binding constraint. A. Target groups Smaller firms: given their lower experience in dealing with suppliers and in business deals in general, smaller firms may be more reluctant to engage with digital suppliers. Furthermore, being smaller often means that firms have fewer opportunities and resources to familiarize themselves with the digital supplier landscape. Firms not located in clusters: firms that are more isolated and less in touch with suppliers and other firms engaging in digitalization efforts will possibly be less aware and less trusting of suppliers. B. Potential sequence of instruments 5 10 18 Small and temporary Marketplaces Supplier development subsidies programs For other challenges, the sequence of interventions moves from programs for lower capa- bility firms to higher capability ones; for the challenge of thin supplier markets, however, the sequence moves from less interventionist to more interventionist programs. The first interven- tion is a hands-off subsidy program for emerging ICT suppliers to help strengthen the size and quality of the supplier market. If needed, policymakers can eventually become more involved by investing time and resources in more directly attempting to develop an ecosystem, first by building a marketplace, then by developing, alongside large firms, supplier development programs that require more extensive selection and training procedures. 83 3 C. Examples of interventions   B2B matching platform: One of the instruments of the Digitaliza tu Pyme program in Chile is Conecta Turismo, which is a business-to-business (B2B) matching platform connecting service  Programs and instruments SMEs from the touristic sector with technological service suppliers to digitalize processes. Through a simple process, after registering on the website, firms indicate for what business function they require the assistance of a supplier in providing digital solutions (web design, marketing, payments, etc.). Then they receive a list of potential suppliers, which can be filtered by region, that can be contacted directly through the platform. Government-driven e-portal: The A*STAR Collaborative Commerce Marketplace (ACCM) program in Singapore is a government-driven e-portal where businesses collaborate to find business and technology synergies. It is a complimentary platform for community building and development, allowing enterprises to network, form relationships, and foster collaborations. Marketplace listings should preferably be technologically driven (e.g., software solutions, robotic systems, warehouse automation, Manufacturing Execution System (MES), enterprise resource planning, cloud hosting, etc.) and B2B in nature. Solutions that have a high level of customizability, scalability, and technological novelty are highly preferred. They must also be commercially ready solutions and not be in the proof-of-concept phase. D. Evidence of impact Although there is significant research on the role of ratings in providing information about quality to buyers (Tadelis 2016), there is limited work evaluating the impact of informational interventions on matching suppliers and buyers in developmental contexts. One recent study examined the hy- pothesis that small firms in developing countries do not use the market for professional business services because firms lack information about the availability of these services and struggle to distinguish the quality of good versus bad providers. After setting up a business service market- place that provided such information, a randomized experiment found that neither the provision of information nor quality ratings had any significant impact on the likelihood that small firms would go on to hire a business service provider (Anderson and McKenzie 2021). If this result translates from business professional service suppliers to IT services suppliers, it suggests that informational constraints may not be binding. As suggested by the authors, an alternative constraint that may be operating is that MSMEs only learn and appreciate the value of professional business services by trying out these services for themselves, even if they know of their existence. E. Strengths Relatively easy and cheap to implement: programs such as the ones described above are not complex. In their simpler forms, they simply need a website with detailed and updated infor- mation on suppliers. (However, including more complex information, such as supplier ratings, can significantly increase costs.) A certification intervention can be outsourced to a third party at a relatively low cost. Operation through market mechanisms: information asymmetry and lack of trust can be re- duced through light-handed interventions that do not distort market signals. Spillovers to the entrepreneurship ecosystem: interventions that aim to increase the number of firms with the capabilities required to provide digitalization services would represent an 84 3 overall improvement to the digital human capital of managers and workers. This would have   positive productivity repercussions for present and future firms across different industries.  Programs and instruments F. Potential drawbacks and risks The risk of government failure: interventions that aim to increase the number of suppliers may lead to wasted resources and market distortions. Even a light intervention such as the creation of an e-portal to connect MSMEs and suppliers, although seemingly less demanding of the public sector, would require constant follow-up and updating to avoid becoming deserted and obsolete. Potential complexity of implementation: it may be hard to identify lack of knowledge or trust in suppliers as a binding constraint. Doing so properly may lead to significantly higher costs in the initial diagnostic phase of the program. G. Elements for good design Degree of intervention in depth of supplier market: program design needs to calibrate the degree of intervention in the market, weighing the importance of market failures against the possibilities of government failure. A diagnosis of a lack of capabilities in the current pool of suppliers may require more heavy-handed intervention to increase the number of providers. Degree of autonomy for firms to make connections: related to the above, how involved will the program be in directly connecting MSMEs to suppliers? An arms-length stance, where the program simply provides information and expects matching to flow naturally from that, is a simple and cost-effective approach but may not yield concrete results. An example of this approach is the Digital Marketplace program managed by the United Kingdom government. Engaging more directly in matchmaking efforts can overcome a lack of results but requires strong public sector capabilities (or a budget available to hire them in the market) and comes with the possibility of government failures. In the context of the supply chain disruptions caused by COVID-19, some governments engaged in more direct matchmaking efforts, as in the case of the Australian state of Queensland and its supplier matching program. Certification done by government or third party: in a certification program, the certifying insti- tution can be either a government body or a third party. The first option might be preferable if the state has the capabilities available to engage in this process, the opportunity cost of public sector employees working directly on this is not very high, and program designers have reason to believe that a more direct association with the government would make the certification more credible for MSMEs. One common form of government-driven certification is procurement certification for SMEs, such as the one run by the Small Business Administration, a United States government agency that runs the certify.sba.gov website. Such procurement-based certification programs can spill over into greater trust of suppliers for private firms. Another possibility is industry-driven certification, with the private sector taking responsibility for the certification process. The advantage of this option is the lower demand for services from the public sector, which would be particularly valuable for certification in the provision of more advanced tech- nological services, where the public sector is less likely to possess the capabilities needed to set up a certification program properly. However, industry-based certification processes may lead to upstream oligopolistic outcomes (Auriol and Schilizzi 2015). Features for certification: what requirements should suppliers need to fulfill to earn the certi- fication? This will be a function of the extent to which MSMEs are skeptical of suppliers’ capa- 85 3 bilities to provide the goods and services needed for digitalization properly. More skepticism   implies greater requirements for certification. Partnerships with industry associations: when engaging in efforts to improve information on  Programs and instruments suppliers to MSMEs and on increasing their levels of trust and interaction, program managers would be well-advised to work closely with suppliers’ industry associations. Such associations can often serve as the starting point for collecting and curating detailed information about providers, which can then be assembled and disseminated in a user-friendly way. For example, SME Digital Tech Hub in Singapore works with SME centers, trade associations, and chambers to help connect SMEs to ICT vendors and consultants. DO DON’T • Ensure that information and match- • Do not operate under the assumption making portals provide lists of that information asymmetry about previous clients of each supplier for suppliers is the main binding con- reference checks. Such information straint. It is important as a first step would ideally include public feedback to elicit from firms information on the from clients. extent to which they are aware of pro- viders and how much they trust them. • Attempt to combine an intervention that provides information on suppli- ers with an intervention that incentiv- izes MSMEs to use their services and provides credentialing of providers. Given the evidence discussed above that lack of knowledge on benefits may be a more binding constraint than information asymmetry, such a combination may yield positive results. A program that exemplifies this approach is the Business Link Pacific program, which serves MSMEs in a number of Pacific nations. 86 3.2 PREFERENCES AND BEHAVIORAL FACTORS 3   Technology adoption can also be impeded by behavioral factors caused by economic moti- vations or by biases. Behavioral factors that have an economic motivation behind them are  Programs and instruments usually associated with workers and their resistance to get fully onboard with a new technol- ogy. Alternatively, behavioral factors can simply be due to biases that require nudging and are usually associated with managers that have conservative risk profiles or are mired in behavioral inertia. Although prevalent, such behavioral challenges have been less explicitly targeted by programs. This makes for a promising avenue for researchers and policymakers to explore. CHALLENGE 4 Behavioral bias Economic decisions are often driven not by rational and optimizing strategies based on com- plete cost-benefit analysis but by behavioral heuristics. Although the economics literature has made extensive progress in applying these concepts to developmental contexts, work relating psychological aspects to technology adoption is more limited. Work on technology adoption has been most common in agriculture, where farmers fail to take advantage of profitable fertilizer investments due to a present bias that leads to procrastination (Duflo, Kremer, and Robinson 2011) and those with an internal locus of control (the belief that one’s own actions can affect future outcomes) more likely to adopt technologies such as fertilizers, improved seeds, and irrigation (Abay, Blalock and Berhane 2017). A. Target groups Older firms: these are prime candidates to be affected by these behavioral biases because firms are more likely to develop inertial decision-making processes over longer spans of time (Hannah and Freeman 1984). Family firms: decisions in companies that rely less on professional management are more subject to the preferences and psychological traits of particular individuals (Meroño-Cerdán, López-Nicolás, and Molina-Castillo 2018). Firms with positive performance: when firms are underperforming, or even worse, when they are facing survival risks, their managers are more likely to take risks (March and Shapira 1987). Consequently, programs seeking to elicit more active behavior from firms should focus on firms with relatively positive performance. B. Potential sequence of instruments 5 4 16 Small and temporary Showcasing examples of Technology demonstration subsidies successful transformation on-site Interventions geared towards ameliorating change-holding behavior can be of two natures. The first is the use of small and temporary subsidies to induce adoption in firms exhibiting time-inconsistent preferences (see the evidence of impact section). The second is the use of demonstration interventions that, by virtue of showcasing successful examples, can pro- pel firms to action. Additionally, programs that collect, process, and share high-frequency 87 3 information on business performance could induce behavioral changes in those interested   in adopting (Cusolito 2021).  Programs and instruments C. Examples of interventions Partnerships for prototyping: the Digitalization Boost program in Denmark is a funding program that incentivizes new industry 4.0 technologies (with commercial purposes) through collabora- tion within firms and between companies and knowledge institutions. With these resources, at least two firms and a knowledge institution can develop later stages of a functional prototype that works in realistic environments in a laboratory or is tested onsite. Provision of services and equipment for experimentation: the Tech Access program in Singapore provides advanced manufacturing and biotech/biomedical scientific services and equipment under the guidance of technical experts in these areas. The beneficiaries can access proto- types of new products, test new applications, and develop solutions using new technologies. Tech Access provides a cost-effective way for MSMEs to make informed equipment investment decisions, thereby mitigating upfront risk and capital expenditure. D. Evidence of impact The literature evaluating the role of risk and behavioral biases as a constraint on adopting new technologies is thin. The reason for this is that the critical question is whether, all things being equal, increases in the ex ante risk or behavioral biases affect technology adoption. Given that such behavioral biases tend not to change over time or vary across people in the same area in some well-defined way, an experiment to test this question remains mostly elusive (Foster and Rosenzweig 2010). Still, there is evidence that, in the context of agriculture in developing countries, offering small, early subsidies (right after harvest) for fertilization adoption induces a much larger adoption response than large subsidies offered later in the season, which is consistent with the present bias. In relation to policy design, this result is encouraging because it indicates that, given a proper diagnosis of change-holding behavior as being a constraint on adoption, the appropriate intervention is much less costly, and entails much lower distortions, than traditional heavy subsidies (Duflo, Kremer, and Robinson 2011). Furthermore, insurance schemes against technological and non-technological (Liu 2013) risks and education (Knight, Weir, and Woldehanna, 2003) encourage producers who are interested in adopting and do not face financial constraints to adopt early. Conversely, they encourage those who face financial constraints to apply for external sources of funding (e.g., commercial loans) to cover the fixed cost of adopting and speed up the process (Gine and Yang 2009). E. Strengths Signaling effects: as firms engage with programs in the creation of prototypes and pilot pro- grams for the adoption of IT technologies, and their peers observe their efforts and steady progress, the latter may become incentivized to adopt new technologies. Demand orientation: pilot programs and prototypes can be geared toward firms’ more im- mediate market needs, which serves as an incentive for more risk-taking behavior related to technology adoption. 88 3 F. Potential drawback and risks   Risk of one-off transactions: the one-off nature of most projects supported by subsidy schemes means that there is a risk that collaboration will not be sustained and, therefore, not lead to  Programs and instruments longer-term behavioral change to more innovative behavior and practices. Difficulty of reaching the intended target group: for a program that intends to target behavioral biases as its main challenge, it is important to be able to have a differential diagnosis of wheth- er investments in IT technologies are held back due to risk aversion or due to other reasons (capabilities, lack of information, etc.). Thus, identifying firms with behavioral biases as binding constraints may prove difficult. G. Elements for good design: Control of funds: when designing the program, policymakers should decide whether the tar- geted firms should receive funds to develop the project directly or whether a third party should control the funding flow according to milestones reached. Assessment of success: an important decision for program designers is what criteria should be included in the program’s ex-post evaluation. For example, should commercialization or full implementation, be a priority? Or is just getting the firm to take the first steps towards adoption the metric of success? As an example, the Tech Access program in Singapore focuses mostly on getting firms to experiment with new technologies, seeing that as an end in itself. Projects should be based on a gradual build-up of technological complexity: because conser- vative views toward investment in technology hamper adoption, the program must take a slow and gradual approach, making sure that firm management is on board with each step taken in the technological ladder. At the margin, projects should focus on more advanced technologies: although the build-up should be gradual, the initial technology chosen should be more advanced than not at the margin, given the assumption that advanced technologies are the ones that tend to lead to more conservative adoption behavior. Prioritizing noninnovative first-time applicants: firms with previous experience in technology adoption are less likely to have behavioral issues as a binding constraint for further techno- logical upgrading. Hence, the priority for programs aiming to tackle behavioral biases should be on first-time adopters. DO DON’T • Tackle behavioral limitations on tech- • If a behavioral constraint has been nology adoption through a variety of identified, do not overemphasize per- interventions: informational, demon- formance outcomes to the detriment strational, prototyping, partnerships, of adoption outcomes when evaluating etc. Explore all possibilities (even if interventions. In the context of behav- through pilots), including their combi- ioral holdups, taking the initial step out nations. of inertial states should be the priority. 89 CHALLENGE 5 3 Worker resistance to adoption   and incentive misalignments A challenge to adopting digital technologies (or most other technologies for that matter) that  Programs and instruments has gone largely unaddressed in the academic literature on economics and policy is workforce resistance to change.24 This is not surprising, given that such resistances are often hard to observe and measure empirically, especially considering that their incidence will vary greatly across different industries, technologies, and production processes. And while more empiri- cal work would be needed to assess how relevant an obstacle this challenge is, both theory (Parente and Prescott 1994; Desmet and Parente 2014) and the business practitioner literature suggest that it could be significant (Richter and Sinha 2020; Ratliff 2021). From an economic standpoint, worker resistance can be associated to loss of rents due to adoption, because of either job loss or lower bargaining power leading to lower salaries. The history of resistance to labor-saving technologies by shop-floor workers indicates that where digital technologies have labor-saving properties, they will also be met with resistance (Mokyr 1990, Lazonick 1990). As digital technologies sharply bring down the cost of communication within the organization, middle managers who primarily served as communication channels between lower-level workers and firm leadership may become redundant (Brynjolfsson and Hitt 2000). Less drastic, but still relevant, is the scenario where workers resist adoption not due to potential redundancy but because it generates rents that accrue mostly to the firm owners instead of to the workers (Aghion et al. 2020). Such disagreements on within-firms distribution of rents in the context of technology adoption can occur due to incentive misalignments. For example, in the context of soccer ball produc- tion in Pakistan, workers were not willing to adopt new production technology due to such a misalignment: their total compensation (which was piece rate-based) would be reduced with the slowing down of their work pace when starting to use the new technology. This resistance to adoption led them to misinform owners about the true efficiency gains of the new technol- ogy (Atkin et al. 2017). When properly identified, incentive misalignment is a source of worker resistance to adoption that can be eliminated through clever policy design. The information systems literature has extensively documented the phenomenon of worker resistance to IT adoption. It can take the form of apathy, sabotage and destructive behavior, and formation of coalitions. The object of resistance can include: the system itself and its features; the significance the new system has to the user, such as loss of power or status; or the implementers themselves (i.e., firm management), as the adoption process can become politicized to the point where the system becomes a pawn in a power struggle between the users and the implementers. Workers’ perceived threats include negative user assessments of the fairness of the exchange between their inputs and the outcomes of their interaction with the new system, and the disutility associated with switching to a new system (switching costs) (Rivard and Lapointe 2012). There are no programs (at least that we are aware of) that explicitly aim at tackling this chal- lenge and that could be used to illustrate it. An initial step would be an explicit acknowledge- ment of the issue in technology adoption advisory programs, such as the ones covered above. 24. Atkin et al. (2017) is one important exception. On how technology adoption can reduce incentive misalignment, see Kelley, Lane, and Schonholzer (2021). 90 3 A. Target groups   Hierarchic, non-flat firms: such firms have an organizational structure that incorporates a greater number of middle managers. Thus, they are natural candidates for internal conflict when the  Programs and instruments increased ease of internal communication derived from digital technologies may make these managers redundant (Brynjolfsson and Hitt 2000). High incidence of low-skilled workers: at the shop-floor level, firms with more workers with low education levels will be more inclined to simply replace them with newly adopted digital technologies (Acemoglu and Restrepo 2019). Although some labor-saving restructuring is probably inevitable for such firms, banking on sweeping dismissals associated with adoption is likely a recipe for worker resistance during the adoption process. Firms that rely heavily on informal contracts: many SMEs do not rely on formalized incentive schemes when managing their workforce. Instead, production is often organized around rela- tional contracts, where relationships between managers and workers are based on informality and trust. The introduction of information technologies can significantly increase the level of monitoring of workers on the part of managers, which can be interpreted as a breach of trust within long-standing relationships and become a source of internal conflict (Verhoogen 2021). B. Potential sequence of instruments 13 12 Managerial development Lump-sum payments programs for workers The lack of academic and applied research, as well as previous policy experiments, make this a hard challenge to address. Nevertheless, one possibility is a standard business advisory service containing a specific instrument that attempts to diagnose and ameliorate issues of incentive misalignment and overall worker resistance for digital technology adoption. This program could be associated with another one that distributes lump-sum payments for workers that demonstrate positive engagement with the technology (Atkin et al. 2017). C. Examples of interventions Business advisory: the ACTIVA Industry 4.0 program in Spain is a specialized and personalized advisory program carried out by accredited consultants with expertise in implementing In- dustry 4.0 projects. The program is carried out with a methodology developed by the General Secretariat for Industry and MSMEs. This program allows companies to have an assessment and a transformation plan that iden- tifies the digital enablers needed and establishes the roadmap for their implementation. The consulting is complemented by workshops demonstrating enabling technologies. D. Evidence of impact Given the incipient stage of the empirical literature on worker resistance and incentive mis- alignments related to technology adoption, there is a scarcity of research that evaluates the impact of interventions geared toward solving the issue. One recent randomized trial that 91 3 investigated how incentive misalignment was preventing adoption found that a lump-sum   payment that was tied to demonstrating successful adoption of the technology to owners and compensated workers for the losses associated with adoption was shown to be effective (Atkin et al. 2017). Thus, a business advisory service must be able to (1) properly diagnose  Programs and instruments such incentive misalignments and (2) propose changes in employee compensation that are compatible with incentives for both owners/managers and employees. In programs that seek to educate managers on how to react, the literature indicates that best results are obtained when managers adopt a strategy of credible dissuasion of complaints (coercion, persuasion), or a strategy of implementation rectification (redesigning the system, training, changing work schedule, etc.) that is congruent with issues raised by workers. Conversely, when managers simply acknowledge dissatisfaction, when they adopt non-congruent rectification, or when dissuasion is non-credible, resistance issues may get amplified (Rivard and Lapoint 2012). E. Strengths Conservative initial approach towards addressing the challenge: because there is no clarity on which instrument would be more appropriate to tackle the issue of worker resistance, using a traditional business advisory service to try to get at it is a prudent first step towards eventually designing a more tailor-made intervention. F. Potential drawbacks and risks Difficulty in diagnosing the challenge as relevant: given the lack of a track record of academic and applied research on the issue of worker resistance for (digital) technological adoption, it is not obvious how program managers could design a way to effectively diagnose this issue as being one of the main binding constraints for digitalization. G. Elements for good design Initial implementation of technology on a small scale: although the lack of programs that directly address the issue of incentive misalignment means that there is no record of best practices, some first principles can be roughly outlined. For instance, to ameliorate employee resistance at both managerial and shop-floor levels, it is important to implement the new technologies gradually and on a small scale. This gives opportunities for workers to grow used to the new processes and, most importantly, voice their concerns and how they believe they can work as complements to the new technologies. Diagnosis by third party: given that incentive misalignment is a sensitive topic between owners, managers, and workers, it is important that the diagnosis be conducted through a neutral third party who is engaged in building a trust-based dialogue between the parties. Identification of rents by third party: similarly, a proper discussion of the additional rents cre- ated for the firm by adopting digital technologies and how they are distributed across different groups needs to be brokered by a neutral, trusted third party. 92 3   DO DON’T • Prioritize getting the diagnostic right, • Don’t adopt technologies at a large  Programs and instruments making sure that worker resistance scale within firms without making sure is occurring and preventing adoption. that they won’t be disruptive to the This will require a good understanding employer-employee relationship. of how the technology interacts with the employer-employee relationship. • Rely on organizations with experience mediating dialogue between manage- ment and workers to act as a neutral third party. 93 3.3 ORGANIZATIONAL MISALIGNMENTS 3   A final set of challenges are related to the lack of capabilities that may hold firms back in their digitalization trajectories. This lack of capabilities can occur at the worker level: MSMEs from  Programs and instruments developing areas usually have to tap into a pool of workers that lack general human capital, and in particular lack the digital human capital necessary for successful adoption. Alternatively, the lack of capabilities may occur at the managerial level. Managers may lack the knowledge of how to better organize production so as to facilitate adoption of technologies and optimize their use. For both challenges, policy has a relevant role in raising the level of knowledge on technical skills and managerial practices, for both workers and managers. CHALLENGE 6 Low digital human capital of workers A. Target group Firms with an older workforce: older employees are, on average, less comfortable with digital technologies, while younger ones often can improve their digital human capital more easily (Behaghel, Caroli and Roger 2014). Thus, even though a small intervention may incentivize younger workers to develop their digital skills, more transformative effects may be achieved by a more deliberate policy of aiming at workers who need a bigger push. Firms with lower general human capital of workers: like the above, the biggest impact of a pro- gram aiming to increase the digital human capital of workers would be achieved by aiming at firms where workers are less productive by virtue of their lower general human capital — which prevents them from improving their digital human capital autonomously. However, levels of general human capital must not be too low; otherwise, the firm will not possess the necessary absorptive capacity to reap the program’s benefits. B. Potential sequence of instruments 1 6 11 Distance technology Vocational training Group technology individual learning for employees learning This sequence of interventions begins with programs at the individual level, which can meet workers at each of their respective levels of technological knowledge. A subsequent step would be to engage in group-level technological learning — because most workers in the firm will have developed a common ground in terms of their digital skills, they can now move ahead in their improving skills as a group. C. Examples of interventions Investment fund for SMEs for training: Digital Now is one of the components of the Mittelstand 4.0 program in Germany. This component is an investment fund for SMEs to exploit the econom- ic potential of digitalization. This instrument offers grants to encourage companies to invest in digital technologies and their employees’ qualifications. The Digital Now funding program runs until the end of 2023. One of its modules is “Investing in employee qualification.” In terms 94 3 of the quality level of the training, providers must be proven by certification according to ISO   9001 or accreditation, according to AZAV. Training centers for digital skills: the Digital Skills Partnership (DSP) program in the United King-  Programs and instruments dom is a multi-stakeholder and cross-sectoral partnership addressing all levels of digital skills through the establishment of local centers across the United Kingdom. They bring together partners from the public, private, and third sectors to upskill the current workforce, advance digital inclusion, and raise awareness of the importance of digital skills regionally. The activities supported by local DSPs have varied, reflecting the local context, digital skills priorities, and funding availability. Common activities include raising awareness of digital oppor- tunities; supporting individuals to enhance their digital skills; increasing businesses’ appre- ciation of the importance of digital skills; and supporting digital inclusion. One of the training projects was pursued by the Southwest DSP, which secured external funding to enable 1000 individuals to access 12-week digital or technical boot camps co-designed with employers that map to known skills shortage vacancies in well-paid roles. D. Evidence of impact Programs that seek to improve the digital human capital of workers will have to rely heavily on training. The labor economics literature on the causal effects of training programs on workers’ labor market outcomes can serve as an informative parameter on how successful these pro- grams can be. While the literature is extensive and finds varying effects (Ashenfelter and Card 1985; LaLonde 1986; Heckman, Ichimura and Todd 1997), a recent meta-analysis (Card, Kluve and Weber 2017) finds that training programs tend to have small average effects in the short run. This is consistent with the fact that participants often suspend their normal efforts while engaged in training, which leads to worse outcomes in the immediate post-program period. However, effects become significantly larger in the medium and long run. E. Strengths Achieving optimal social levels of training: it has been well-documented that without public policy interventions, firms undersupply employee training compared to optimal social levels because they cannot internalize the benefits of training given employee mobility to competitors (Brunello and de Paola 2004). Programs that provide digital training for employees can overcome this issue. Scale: much of the digital human capital that firms need their employees to acquire is gener- al-purpose, cutting across different sectors. Consequently, a training program for employees focusing on digital skills can be designed at a large enough scale to justify its fixed costs. F. Potential drawback and risks Poor match between supply and demand for services: Program designers must ensure that the training services provided are the ones that firms need. Problems can arise when a standardized set of training services is offered regardless of firms’ needs, which may lead to low take-up Lack of coordination among service providers: Several agencies and actors are typically in- volved in providing training. Therefore, the government must ensure a coordinated, systemic approach to providing training services integrated with the availability of other instruments. 95 3 G. Elements for good design   Selection mechanism: firms can be selected through competition, to identify firms most likely to meet program goals, or automatically through entitlement criteria. Selection can be done  Programs and instruments via a continuous application process, whereby proposals can be received at any time, or via periodic calls for applications with fixed deadlines. Competition-based programs tend to have more positive firm outcomes (Colombo, Grilli and Murinu 2011; Crespi and Maffioli 2014). Funding versus direct provision: when training services are subsidized, the government usually provides the SMEs a voucher or grant to purchase the services from a third party. For example, the Walloon region of Belgium sells for 15 EUR vouchers for employee training that are worth 30 EUR, with companies being allowed to purchase up to 800 vouchers and use them for courses in accredited training institutions. When the government provides the service directly, they generally have centers with advisers at the regional and local levels that deliver the services to SMEs. In this direct provision model, it is critical that the training centers have knowledge and credibility and can quickly add value to SMEs. This model is followed by the Mittelstand centers in Germany. Depth versus coverage: program designers will have to decide whether to prioritize the provi- sion of training to the largest possible number of firms or to provide services to fewer firms but engage with them over time to follow up on their digitalization efforts and constantly monitor their training needs. The latter configuration would require a more direct selection of partici- pating firms, which would require objective criteria for selection (such as export potential) to avoid program capture. Narrower coverage based on competitive criteria may incur the risk of “picking winners” — although selected firms may be more capable of using the resources to suc- cessfully train employees, they possibly could have done so regardless of the subsidy (meaning there is no additionality) (Cantner and Kosters 2012; Ruegg and Feller 2003; Cirera et al. 2020). Improving information on training services: given the profusion of public and private training providers and firms consequently not having clarity on which ones to rely on, it may be op- portune to provide a vetted list of training providers that are known to offer quality services. Partnering with private sector intermediaries: it is important to rely on the help of actors, such as business associations and large customer companies, that regularly interact with SMEs to be able to reach and steer them toward the training programs (Schnedeider 2015). For exam- ple, the Mittelstand 4.0 Competence Centers leverages partnerships with multipliers, such as associations and business chambers, to make its training offers widely known. DO DON’T • Focus on additionality is important. • When it comes to working with a Given their growth needs, fast-grow- training provider, firms need guidance ing firms will tend to provide employee on how to choose a partner. Don’t training regardless of market failures. trust SMEs to make this decision on Thus, programs should target firms their own. that have more binding constraints. • One possible mechanism is requiring matching funds with lower rates for firms in greater need. 96 CHALLENGE 7 3 Inadequate   managerial practices In the last 20 years, a substantial literature on the overlap of economics, management, and infor-  Programs and instruments mation systems has described how firms struggle with the adoption of digital technologies due to a lack of investments in the adoption of complementary practices. (Relevant works include Bresnahan, Brynjolfsson and Hitt 2002; Tambe, Hitt, and Brynjolfsson 2012; Crespi, Criscuolo, and Haskel 2007; and Bloom, Sadun, and Van Reenen 2012.) Many of the more traditional man- agerial practices that were in place before the IT revolution reflected the high cost of informa- tion processing at the time. However, as computers became cheaper and more powerful, their adoption meant that such practices became obsolete — in turn, achieving the full potential of these technologies required the invention and adoption of new processes, procedures, and or- ganizational structures (Brynjolfsson and Hitt 2000). These new practices may take many forms depending on the company and industry context. However, they often include a more diversified and customized product portfolio, decentralization, more worker autonomy, and frequent inter- actions with customers and suppliers (Ichniowski and Shaw 2003). More broadly, this struggle to adopt adequate management practices in the context of digitalization efforts can be seen as an instance of the shortage of managerial capabilities that often afflicts MSMEs (Bloom et al. 2013). A. Target firms Older firms: older firms will have more inertia in their work practices and thus will have a harder time implementing necessary complementary practices when adopting digital technologies (Hannah and Freeman 1984). Family firms: family firms, especially ‘dynastic’ family firms that combine family ownership and control, are less likely to adopt structured management practices (Tsoutsoura 2021). B. Potential sequence of instruments 6 13 12 Vocational training Managerial development Lump-sum payments for employees programs for workers The proposed sequence of interventions considers the need for both workers and managers to become more acquainted with the organizational practices required to optimize digital tech- nology adoption. Using groups (Iacovone, Maloney and McKenzie 2022), either horizontally or vertically (Dalton et al. 2021), can help move (Cai and Szeidl 2018; Fafchamps and Quinn 2018) beyond business training or consulting-directed managers to rely more on insourcing workers with functional expertise, outsourcing tasks to professional specialists, or outsourcing tasks to professional specialist (Anderson and McKenzie 2021) plants. Finally, a program of lump-sum payments for workers to incentive practice adoption is also suggested, following the logic of programs designed for technology adoption in the context of incentive misalignments on the part of workers (Atkin et al. 2017). C. Examples of interventions Advisory program for complementary digital capabilities: The I’m Star program in Korea is a comprehensive program for small firms to enter and develop content for online sales. The 97 3 program consists of more than 13 instruments tailored according to the firm’s needs: education,   training, vouchers, and public goods to enter new markets online, develop digital content to promote sales, develop capabilities for increasing the firm’s presence on e-commerce, etc.  Programs and instruments Relevant instruments related to complementary practices include: • Commercialization consulting support: this line offers support for the commercialization of products to make them suitable for online sales. • Support for promotion and advertisement of products: this line offers support for the prod- uct’s digital design (images, videos, etc.) to increase online sales. Model plant: one of the components of the Smart Factory Korea program aims to enhance the performance of smart factory building companies, such as improving the productivity of smart factory building through corporate diagnosis and advice from specialized consulting organizations. The mechanism is providing funding for on-site diagnosis of smart factory building companies and solving problems in the field during the construction process through consulting methodology and tools of specialized consulting organizations. D. Evidence of impact Evidence on the effectiveness of management advisory has been accumulating over the years. An intervention with Indian textile firms found that free consulting on management practices increased productivity by 17% in the first year through improved quality and efficiency and re- duced inventory (Bloom et al. 2013). Similarly, an intervention in Mexico found positive effects of one year of management consulting on total factor productivity, entrepreneurial spirits, and return on assets. The most prominent improved management practices were marketing, financial accounting, and long-term business planning (Bruhn, Karlan, and Schoar 2018). Sub- sequent work has found that such interventions may be even more effective if they are done for groups of firms to allow for peer learning (Iacovone, Maloney and McKenzie 2022) and that both insourcing and outsourcing of management tasks can be more effective than training and consulting (Anderson and McKenzie 2021). Adoption of digital technology could itself induce managerial changes — e.g. changes in workflow resulting from digitalization required modern- izing human resources (HR) practices (Roberts and Shaw 2022). E. Strengths Benefits that go beyond digitalization: although the introduction of complementary capabil- ities may at first serve the main objective of optimizing the benefits of the adoption of digital technologies, these capabilities can often prove useful in the medium and long term in terms of improving firm competitiveness (Bloom et al. 2013; Cirera and Maloney 2017). Furthermore, exposure to such processes of capability acquisition may lead firms to be able to independent- ly diagnose and implement new capabilities as their competitive environment evolves — thus developing what the management literature terms dynamic capabilities. Spillovers: relatedly, as employees and managers become versed in new practices, and ac- quire greater flexibility and exposure to organizational change, this can have positive impacts for firms outside the initial scope of the program through labor market mobility and informal interactions (Arrow 1962; Saxenian 1994). 98 3 F. Potential drawback and risks   High cost: a proper diagnostic and subsequent implementation of the complementary capabili- ties that the firm needs to develop may require extensive involvement from outside consultants,  Programs and instruments who would need to evaluate tasks across many functions within the firm. This could lead to rising program costs, at least compared to programs focusing on informational interventions or specific business functions. Adequate supply of consultants: a necessary condition for the success of programs that look to increase managerial capabilities in MSMEs (either through advisory services, insourcing of specialized workforce or outsourcing of tasks) is the availability of skilled consultants. Howev- er, MSMEs may not have access to a deep pool, especially if they are located in less dynamic areas. Resorting to consultants from different regions may lead to significant cost increases. G. Elements for good design Diagnostic versus diagnostic + intervention: a simpler version of a program might simply provide firms with a diagnostic of the complementary capabilities required and leave firms with the option of implementing them by themselves. A more comprehensive version might go beyond that to include instruments that implement the required capabilities. For example, the Industry 4WRD Readiness Assessment program in Malaysia assists firms in assessing their capabilities and readiness to adopt Industry 4.0 technologies and processes by identifying the gaps and areas of improvement for Industry 4.0 adoption as well as opportunities for productivity im- provement and growth. Only in a separate module do the firms have the option of applying for grants to attempt to implement the plans from the diagnostic phase. Focus on one versus many functions: another decision related to the program’s complexity is how many business functions to include in the design. The simpler alternative is to focus solely on core business functions. For example, the Transfer Office for IT Security component of the Mittelstand 4.0 program in Germany establishes an IT security transfer office for medi- um-sized companies, providing tailor-made measures to improve the beneficiary’s IT security. Securing employee buy-in: adopting these complementary practices may involve significant changes to how things get done in the firm. This attempt to change long-established practices may lead to significant employee backlash because employees often resist giving up routines they have perfected over the years (Lazonick 1990). It is then important to give voice to work- ers (possibly within a continuous improvement or kaizen framework), so they feel that their input and their experiential knowledge of production processes are being taken into account (Kochan et al. 2019). Paying attention to hard-dying old habits: alternatively, while employees may be on board with enacting changes to production processes to make them more complementary to new digital technologies, a lot of the heuristics and tacit knowledge that go into employees’ day- to-day tasks may remain influenced by the original routines. Thus, the proper adoption of new complementary practices may be unintentionally hampered by these habits. Program design should make sure to probe possible vestiges of old practices. Finding the key complementary practice: the focal firm may face many potential complementary practices that it could adopt to enhance its digitalization process. Attempting to adopt many of them simultaneously may lead to subpar performance, wasted resources, and disgruntled employees. Ideally, the program should be able to identify 1 or 2 practices that present the 99 3 biggest “bang for the buck” in unlocking the potential of the newly adopted digital technologies.   The key complementary practice is often HR-related (Aral, Brynjolfsson and Wu 2012).  Programs and instruments DO DON’T • Map out the size and quality of the • Although interventions at the man- business consultant market available agement level may be the focus, do to MSMEs. A program that is ambi- not forget about acquiring workers’ tious to the point that it stresses buy-in to introduce changes in man- the available pool of consultants agement practices because they will can lead to MSMEs receiving subpar often need worker involvement and services. cooperation. 100 ANNEX 1 LIST OF ANALYZED DIGITAL DIAGNOSTIC TOOLS Business Country Digital Diagnostic Tool Category Firm size sector Developer(s) Australia Digital Readiness Light-touch / Micro / Small Horizontal Australian Department of Assessment Tool Stand-alone / Medium Industry, Science Energy and Resources Baltic InnoCAPE Digital Maturity Light-touch / Micro / Small Horizontal InnoCAPE Consortium States / Assessment Tool Integrated / Medium Scandinavia Belgium DigiScan Light-touch / Micro / Small Industry Agoria Integrated / Medium DigiScore Light-touch / Micro / Small Horizontal Agence du Numerique (AdN) Stand-alone / Medium Digital Maturity Diagnosis Intensive / All Horizontal Digital Wallonia Integrated Binder Djiker Otte Digital and Innovation (BDO) Digital Maturity Diagnosis - Light-touch / All Construction Digital Wallonia Construction Sector Integrated Binder Djiker Otte Digital and Innovation (BDO) Confederation Construction Wallone (CCW) Digital Maturity Diagnosis - Light-touch / Micro / Small Horizontal e-invoicing Integrated Digital Matyrity Diagnosis - Intensive / Micro / Small Horizontal Retail Sector Stand-alone Indice de Sante Digitale Intensive / All Horizontal Safeonweb (Digital Health Index) Stand-alone Quick Scan Industry 4.0 Light-touch / Small / Manufacturing Digital Wallonia Online Integrated Medium Canada Digital Maturity Assessment Light-touch / Micro / Small Horizontal Business Development Bank of Stand-alone / Medium Canada Digital Maturity Self- Light-touch / All Construction Klynveld Peat Marwick Assessment Tool Stand-alone Goerdeler (KPMG) Canadian Construction Association (CCA) Denmark Digital Maturity Assessment Light-touch / All Horizontal Aarhus University Tool Stand-alone France Auto-Diagnostic Numerique Light-touch / All Horizontal Uniformation Stand-alone Diag-Numerique Light-touch / Micro / Small Horizontal Mouvement des Entreprises de Stand-alone / Medium France (MEDEF) Diagnostic DIMM.UP Light-touch / Various Horizontal DIMM.UP Stand-alone 101 Business Annex 1  Country Digital Diagnostic Tool Category Firm size sector Developer(s) France Diagnostic de Maturite Light-touch / All Horizontal BlueMarketing (cont.) Digitale Stand-alone Digitalometre Intensive / Micro / Small Horizontal Bpifrance Stand-alone / Medium  List of analyzed Digital Diagnostic Tools Digital Maturity Assessment` Light-touch / All Horizontal Axway Integrated Digital Maturity Assessment Light-touch / Small / Manufacturing Braincube Stand-alone Medium / Large EvalNumPME Light-touch / Micro / Small Horizontal Cofederation de Petites est Stand-alone / Medium Moyennes Entreprises (CPME) CINOV Numerique Le Grand R Light-touch / Small / Manufacturing Visiativ Stand-alone Medium / Large Outil D’Autodiagnostic Light-touch / All Horizontal Solidataech Numerique Stand-alone Smart Diag’ Intensive / Micro / Small Horizontal Operateur de Competence des Stand-alone / Medium Entreprises de Proximite Test de Maturite Numerique Light-touch / Micro / Small Horizontal CNMA Grand Est Stand-alone / Medium Germany Digital Maturity Index Intensive / Small / Manufacturing acatech (National Academy of Integrated Medium / Science and Engineering) Large Digital Readiness Light-touch / Small / Horizontal Consultport Assessment Stand-alone Medium / Large Digital Supply Chain Maturity Intensive / Small / Horizontal SAP Assessment Stand-alone Medium / Large Digital Transformation Intensive / Small / Horizontal Frauhofer IPK Assessment Integrated Medium / Large Festo Industry 4.0 Quick Light-touch / Small / Manufacturing Festo Check Integrated Medium / Large I4.0 Readiness Assessment Intensive / Small / Manufacturing TUV SUD Integrated Medium / Large Industry 4.0: Online Self- Intensive / Micro / Small Manufacturing IMPULS Foundation of Check for Buiness Stand-alone / Medium the German Engineering Foundation IW Consult (a subsidiary of the Cologne Institute for Economic Research Institute for industrial Management at RWTH Aachen University Industry 4.Now Maturity Light-touch / Small / Manufacturing SAP Assessment Stand-alone Medium / Large 102 Business Annex 1  Country Digital Diagnostic Tool Category Firm size sector Developer(s) Ireland Digital Maturity Index Light-touch / All Horizontal Organization Development Tool Integrated Institute (ODTI) Japan Digital Maturity Assessment Light-touch / All Horizontal Konica Minolta Integrated  List of analyzed Digital Diagnostic Tools Singapore The Smart Industry Light-touch / Small / Manufacturing Singapore Economic Board Readiness Index (SIRI) Integrated Medium / (EBD) Large TUV SUD Spain Advanced Digital Self- Light-touch / Medium / Manufacturing Spanish Mistry of Industry, Diagnostic Tool (HADA) Stand-alone Large Commerce and Tourism Digital Maturity Assessment Light-touch / Micro / Small HoReCa The Economic Development Tool (DMAT) Integrated Department of Fomento San Sebastian Sweden Digital Maturity Assessment Light-touch / Micro / Small Horizontal Combient Integrated UK Digital Maturity Assessment Light-touch / Micro / Small Horizontal Aveva Stand-alone Digital Maturity Assessment Light-touch / Micro / Small Horizontal Bilfinger Stand-alone Digital Maturity Assessment Light-touch / Micro / Small Horizontal Ionology Tool Integrated Digital Maturity Model Intensive / Micro / Small TelCo Deloitte (DMM) Integrated / Medium TM Forum Digital Maturity Self- Intensive / Small / Healthcare National Health Services (NHS) Assessment Integrated Medium Digital Readiness Intensive / Micro / Horizontal Ernst & Young (EY) Assessment Stand-alone Medium Industry 4.0 /Digital Light-touch / Micro / Small Manufacturing PricewaterhouseCoopers Operations Self-Assessment Integrated / Medium (PwC) USA KPMG Connected Enterprise Intensive / All Horizontal KPMG Maturity Assessment Integrated Diagnostic Agile 360 Light-touch / All Horizontal McKinsey & Company Stand-alone Digital Acceleration Index Intensive / Medium / Horizontal Boston Consulting Group (BCG) (DAI) Integrated Large Digital Capabilities Intensive / Medium / Horizontal McKinsey & Company Integrated Large Digital Culture Index Intensive / Medium / Horizontal McKinsey & Company Integrated Large Digital Infrastructure Intensive / Medium / Horizontal McKinsey & Company Assessment Integrated Large Data Management Maturity Light-touch / Small / Horizontal CMMI Institute (DMM) Assessment Integrated Medium / Large Digital Data Maturity Light-touch / Micro / Small Media Google Assessment (News) Stand-alone Deloitte 103 Business Annex 1  Country Digital Diagnostic Tool Category Firm size sector Developer(s) USA(cont.) Digital Data Maturity Light-touch / Micro / Small Retail Google Assessment (Retailers) Integrated Deloitte Digital Marketing Maturity Light-touch / Micro / Small Horizontal Google Assessment Integrated / Medium Boston Consulting Group  List of analyzed Digital Diagnostic Tools Digital Maturity Assessment Intensive / Small / Wholesalers NuORDER Tool Integrated Medium OUTGROW Digital Marketing Light-touch / Medium / Horizontal McKinsey & Company Capabilities Assessment Stand-alone Large Digital Maturity Quick Index Light-touch / All Horizontal Emerson Electric Stand-alone Digital Maturity Quiz Light-touch / All Horizontal West Monroe Stand-alone Digital Quotient Light-touch / Small / Horizontal McKinsey & Company Stand-alone Medium / Large Digital Readiness Light-touch / Micro / Small Horizontal Aon Assessment Stand-alone / Medium Digital Readiness Light-touch / Micro / Small Horizontal GoGestalt Assessment Tool Stand-alone / Medium Digital Readiness Light-touch / All Manufacturing Oden Technologies Assessment Tool Stand-alone Digital Readiness Survey Light-touch / Medium / Horizontal Bain and Company Stand-alone Large Digital Technology Light-touch / Medium / Horizontal McKinsey & Company Assessment Stand-alone Large Digital Transformation Intensive / Small / Horizontal International Data Corporation Assessment Stand-alone Medium / (IDC) Large Digital Transformation in Light-touch / All Manufacturing Northwestern University Operations Stand-alone The Kubernetes Maturity Intensive / Medium / Horizontal Altoros Assessment Integrated Large The Cisco Digital Readiness Light-touch / All Horizontal Cisco Index Rapid Assessment Stand-alone Mercer Digital Readiness Light-touch / All Horizontal Mercer Assessment Stand-alone Source: Digitrans team 104 ANNEX 2 LIST OF ANALYZED PROGRAMS AND INSTRUMENTS Selection of instruments for the analysis was guided by the objective to obtain information about effective interventions implemented in most mature digital ecosystems globally. The scope of the review included digitalization leaders from Europe, North America, South Amer- ica, and Asia (Table A2.1). From each continent from 2 up to 5 firm-support ecosystems were analyzed TABLE A2.1 Countries included in analysis of to account for differentiated approach to address programs and instruments  digitalization challenges in various socio-economic North South contexts. Instruments were identified and selected Europe America America Asia to an in-depth review based on information from Germany  Canada  Chile  Malaysia desk research. Owners and managers of programs Denmark  United States  Uruguay  Korea  and instruments analyzed in-depth were contacted Portugal  New Zealand  to obtain additional information on effectiveness Spain  Singapore  and impact.  United Kingdom Over 35 programs encompassing almost 100 instru- ments were revised for the purpose of this report. Source: Digitrans team. Table A2.2 summarizes key information about the revised interventions, and the Excel database – an additional Annex to this report, specifies more details of the reviewed ones. The database would be useful to inform the design instru- ments for addressing complementary capabilities for digitalization of MSMEs. TABLE A2.2 List of instruments and programs Program Instrument Instrument Managing Firm Business Digital Country name name type Short description authority size sector sophistication Canada Digital Grow Your 5. Small and Grow Your Business Innovation, Micro / Horizon- Pre-Foun- Adoption Business temporary Online is exclusively for Science Small tal dational / Program Online subsidies businesses that haven’t and Foundational taken advantage of Economic eCommerce opportuni- Devel- ties yet. Support in this opment stream comes in the form Canada of microgrants of around $2,400 to help with digital adoption costs and will be available for up to 90,000 small businesses. The grants can be used for various activities, including website devel- opment, search engine optimization (SEO), social media advertising, and fees for eCommerce platforms. Firms also have access to a network of digital experts. 105 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Digital Boost Your 11. Manage- Grants that can be used Innovation, Micro / Horizon- Pre-Foun- Adoption Business rial devel- for: coverage of 90% of Science Small tal dational / Program Technology opment the cost of developing a and Foundational (cont.) program digital plan; Interest-free Economic loans to support the Devel- implementation of the opment  List of analyzed programs and instruments plan; wage subsidies Canada to hire Canadian youth to help with the digital/ online elements of the business on short-term placements. Participants will also have access to a network of advisors to help with the design and implementation of digital plans.  Chile Digitaliza Pymes de 12. Market- E-commerce platform for Ministry of Micro / Horizon- Pre-Foun- tu Pyme Barrio places micro, small and medium Economy Small / tal dational / retailers, which providessupported Medium Foundational marketing and B2C tools. by CORFO and SERO- Pymes en 15. Individ- Collaborative network of TEC Línea ual tech. advisors, workshops, and learning courses, providing basic digital skills and other advisory services. After receiving the courses and being certified, the entrepreneur can apply to business advisory services, some of them are free. Ruta Digital 1. Distance Courses for basic digital individual skills. The courses are tech. learn- offered after the volun- ing tary-self assessment online and tailored according to the results. Conecta 12. Market- B2B matching platform, Tourism Turismo places connecting service SMEs from the touristic sector with technological ser- vice suppliers to digitalize processes. Chile Digita- Digitaliza 15. Individ- Voucher to purchase digi- Sercotec Micro / Horizon- Pre-Foun- liza tu tu Almacén ual tech. tal equipment and access Small / tal dational / Almacén learning technical assistance, Medium Foundational training, and market advisory to digitalize the business process. This program is oriented to micro and small retailers. Denmark Digita- Digitaliser- 10. Group Support for collaboration Danish Micro / Horizon- Intermediate / lization ings tech. learn- within firms and between Business Small / tal High Boost Boost - ing companies and knowl- Agency Medium Classic edge institutions. (Erh- vervsh- usene) 106 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Digita- Digitaliser- 3. Assis- Funding for implemen- Danish Micro / Horizon- Intermediate / lization ings tance in tation of technologies Business Small / tal High Boost Boost - access to allowing to maintain Agency Medium (cont.) COVID-19 finance business activity during (Erh- the pandemic. vervsh- usene)  List of analyzed programs and instruments Denmark Sprint: Digital 15. Individ- Vouchers for business Danish Small / Horizon- Foundational / Digital Vouchers ual tech. advisory with technology Design Medium tal Intermediate learning extension services. Center Denmark SMV: Digital 4. Web- Vouchers for advice to Danish Micro / Horizon- Foundational Digital conversion based enhance SME’s digital Business Small tal - vouchers technology conversion and for devel- Agency demonstra- oping e-Commerce. (Erh- tion vervsh- usene) Learn:Dig- 1. Distance On-line course to support ital individual creation of e-Commerce (Laer:Dig- tech. learn- capabilities in firms. ital) ing Germany Mittel- Com- 14. Tech- Network of regional Federal Small / Manufac- Intermediate / stand 4.0 petence nological information points and Minis- Medium turing High Centers testbeds technological demon- try for (Kompe- strators for testing new Economic tenz-zen- solutions in practice Affairs and tern) (test-before-invest). Energy (BMWI) Transfer 9. Technol- Network of regional IT Medium Horizon- office for ogy demon- security transfer offices tal cyber-se- stration for medium-sized curiry off-site companies. It creates a nationwide supply, providing tailor-made measures to improve the beneficiary’s IT security. Digital Now 3. Assis- Grants and matching Micro / Foundational / (Digital tance in grants for financing tech- Small / Intermediate Jetzt) access to nology acquisition and Medium finance trainings. Germany Go-Dig- Digitized 15. Individ- Funding for consulting Federal Micro / Horizon- Foundational ital business ual tech. and implementation Minis- Small / tal processes learning services incl.: shipping try for Medium and returns management, Economic logistics, warehous- Affairs and ing, electronic payment Energy methods. (BMWI) Germany AI for AI for SMEs 17. Exper- This program provides Federal Small / Specific, Intermediate / SMEs (KI imental funding for R&D and Ministry of Medium priority High für KMU) devel- implementation of AI Educa- sectors opment technologies in SMEs, tion and programs incl. high-risk industrial Research research and pre-com- (BMBF) petitive development projects in the field of AI, which could not be carried out or only with significant delay without funding. 107 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Germany SME SME inno- 17. Exper- This program provides Federal Small / Horizon- High innova- vative: ICT imental funding for research on Ministry of Medium tal tive: ICT (KMU-in- devel- the field of ICTs for SMEs. Educa- (KMU-in- novativ: opment The projects funded tion and novativ: IKT) programs should outline the novelty Research IKT) about the state-of-the- (BMBF)  List of analyzed programs and instruments art in R&D in ICTs. Korea Smart Support 3. Assis- High value grants for tar- Korea Small / Manufac- High Factory for the tance in gets consortiums between Smart Medium turing Korea construc- access to producers and suppliers to Manu- tion of new finance develop new smart facil- facturing and the ities or upgrade existing Office upgrading ones. Up to 50% of public of existing co-financing of investment smart and operational costs. factories Support for 18. Supplier Support for collaboration Small / SMEs win- devel- between SMEs and large Medium win smart opment firms to joint production / Large factory program facilities by co-financing creation of new or expan- sion of existing smart factories between large firms and SMEs. Support for 3. Assis- Funding for developing Small / Specific, building tance in smart factories for groups Medium priority special- access to of SMEs within specific sectors ized smart finance economic activities with facilities high export potential. Support 15. Individ- Funding to introduce spe- Horizon- Intermediate / for Smart ual tech. cialized analysis solutions tal High Factories learning for SME manufacturing based on data such as AI and big DNA (Data, data and support for the Network, construction of automa- AI) tion equipment, control- lers, sensors, etc. Smart Fac- 3. Assis- Covering costs of certi- Manufac- tory Level tance in fication aimed to expand turing Certifica- access to the supply and expansion tion System finance of smart factories by confirming and verifying SMEs’ smartization. Estab- 14. Tech- Financing for establish- Medium Specific, High lishment nological ment and operation of a / Large priority of smart testbeds testbed that conducts sectors factory technology development testbed by and demonstration related region to smart factories in each of the four Korean regions. Reinforce- 15. Individ- Enhance smart factory Small / Horizon- Foundational / ment of ual tech. performance through Medium tal Intermediate smartiza- learning mentoring and consulting. / Large tion capa- bilities Cloud- 17. Exper- Creation of a cloud com- Associ- Intermediate / based imental puting solution dedicated ation of High solution devel- to needs of group of SMEs, SMEs develop- opment in collaboration with pro- ment programs vider of digital services and technologies. 108 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Korea Sup- Non-face- 4. Web- Vouchers to enable SMEs Ministry of Small / Horizon- Pre-Foun- port for to-face based to merge to online ser- SMEs and Medium tal dational / remote service technology vices, such as video con- Startups Foundational work voucher demonstra- ferencing and telecom- (MSS) support tion muting, personnel, labor, and security consulting  List of analyzed programs and instruments for the introduction of a non-face-to-face system. Korea I’m Star Online 1. Distance Customized non-face-to- Ministry of Small Service Pre-Foun- - Small advance- individual face education support SMEs and dational / Business ment tech. learn- for small business Startups Foundational Online education ing owners to enter online, (MSS) Sales such as shopping mall Support entry procedures and Project operations. Commer- 8. Distance Business advisory and cialization managerial technology extension consulting learning services to support com- support mercialization of prod- ucts, in order to make the suitable for online sales. Dedicated 15. Individ- Development support for Seller ual tech. online product registra- Matching learning tion and detailed page Support production, promotion, and sales of small busi- ness products. Assis- tance in creating website. Support for 15. Individ- This line offers support promotion ual tech. for the product’s digital and adver- learning design (images, videos, tisement of etc.) in order to increase products online sales. The support encourages the production and diffusion of Q-Tone advertisement content, a 2 minutes TV spot that raises awareness for small business products. V-com- 8. Distance Support for sales of small merce managerial business products in the learning YouTube Creator Channel Digital 3. Assis- Construction of facilities Commerce tance in and equipment (studios, Specialized access to editing rooms, etc.) and Institution finance business support neces- sary for entering and sup- porting the online market for digital transformation of small business owners. Establish- 15. Individ- This component provides Foundational / ment and ual tech. demonstration and oper- Intermediate operation learning ation of online-to-offline of a smart (O2O) linked stores for flagship small businesses, com- store bining content experience and product sales. A pri- vate specialized operator is selected to operate as an intermediary (called professional operators). 109 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Korea I’m Star Support for 18. Supplier Promotion of the opening Ministry of Small Service Foundational / (cont.) - Small entering devel- of special exhibitions for SMEs and Intermediate Business online opment small businesses and Startups Online shopping program support product sales in (MSS) Sales malls online shopping malls; Support Support for creating  List of analyzed programs and instruments Project detailed pages that (cont.) systematically organize and market the features of the product to induce purchases. TV home 3. Assis- Support for broadcast shopping tance in entry fees, video produc- and T-com- access to tion costs etc. connected merce finance with mentoring in broad- entry cast preparation and support management of publicity after broadcast. V-Com- 3. Assis- Co-financing of building merce tance in firm capabilities to offer store entry access to video production and support finance sales support, incl. video planning and production. Support for 19. Indi- Trainings and advisory entering vidual regarding market entry overseas managerial to foreign e-commerce shopping learning markets, with dedicated malls training programs on US, China, Japan, and South- east Asia. Live com- 8. Distance Real-time sales and pro- merce pro- managerial motion of small business duction and learning products through live operation commerce for non-face- support to-face digital conversion for small business own- ers through live com- merce operation, direct production support for small business owners. Live 15. Individ- Support in promotional content ual tech. videos that can be used creation Learning for online sales platforms, support SNS and offline, YouTube advertisements, and TV advertisements. The videos’ production can be tailored using influencers, reviewed by experts and famous celebrities. Malaysia SME SME 3. Assis- Matching grant consist- Ministry of Small / Horizon- Foundational Business Business tance in ing of 50% of co-funding Finance Medium tal Digital- Digitalisa- access to to adopt digitalization / Large isation tion Grant finance of processes in: digital Grant marketing and sales, electronic point of sales (e-POS), HR Payroll Sys- tem/CRM, procurement, e-Commerce. remote working, and ERP/ accounting and tax. 110 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Malaysia #Smart Smart 5. Small and Small scale grants (up Malaysia Small / Service Foundational / Automa- Automation temporary to USD 25k) to assist Digital Medium Intermediate tion Grant Grant - subsidies service companies in Economy (SAG) MDEC adopting digital solutions Corpo- to increase revenue, save ration costs, reduce worktime, (MDEC)  List of analyzed programs and instruments and create new sources of growth. Smart 3. Assis- Matching grant for cov- Malaysia Horizon- Intermediate Automation tance in ering up to 50% and USD Devel- tal Grant - access to 100k of investments in opment MIDA finance automatization projects Authority in firms which have (MIDA) already implemented digitalization in their business model. Malaysia Global Nurturing 18. Supplier Support to local scale-up Malaysia Small / Tech- High Tech Global devel- technology companies to Digital Medium nology Fund Champions opment go into their first global Economy / Large compa- (GTF) program market or expand their Corpo- nies existing global market ration presence via technol- (MDEC) ogy innovation and commercialization of market-driven product/ service. Investment 17. Exper- Financing for foreign in R&D imental technology companies to devel- set up Centre of Excel- opment lence to conduct high programs value technology innova- tion/R&D activities lead- ing to the development and commercialization of market driven, innovative product/service for the global market as well as contribute to ecosystem development. Catalyzing 11. Manage- Assistance to Technology Horizon- Intermediate / Digital rial devel- Accelerators, who can tal High Innovation opment help startup companies Ecosystem program aiming to expand their operations to inter- national markets. The technology accelerator should include providing funding, mentorship, and market access. Malaysia Industry- Readiness 2. Informa- The Readiness Assess- Ministry of Small / Manufac- Foundational / 4WRD Assess- tional and ment consists of an Inter- Medium turing Intermediate ment (RA) promotional on-site visit (min 2 national / Large campaign days/2 auditors). After 1.5 Trade and months and receiving the Industry RA report, the benefi- (MITI) ciary can apply for a 1year intervention. 111 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Malaysia Indus- Domestic 17. Exper- The Industry4WRD DISF Ministry of Small / Manufac- Intermediate / (cont.) try4WRD Investment imental provides matching grant Inter- Medium turing High (cont.) Strate- devel- in the form of 60:40 ratio national / Large gic Fund opment based on a reimburs- Trade and (DISF) programs able basis to Malaysian Industry companies which have (MITI)  List of analyzed programs and instruments undertaken the Indus- try4WRD RA programme. This grant funds: R&D activities; training activ- ities; modernization and upgrading of facilities and equipment; licens- ing or purchase of new/ high technology; and obtaining international standards/certification. Portugal Coaching Coaching 11. Manage- A voucher-like incentive Minis- Small / Horizon- High 4.0 4.0 rial devel- to boost the technological try for Medium tal opment component of companies Digitali- program by supporting zation and process development Admin- and organizational istrative competencies that drive Modern- the business model’s ization digital transformation. Portugal Techno- Techno- 14. Tech- A “safe space” in which National Small / Horizon- Foundational / logical logical nological companies can test Innovation Medium tal Intermediate Free Free Zones testbeds innovative products, Agency Zones (TFZ) services, business (ANI) (TFZ) models, and delivery mechanisms without immediately incurring all the normal regulatory consequences related to the activity in question. New Digital Spotlight 7. Show- A series of real world, Ministry of Micro / Horizon- Pre-Foun- Zealand Boost Series casing in-depth video case Business, Small tal dational / examples of studies where small Innova- Foundational successful business owners, who tion and transfor- have recently trans- Employ- mation formed their business by ment adopting digital tools and digital ways of working, share their experiences. Educate 1. Distance A free digital skills train- individual ing platform available to tech. learn- any small business owner ing (or their employees) who wants to boost or acceler- ate their digital trans- formation. The training platform offers over 500 video tutorials and Q&A sessions, daily live work- shops with experts and live helpdesk support. 112 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication New Digital Checkable 8. Distance An extension of Digital Ministry of Micro / Horizon- Pre-Foun- Zealand Boost managerial Boost Educate, the Business, Small tal dational / (cont.) (cont.) learning Checkable solution is Innova- Foundational focused on supporting tion and small businesses to Employ- priorities where they ment  List of analyzed programs and instruments should focus their time and effort on digitalizing their business for the best return. The solution will provide business owners with a customized Digital Action Plan. Digital 10. Group Provides additional Facilitation tech. learn- support and encourage- Scheme ing ment to small businesses to adopt and practi- cally use Digital Boost. Designed for those who are at the initial stages of their learning journey, business intermediaries are funded to provide facilitated learning support to groups of small businesses in their community. Alliance 2. Informa- A collaborative effort Aotearoa tional and between the government promotional and larger-scale private campaign sector organizations focused on motivating and inspiring small busi- nesses, individuals, and communities across New Zealand to lift their use of digital technologies. Each Alliance partner has committed to providing specific offers or services to help businesses succeed in their digital journey. Singapore SMEs Go Digital 3. Assis- Uplifting the digital Infocomm Small / Retail Pre-Foun- Digital Resilience tance in capabilities of enter- Media Medium and Con- dational / Bonus access to prises most affected by Devel- sumer Foundational (DRB) finance safe distancing require- opment Services ments in the post-pan- Authority demic reopening of the (IMDA) economy. Financing for advisory and purchase of Business Process Solu- tions, Digital Presence, e-Commerce and Data Mining and Analytics. 113 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Singapore SMEs Go Digital 12. Market- Repository of digital Infocomm Small / Horizon- Pre-Foun- (cont.) Digital Solutions places platforms, offering Media Medium tal dational / (cont.) for Safe solutions to the business Devel- Foundational Reopening to safely operate in a opment social distancing context Authority or reopening, incl. online (IMDA)  List of analyzed programs and instruments collaboration, virtual meetings, e-Signatures, remote workforce man- agement, visitor man- agement and payment technologies. Pre-Ap- 9. Technol- Support for MSMEs to Foundational proved ogy demon- adopt pre-approved Solutions stration solutions assessed to be off-site market-proven, cost-ef- fective and supported by reliable vendors. PSG supports the adoption of pre-scoped IT solutions, equipment and consultancy services that improves productiv- ity, aligned with digital transformation plans Grow 8. Distance Under this initiative, Digital managerial SMEs can participate learning in Business-to-Busi- ness (B2B) and Busi- ness-to-Consumer (B2C) e-commerce platforms to sell overseas with- out a need for physical presence. Chief 20. CTO Access a one-stop web Intermediate / Technology extension application where SMEs High Officer-as- scheme can assess their digital a-Service maturity and digitali- (CTOaaS) zation needs, followed up with tailored recom- mendations on digital solutions they can adopt, including the relevant grant support available, based on their identified needs, enterprise profile and business goals. For SMEs that need more in-depth digital advisory, access to a shared pool of CTO-equivalents, or Dig- ital Consultants to help SMEs identify their dig- italization needs, select suitable digital solutions and project manage the solution implementation. 114 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Singapore SMEs Go Standards 2. Informa- A nationwide E-invoicing Infocomm Micro / Horizon- Foundational / (cont.) Digital tional and method that facilitates Media Small / tal Intermediate (cont.) promotional the direct transmission Devel- Medium campaign of invoices in a structured opment digital format across Authority finance systems. (IMDA)  List of analyzed programs and instruments Industry 2. Informa- Access to an easy-to-use, Micro / Digital tional and step-by-step guide on Small Plans (IDP) promotional the digital solutions to campaign adopt at each stage of growth and the training programmes to enhance employees’ digital skills. Start Digi- 4. Web- Vouchers for foundational Pre-Foun- tal Pack based and easy-to-deploy dational / technology digital solutions to newly Foundational demonstra- incorporated SMEs, or tion those that have yet to digitalize, to help them start their digitalization process. SME Digital 16. Technol- Specialist digital technol- Small / Intermediate / Tech Hub ogy demon- ogy advisory to SMEs with Medium High stration more advanced digital on-site needs, such as data ana- lytics and cybersecurity. Advanced 20. CTO Grants to adopt more High Digital extension advanced solutions to Solutions scheme deepen firms’ capabili- (ADS) ties, strengthen business continuity measures and build longer term resilience. Solutions supported under ADS address common firm-level challenges at scale, help enterprises to adopt cutting-edge technologies and enable enterprises to transact more seamlessly within or across sectors. Singapore Tech Tech 14. Tech- Providing access A*STAR Small / Manufac- High Access Access nological advanced demonstration Medium turing testbeds labs and technological testbeds Tech Access provides a cost-effective way for SMEs to make informed equipment investment decisions, thereby mitigating upfront risk and capital expenditure. 115 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Singapore Collab- Collab- 12. Market- The A*STAR Collab- A*STAR Small / Horizon- Intermediate / orative orative places orative Commerce Medium tal High Com- Commerce Marketplace (ACCM) is / Large merce Market- a government-driven Market- place e-portal where busi- place (ACCM) nesses collaborate to find  List of analyzed programs and instruments (ACCM) business and technol- ogy synergies. It is a complimentary platform for community building & development, allowing enterprises to network, form relationships & foster collaborations. Singapore Digital Digital 19. Indi- Support for companies Infocomm Micro / Horizon- High Leaders Leaders vidual who are ready to integrate Media Small / tal Pro- Pro- managerial digital technology into their Devel- Medium gramme gramme learning core business strategy to opment (DLP) (DLP) develop new digital capa- Authority bilities, so that they can (IMDA) develop innovative busi- ness models and capture new growth opportunities. Three support modules: 1. Build an in-house core digital team; 2. Engage consultancy services to develop digital road- maps; 3. Develop Proof of Concepts (POCs) for new markets/ customer segments. Singapore Heart- Heartlands 3. Assis- Support for integrating Enterprise Micro / Services Pre-Foun- lands Go Go Digital tance in technologies necessary Singapore Small dational / Digital access to for e-Commerce by (ESG) Foundational finance co-financing services of leading providers of e-payment and digital commerce solutions. Spain Acelera Offices and 2. Informa- A network of physical Ministry of Micro / Horizon- pyme portal tional and and virtual information Economic Small / tal promotional points to raise knowledge Affairs Medium campaign of digital transformation and Digital and disseminate knowl- Transfor- edge on practical benefits mation for firms. Spain Cloud Supply 6. Voca- Funding for firms devel- red.es Micro / Tech- High Comput- develop- tional oping cloud services Small nology ing ment training for to finance personnel, compa- employees subcontract technologi- nies cal services and contract technological evaluators from third parties. Demand 12. Market- Co-financing adoption Micro / Horizon- Foundational / develop- places of cloud-based comput- Small / tal Intermediate ment ing tools from a curated Medium database of certified providers. The instrument supports adoption of 10 broadly used technolo- gies in general business functions. 116 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Spain Asesores Advisory 2. Informa- Specialized and person- red.es Micro / Horizon- Pre-Foun- Digitales vouchers tional and alized advisory support Small tal dational / (Digital promotional in creating firm-level Foundational Advisors) campaign transformation plan, followed-up by contacting enterprise with certified  List of analyzed programs and instruments providers of digital services. Spain Ofici- Digital 2. Informa- Voucher for education red.es Micro / Horizon- nas de Transfor- tional and and training for SMEs, Small tal Transfor- mation promotional establishing collab- mación Offices campaign orations and clusters. Digital Instrument focuses on (Digital popularizing awareness Transfor- of digital transformation mation among the highest num- Offices) ber of SMEs possible. Spain Industria ACTIVA 19. Indi- Specialized and person- General Small / Manufac- Intermediate / Conect- Industria vidual alized advisory program Secre- Medium turing High ada 4.0 4.0 managerial carried out by accredited tariat of learning consultants with expertise Industry in the implementation and SMEs of Industry 4.0 projects. The program is carried out with a methodology developed by the General Secretariat for Industry and SMEs. This program allows com- panies to have an assess- ment and a transformation plan that identifies the digital enablers needed and establishes the roadmap for their imple- mentation. The consulting is complemented by workshops demonstrating enabling technologies. ACTIVA 15. Individ- Support the incorporation Horizon- Financia- ual tech. of knowledge, technol- tal cion Learning ogies, and innovations targeting the digitaliza- tion of processes and the creation of technologi- cally advanced products and services with greater added value (complex digital solutions). ACTIVA 17. Exper- Facilitating connection Micro / Specific, High Retos imental between startups and Small priority Industri- devel- recognized industrial sectors ales (12 opment companies in Spain challenges programs and thus promote open of the innovation. Industry After identifying the 4.0) technological challenges of 12 large industrial com- panies, the call seeks to select as many startups that respond to them through solutions based on the digital enablers of Industry 4.0. 117 Program Instrument Instrument Managing Firm Business Digital Annex 2  Country name name type Short description authority size sector sophistication Spain Industria ACTIVA 15. Individ- Business advisory and General Micro / Horizon- Pre-Foun- (cont.) Conect- Cibersegu- ual tech. technology exten- Secre- Small tal dational / ada 4.0 ridad learning sion services aimed to tariat of Foundational (cont.) increase cybersecurity Industry among SMEs through and SMEs diagnosis, implementa-  List of analyzed programs and instruments tion, and post-implemen- tation support. ACTIVA 19. Indi- Provision of customized Crec- vidual consulting for SMEs - 50 imiento managerial hours with advisor per learning firms - to analyze needs and address low-hanging fruit in digitalization. Spain AI Tech- AI Tech- 17. Exper- Comprehensive financial res.es Small / Horizon- High nological nological imental support for experimental Medium tal develop- develop- devel- development and testing / Large ment ment opment for digital enabling tech- programs nologies based on artifi- cial intelligence or other pre-determined digital enabling technologies. Uruguay Digital Digital 1. Distance Firms engage in self-di- National Micro / Horizon- Pre-Foun- Mode Mode individual agnosis of their digital Devel- Small tal dational / tech. learn- needs and will receive opment Foundational ing recommendations such Agency as: advisory through (ANDE) online content; support to choose and acquire basic digital solutions from the market; digital transformation plans for more advanced firms; capabilities to learn to handle proposed recom- mendations. United Help to Advice and 1. Distance Advice and Learning Depart- Micro / Horizon- Pre-Foun- Kingdom Grow: learning individual section includes useful ment for Small tal dational / Digital tech. learn- information about how Business, Foundational ing to help identify your Energy software needs, finding and the best supplier, getting Industrial the best return on your Strategy investment and how to integrate software into your business. Discount 12. Market- Firms can apply for a places discount in the purchase of the following software types from approved suppliers: Customer Relationship Manager (CRM); Digital Account- ing; eCommerce. Source: Digitrans team. 118 ANNEX 3 EVIDENCE OF IMPACT OF INSTRUMENTS FOR ADDRESSING GAPS IN COMPLEMENTARY CAPABILITIES Type of evidence objectives related to private sector including monitoring reports, self- of instruments for digitalization of other relevant objectives related of instruments for other relevant reported or qualitative evidence experimental impact evaluation experimental impact evaluation instruments for digitalization of 2. Other types of evaluations of 4. Other types of evaluations of 5. Non-evaluation documents, 3. Experimental or quasi- 1. Experimental or quasi- catching-up MSMEs catching-up MSMEs to private sector Instruments References 1. Distance tech. (3) Comin, Trumbull and Yang 2016; Gupta, individual learning • Ponticelli and Tesei 2020; Ichimura and Todd 1997; 2. Informational (1) Dalton et al 2021; (3) Casaburi et al. 2018; and promotional campaigns • • de Andrade, Bruhn, and McKenzie 2014; Anderson and Feder 2004; (4) Jensen 2010; Dupas 2011; 3. Assistance in access to finance • • (1) Drexler et al 2014; (3) Liu 2013; Gine and Yang 2009; 4. Web-based (3) McKenzie 2013; Comin, Trumbull; Casaburi technology demonstration • et al. 2018; 5. Small and (3) Duflo, Kremer, and Robinson 2011; Atkin temporary subsidies • • et al 2017; and Yang 2016 (4) Foster and Rosenzweig 2010; 6. Vocational training for employees • (3) Card, Kluve and Weber 2017; Ichimura and Todd 1997; (2) Rivard and Laponite 2012; 7. Showcasing (1) Brooks et al 2018; (2) Oster and Thornton examples of successful transformation • • • • 2012; (3) Casaburi et al. 2018; del Mel, McKenzie and Woodruf 2013; (4) Foster and Rosenzweig 2010; 8. Distance managerial learning • • (1) Anderson and McKenzie 2021; (3) Knight, Weir, and Woldehanna, 2003 9. Technology (2) Cirera et al. 2020; Schneider 2015; demonstration off- site • 119 Type of evidence Annex 3  objectives related to private sector including monitoring reports, self- of instruments for digitalization of other relevant objectives related of instruments for other relevant reported or qualitative evidence experimental impact evaluation experimental impact evaluation instruments for digitalization of 2. Other types of evaluations of 4. Other types of evaluations of 5. Non-evaluation documents, 3. Experimental or quasi- 1. Experimental or quasi- catching-up MSMEs catching-up MSMEs  Evidence of impact of instruments for addressing gaps in complementary capabilities to private sector Instruments References 10. Group technology (1) Iacovone, Maloney and McKenzie 2022; learning • • Cirera et al. 2020; Saphira, Youtie and Kay 2011; 11. Managerial (1) Vladiva 2015; (3) Rivard and Lapoint 2012; development programs • • Bloom et al. 2013; de Andrade, Bruhn, and McKenzie 2014 12. Marketplaces (1) Anderson and McKenzie 2021; (3) Tadelis • • 2016; Aurol and Schilizzi 2015; Williamson 1985; 13. Lump-sum payments for workers • • (3) Banerjee et al 2011; Athin et al. 2017; (5) McKinsey 2015; 14. Technological testbeds • • (2) Cirera et al. 2020; Saphira, Youtie and Kay 2011; (3) de Giorgi and Rahman 2013; 15. Individual (1) Roberts and Shaw 2022; (3) de Giorgi and technology learning • • Rahman 2013; del Mel, McKenzie and Woodruf 2013; 16. Technology (3) Foster and Rosenzweig 2010; (4) demonstration on-site • • Schnedeider 2015; 17. Experimental (2) Middelstand; (3) Benhassine et al. 2018; development programs • • 18. Supplier (2) Benton, Prainski, Fan 2020; (4) Arraiz, development programs • • Henriquez, Stucci 2012; 19. Individual managerial learning • • (1) Sonobe et al. 2016; (3) Bruhn, Karlan and Schoar 2018; 20. CTO extension schemes • • (2) IMDA 2020; (3) Benhassine et al. 2019; Source: Digitrans team Note: IE = Impact evaluation 120 BIBLIOGRAPHY Abay KA, Blalock G, Berhane G. 2017. Locus of Control and Behaghel L, Caroli E, Roger M. 2014. Age-biased Technical and Technology Adoption in Developing Country Agriculture: Organizational Change, Training and Employment Pros- Evidence From Ethiopia. Journal of Economic Behavior & pects of Older Workers. Economica 81(322):368 – 89. Organization. 143:98 – 115. Bellantuono N., Nuzzi A., Pontradolfo P., Scozzi B. 2021. Dig- Abdrazakova T, Salihudin S. 2022. COVID-19 and Technol- ital Transformation Models for the I4.0 Transition: Les- ogy Adoption in Small and Medium-Sized Enterprises: The sons from the Change Management Literature. Sustainabil- Impact and Way Forward. World Economic Forum White ity 13(12941). Paper (World Economic Forum, Geneva, 2022). Benhassine N, McKenzie D, Pouliquen V, Santini M. 2018. Does Acemoglu D, Restrepo P. 2019. Automation and New Tasks: Inducing Informal Firms to Formalize Make Sense? Experimen- How Technology Displaces and Reinstates Labor. Journal of tal Evidence from Benin. Journal of Public Economics. 157:1 – 14. Economic Perspectives. 33(2):3 – 30. Bennett RJ. 2012. Government Advice Services for SMEs: Aghion P, Antonin C, Bunel S, Jaravel X. 2020. What Are the Some Lessons from British Experience. Government, SMEs Labor and Product Market Effects of Automation? New and Entrepreneurship Development. Routledge. Evidence from France. Research Briefs in Economic Policy. BenYishay A, Mobarak AM. 2019. Social Learning and Incen- (225):1 – 2. tives for Experimentation and Communication. Review of Albaz A, Dondi M, Rida T, Schubert J. 2020. Unlocking Growth Economic Studies 86 (3): 976 – 1009. in Small and Medium-Sized Enterprises. McKinsey Quarterly. Bloom N, Scur D, Sadun R, Van Reenen J, Lemos R. 2021. World Anderson JR, Feder G. 2004. Agricultural Extension: Good Management Survey at 18: Lessons and the Way Forward. Intentions and Hard Realities. The World Bank Research Harvard Business School Working Paper 21 – 093. Observer. 19(1):41 – 60. Bloom N, Eifert B, Mahajan A, McKenzie D, Roberts J. 2013. Anderson SJ, McKenzie D. 2021. What Prevents More Small Does Management Matter? Evidence from India. The Quar- Firms from Using Professional Business Services? An terly Journal of Economics 128 (1): 1 – 51. Information and Quality-Rating Experiment in Nigeria. Bloom N, Iacovone L, Pereira – Lopez M, Van Reenen J. 2022. Washington, DC: World Bank. Management and Misallocation in Mexico. National Bureau Aral S, Brynjolfsson E, Wu L. 2012. Three-Way Complementar- of Economic Research Working Paper 29717. ities: Performance Pay, Human Resource Analytics, and Bloom N, Sadun R, and Van Reenen J. 2012. The Organization Information Technology. Management Science. 58(5):913 – 31. of Firms Across Countries. The Quarterly Journal of Econom- Aridi, A., Kapil, N. (2019). Innovation Agencies: Cases from Develop- ics 127(4): 1663 – 1705 ing Economies. World Bank. Bloom N. and Van Reenen J. 2010. Why Do Management Prac- Arrow K. 1962. Economic Welfare and the Allocation of tices Differ Across Firms and Countries? Journal of Economic Resources for Invention. In The Rate and Direction of Inven- Perspectives 24(1): 203 – 24 tive Activity: Economic and Social Factors, pp. 609 – 26. Prince- Borowiecki M, Pareliussen J, Glocker D, Kim EJ, Polder M, Rud ton University Press I. 2021. The Impact of Digitalization on Productivity: Firm- Ashenfelter O. and Card D. 1985. Using the Longitudinal Struc- level Evidence from the Netherlands. OECD Economics ture of Earnings to Estimate the Effect of Training Programs. Department Working Papers No. 1680. The Review of Economics and Statistics 67(4): 648 – 660 Branca TA, Fornal B, Colla V, Murri MM, Steppa E, Schroeder Atkin D, Chaudhry A, Chaudry S, Khandelwal AK, and Verhoo- AJ. (2020). The Challenge of Digitization in the Steel Sec- gen E. 2017. Organizational Barriers to Technology Adop- tor. Metals 10(2). tion: Evidence from Soccer-Ball Producers in Pakistan. The Bresnahan TF, Brynjolfsson E, Hitt LM. 2002. Information Tech- Quarterly Journal of Economics 132(3):1101 – 64. nology, Workplace Organization, and the Demand for Audretsch DB, Feldman MP. 1996. R&D Spillovers and the Skilled Labor: Firm-Level Evidence. The Quarterly Journal of Geography of Innovation and Production. The American Eco- Economics. 117(1):339 — 76. nomic Review. 86(3):630 – 40. Bresnahan TF, Trajtenberg M. 1995. General Purpose Technolo- Auriol E, Schilizzi SGM. 2015. Quality Signaling through Certifi- gies ‘Engines of Growth’? Journal of Econometrics. 65(1):83 — 108. cation in Developing Countries. Journal of Development Eco- Brooks W, Donovan K, Johnson T R. 2018. Mentors or Teach- nomics. 116:105 – 21. ers? Microenterprise Training in Kenya. Applied Economics Azevedo A, Almeida AH. 2021. Grasp the Challenge of Digi- 10(4) 196 — 221 tal Transition in SMEs—A Training Course Geared towards Bruhn M, Karlan D, Schoar A. 2018. The Impact of Consulting Decision-Makers. Education Sciences 11(151). Services on Small and Medium Enterprises: Evidence from Baldwin J. 2002. Impediments to advanced technology adop- a Randomized Trial in Mexico. Journal of Political Economy. tion for Canadian manufacturers. Research Policy 31(1): 1 – 18 126(2):635 — 87. Baldwin J, Lin Z. 2002. Impediments to Advanced Technol- Brunello G, de Paola M. 2004. Training and the Density of Eco- ogy Adoption for Canadian Manufacturers. Research Pol- nomic Activity: Evidence from Italy. IZA Discussion Papers 1173. icy. 31(1):1 – 18. Brynjolfsson E, Hitt LM. 2000. Beyond Computation: Infor- Bandiera O, Rasul I. 2006. Social Networks and Technology mation Technology, Organizational Transformation and Adoption in Northern Mozambique. The Economic Journal Business Performance. Journal of Economic Perspectives. 116 (514): 869 – 902. 14(4):23 — 48. Beaman L., Karlan D., Thuysbaert B. and Udry C. 2013. Profitabil- Brynjolfsson E, Milgrom P. 2013. Complementarity in Organiza- ity of Fertilizer: Experimental Evidence from Female Rice tions. The Handbook of Organizational Economics. 11 — 55. Farmers in Mali. American Economic Review 103(3): 381 – 86 Brynjolfsson E, Rock D, Syverson C. 2018. Artificial Intelligence Beaman L, BenYishay A, Magruder J, Mobarak AM. 2018. Can and the Modern Productivity Paradox: A Clash of Expecta- Network Theory-based Targeting Increase Technology tions and Statistics. In The Economics of Artificial Intelligence: Adoption? Working Papers 1062, Economic Growth Center, An Agenda, pp. 23 — 57. University of Chicago Press Yale University. Buffington C, Jarmin RS. 2017. The Management and Organiza- Beaman L, BenYishay A, Magruder J, Mobarak AM. 2021. Can tional Practices Survey (MOPS): An Overview. Journal of Eco- Network Theory-based Targeting Increase Technology nomic and Social Measurement 42(1). Adoption? American Economic Review 111(6). Business Development Bank of Canada. (2022). 2021 Annual Beaman L, Karlan D, Thuysbaert B, Udry C. 2013. Profitability of Report. Fertilizer: Experimental Evidence from Female Rice Farm- Cai J, Szeidi A. 2018. Interfirm Relationships and Business Per- ers in Mali. American Economic Review. 103(3):381 – 86. formance. The Quarterly Journal of Economics 133(3): 1229 — 82. Beaman, L., Magruder J., and Robinson J. 2014. Minding Small Cantner U, Kösters S. 2012. Picking the Winner? Empirical Evi- Change among Small Firms in Kenya. Journal of Develop- dence on the Targeting of R&D Subsidies to Start-ups. ment Economics 108 (C): 69 – 86. Small Bus Econ. 39(4):921 — 36. 121 Card D, Kluve J, Weber A. 2018. What Works? A Meta Analysis of Digital Innovation Hubs Working Group. 2018. Report from the Bibliography Recent Active Labor Market Program Evaluations. Journal Working Group Meeting on Access to Finance. European Com- of the European Economic Association. 16(3):894 — 931. mission DG CONNECT. Casaburi L, Kremer M, Mullainathan S, Ramrattan R. 2014. Har- Dietz M, Vinayak HV, Lee G. 2017. Bracing for Seven Critical nessing ICT to Increase Agricultural Production: Evidence Changes as Fintech Matures. McKinsey Quarterly. From Kenya. Duflo E, Kremer M, Robinson J. 2011. Nudging Farmers to Use Catlin T., Scanlan J. and Willmott P. (2015). Raising your Digital Fertilizer: Theory and Experimental Evidence from Kenya. Quotient. McKinsey Quarterly June 2015. American Economic Review 101, no. 6 (October):2350 — 90. Cirera X, Comin D, Cruz M. 2022. Bridging the Technological Dupas P. 2011. Health Behavior in Developing Coutries. Annual Divide: Technology Adoption by Firms in Developing Coun- Review of Economics 3 : 425 — 449 tries. The World Bank Productivity Project. Washington, DC: Drexler A, Fischer G, Schoar A. 2014. Keeping It Simple: Finan- World Bank. cial Literacy and Rules of Thumb. Applied Economics 6(2): Cirera X, Comin D, Cruz M, Lee K. 2022. Technology Within and 1 — 31 Across Firms. Policy Research Working Paper No. WPS 9476 Edquist C. 2011. Design of Innovation Policy through Diagnos- Washington, DC: World Bank Group. tic Analysis: Identification of Systemic Problems (or Fail- Cirera X, Frias J, Hill J, Li Y. 2020. A Practitioner’s Guide to Innova- ures). Industrial and Corporate Change. 20(6):1725 — 53. tion Policy: Instruments to Build Firm Capabilities and Acceler- Eggers F. 2020. Masters of disasters? Challenges and oppor- ate Technological Catch-Up in Developing Countries. Washing- tunities for SMEs in times of crisis. Journal of Business ton, DC: World Bank. Research 116: 199 — 208 Cirera X, Maloney WF. 2017. Developing-Country Capabilities and Eller R., Alford P, Kallmunzer A, Peters M. 2020. Anteced- the Unrealized Potential of Technological Catch-Up. Washing- ents, Consequences, and Challenges of Small and Medi- ton, DC: World Bank. um-sized Enterprise Digitization. Journal of Business Cole S., and Fernando A.N., 2021. ‘Mobile’izing Agricultural Research 112:119 — 27. Advice Technology Adoption Diffusion and Sustainability. European Investment Bank. 2019. Financing the Digitalization of The Economic Journal. 131(633):192 — 219. Small and Medium-sized Enterprises. The Enabling Role of Digi- Cole S., Fernando A. N., Stein D., Shah V., Litvine L., Tobacman J., tal Innovation Hubs. Chattopadhayay R. 2016. Reduced-form impacts of formal Eurostat. 2018. Community Survey on ICT Usage and e-Commerce insurance against weather shocks. International Initiative for in Enterprises. Impact Evaluation Fafchamps M, Quinn S. 2018. Networks and Manufacturing Cole S., Fernando A. N., Stein D., Tobacman J. 2020. Field com- Firms in Africa: Results from a Randomized Field Experi- parisons of incentive-compatible preference for elicita- ment. The World Bank Economic Review. 32(3):656 — 75. tion techniques. Journal of Economic Behavior & Organiza- Foster AD, Rosenzweig MR. 2010. Microeconomics of Technol- tion. 172:33 — 56 ogy Adoption. Annual Review of Economics. Colombo M. G., Grilli L., and Murinu S. 2011. Venture Capital Friederici N. 2022. The Digital Transformation of Small Business Financing and the Growth of High-Tech Start-Ups: Disen- Support. Caribou Digital UK, Mastercard Center for inclu- tangling Treatment from Selection Effects. Research Pol- sive Growth icy 40:1028 — 1043 Ghobakhloo M, Sai Hong T, Sabouri MS, Zulkifli N. 2012. Strat- Comin D, Trumbull G, Yang L. 2016. Frauenhofer: Innovation in egies for Successful Information Technology Adoption in Germany. Drivers of Competitiveness 409 — 44. Small and Medium-sized Enterprises. Information 3:36 — 67. Correa P. 2017. Public Expenditure Reviews in Science, Technology, Gin´e X, Yang D. 2009. Insurance, Credit, and Technology Adop- and Innovation. Washington, DC: World Bank. tion: Field Experimental Evidence from Malawi. Journal of Crespi G, Criscuolo C, Haskel J. 2007. Information Technology, Development Economics 89 (1):1 — 11. Organisational Change and Productivity. CEPR Discussion Gray C. 2002. Entrepreneurship, Resistance to Change and Paper No. DP6105. Growth in Small Firms. Journal of Small Business and Enter- Crespi GA, Maffioli A. 2014. Design and Evaluation of Fiscal prise Development 9(1):61 — 72. Incentives for Business Innovation in Latin America: Les- Grindle MS, Hilderbrand ME. 1995. Building Sustainable Capac- sons Learned After 20 Years of Experimentation. In Science, ity in the Public Sector: What Can Be Done? Public Adminis- Technology and Innovation Policies for Development: The Latin tration and Development 15(5):441 — 463. American Experience, eds. G Crespi, G Dutrénit, pp. 225 — 53. Gugerty MK, Karlan D. 2018. Ten Reasons Not to Measure Cham: Springer International Publishing. Impact—and What to do Instead. Stanford Social Innovation Cusolito A. 2021. The Economics of Technology Adoption: Key Driv- Review. ers, Barriers and Policy Instruments to Boost Take-up. Gupta A, Ponticelli J, Tesei A. 2020. Information, Technology Dalton PS, Rüschenpöhler J, Uras B, Zia B. 2021. Curating Local Adoption and Productivity: The Role of Mobile Phones in Agri- Knowledge: Experimental Evidence from Small Retailers culture. NBER Working Paper 27192. in Indonesia. Journal of the European Economic Association. Hannan M. T., and Freeman J. 1984. Structural Inertia and 19(5):2622 — 57. Organizational Change. American Sociology Review 49(2): David PA. 1990. The Dynamo and the Computer: An Historical 149 — 164 Perspective on the Modern Productivity Paradox. The Ameri- Haltom R, Bloom N. 2014. Econ Focus. Federal Reserve Bank can Economic Review. 80(2):355 — 61. of Richmond. https://worldmanagementsurvey.org/data/ de Andrade GH, Bruhn M, McKenzie D. 2016. A Helping Hand or international-mops/ the Long Arm of the Law? Experimental Evidence on What Hameed MA, Counsell S, Swift S. 2012. A Meta-analysis of Rela- Governments Can Do to Formalize Firms. The World Bank tionships between Organizational Characteristics and IT Economic Review. 30(1):24 — 54. Adoption in Organizations. Information & Management 49(5). de Giorgi G, Rahman A. 2013. SME’s Registration: Evi- Heckman JJ, Ichimura H, Todd PE. 1997. Matching as an Econo- dence from an RCT in Bangladesh. Economics Letters. metric Evaluation Estimator: Evidence from Evaluating a 120(3):573 — 78. Job Training Programme. The Review of Economic Studies. de Janvry A., Emerick K., Sadoulet E. and Dar M. H. 2016. Tech- 64(4):605 — 54. nological Innovations, Downside Risk, and the Modern- Hudson B, Hunter D, Peckham S. 2019. Policy Failure and the ization of Agriculture. American Economic Review 106(6): Policy-implementation Gap: Can Policy Support Programs 1537 — 61 Help? Policy Design and Practice 2(1). de Mel S, McKenzie D, Woodruff C. 2013. The Demand for, and Huizingh E, Mulder M. 2015. Effectiveness of Regulatory Inter- Consequences of, Formalization among Informal Firms in ventions on Firm Behavior: A Randomized Field Experiment Sri Lanka. American Economic Journal: Applied Economics 5(2): with e-Commerce Firms. Small Business Economics 45(4). 122 — 50. Iacovone L, Maloney W, McKenzie D. 2022. Improving Manage- Desmet K, Parente SL. 2014. Resistance to Technology Adop- ment with Individual and Group-Based Consulting: Results tion: The Rise and Decline of Guilds. Review of Economic from a Randomized Experiment in Colombia. The Review of Dynamics. 17(3):437 — 58. Economic Studies 89(1), 346 — 71. 122 Ichniowski C, Shaw K. 2003. Beyond Incentive Pay: Insid- Lichtenhaler U. 2019. Building blocks of successful digital Bibliography ers’ Estimates of the Value of Complementary Human transformation: Complementing technology and market Resource Management Practices. Journal of Economic Per- issues. International Journal of Innovation and Technology spectives. 17(1):155 — 80. Management. Intarakumnerd P, Goto A. 2016. Technology and Innovation Poli- Liu EM. 2013. Time to Change What to Sow: Risk Preferences cies for Small and Medium-Sized Businesses. Asian Develop- and Technology Adoption Decisions of Cotton Farm- ment Bank Working Paper 578. ers in China. The Review of Economics and Statistics 95 (4): Ivančić L, Vukšić L, Bosilj V, Spremić, M. 2019. Mastering the 1386 — 1403. Digital Transformation Process: Business Practices and March JG, Shapira Z. 1987. Managerial Perspectives on Risk Lessons Learned. Technology Innovation Management Review, and Risk Taking. Management Science. 33(11):1404 — 18. 9(2):36 — 50. Martin S, Scott JT. 2000. The Nature of Innovation Market Fail- Jay L, Schaper M. 2003. Which Advisers Do Micro-Firms Use? ure and the Design of Public Support for Private Innova- Some Australian Evidence. Journal of Small Business and tion. Research Policy. 29(4):437 — 47. Enterprise Development 10(2). McKenzie D, Bloom N, Eifert B, Mahajan A, Roberts J. 2013. Jensen R. 2010. The (Perceived) Returns to Education and the Does Management Matter? Evidence from India. The Quar- Demand for Schooling. The Quarterly Journal of Economics. terly Journal of Economics 128(1): 1 — 51. 125(2):515 — 48. McKenzie D. 2020. Small business training to improve manage- Juhasz R, Squicciarini M, Voigtlander N. 2020. Technology Adop- ment practices in developing countries: re-assessing the tion and Productivity Growth: Evidence from Industrialization evidence for ‘training doesn’t work’. Oxford Review of Eco- in France. American Economic Review. nomic Policy 37(2): 276 — 301. Kane GC, Palmer D, Philips AN, Kiron D, Buckley N. 2015. Strat- Miorner J, Rissola G, Sorvik J, Wernberg J. 2019. Putting Digi- egy, not Technology, Drives Digital Transformation. MIT tal Innovation Hubs into Regional Context. JRC Technical Sloan Management Review Research Report. Reports. European Commission. Kane GC, Palmer D, Philips AN, Kiron D, Buckley N. 2017. MIT Sloan. 2017. Digital Transformation: A Roadmap for Business Achieving Digital Maturity. MIT Sloan Management Review Organizations. Findings from Phase 1 of the Digital Transfor- Research Report. mation Study conducted by the MIT Center for Digital Busi- Kelley EM, Lane G, Schönholzer D. 2021. Monitoring in Target ness and Capgemini Consulting. Contracts: Theory and Experiment in Kenyan Public Tran- Mokyr J. 1990. The Level of Riches: Technological Creativity sit. Center for Effective Global Action Working Paper Series and Economic Progress. The Economic Journal 101(407): 996. Paper No. 150. Meroño-Cerdán AL, López-Nicolás C, Molina-Castillo FJ. 2018. Kesley JB, Oliva P, Severen C, Walker E, Bell S. 2015. Technology Risk Aversion, Innovation and Performance in Family Firms. Adoption under Uncertainty: Take-up and Subsequent Invest- Economics of Innovation and New Technology. 27(2):189 — 203. ment. NBER Working Paper 21415. Munshi, K. 2004. Social Learning in a Heterogeneous Population: Kilimis P, Zou W, Lehmann M, Berger U. 2019. A Survey on Digitali- Technology Diffusion in the Indian Green Revolution. Journal zation for SMEs in Brandenburg, Germany. IFAC Papers 52(13). of Development Economics 73, no. 1 (February): 185 — 213 Knight J, Weir S, Woldehanna T. 2003. The Role of Education in Nadkarni S, Pruegel R. 2021. Digital Transformation: A Review, Facilitating Risk-taking and Innovation in Agriculture. The Synthesis and Opportunities for Future Research. Manage- Journal of Development Studies. 39(6):1 — 22. ment Review Quarterly 71(2). Koch L, Simmler M. 2020. How Important are Local Knowledge OECD. 2017. Strengthening SME Capabilities through a Sustainable Mar- Spillovers of Public R&D and What Drives Them? Research ket for Business Development Services. Project Report—OECD Eur- Policy. 49(7):104009. asia Competitiveness Programme. Kochan TA, Yang D, Kimball WT, Kelly EL. 2019. Worker Voice OECD. 2018. Leveraging Business Development Services for SME in America: Is There a Gap between What Workers Expect Productivity Growth—International experience and Implica- and What They Experience? ILR Review. 72(1):3 — 38. tions for United Kingdom Policy. Centre for Entrepreneurship, Kohnke O. 2016. It’s Not Just About Technology: The People SMEs, Regions, and Cities Initiative (CFE). Side of Digitization. Shaping the Digital Enterprise 69 — 91. OECD. 2019. Measuring the Digital Transformation: A Roadmap for Springer. the Future. Kuriakose S. (2022). Results from the 4th Round of COVID-19 Busi- OECD. 2020. Digital Business Diagnostic Tools for SMEs and Entre- ness Pulse Survey. World Bank. preneurship: A Review of International Policy Experiences. Labelle R. 2019. ICT Policy Formulation and e-Strategy Devel- OECD SME and Entrepreneurship Papers 21. opment: A Comprehensive Guidebook. UNDP-APDIP ICT4D OECD. 2021. Spurring Growth and Closing Gaps through Digitiza- Series. tion in a Post-COVID World: Policies to Lift All Boats. An OECD LaLonde RJ. 1986. Evaluating the Econometric Evaluations of Contribution to the G20 Italian Presidency 2021 Training Programs with Experimental Data. The American Office of Fair Trading. 2009. Government in Markets. Why Compe- Economic Review. 76(4):604 — 20. tition Matters—A Guide for Policymakers. UK Government. Lambrecht J, Pirnay F. 2015. An Evaluation of Public Support Olafsen E, Cook PA. 2016. Growth Entrepreneurship in Developing Measures for Private External Consultancies to SMEs Countries: A Preliminary Literature Review. World Bank. in the Walloon Region of Belgium. Entrepreneurship and Olejniczak K, Raimondo E, Kupiec T. 2016. Evaluation Units as Regional Development 17(2). Knowledge Brokers: Testing and Calibrating an Innovative Lazonick W. 1979. Industrial Relations and Technical Change: The Framework. Evaluation. 22(2):168 — 89. Case of the Self-Acting Mule. Cambridge Journal of Economics Osiyevskyy O, Dewald J. 2014. Inducements, Impediments, 3(3):231 — 62. and Immediacy: Exploring the Cognitive Drivers of Small Lazonick W. 1990. Organizational Capabilities in American Business Managers’ Intentions to Adopt Business Model Industry: The Rise and Decline of Managerial Capitalism. Change. Journal of Small Business Management 53(4). Business and Economic History 19:35 — 54. Oster E, Thornton R. 2012. Determinants of Technology Adop- Lee J. 2004. Discriminant Analysis of Technology Adoption tion: Peer Effects in Menstrual Cup Take-Up. Journal of the Behavior: A Case of Internet Technologies in Small Busi- European Economic Association. 10(6):1263 — 93. nesses. Journal of Computer Information Systems 44(4); 57 — 66. Parente SL, Prescott EC. 1994. Barriers to Technology Adop- Lemos R, Choudhary A, Van Reenen J, Bloom N. (2016). Man- tion and Development. Journal of Political Economy. agement in Pakistan. First Evidence from Punjab. International 102(2):298 — 321. Growth Centre. Premkumar G. 2003. A Meta-Analysis of Research on Informa- Li X, Kauffman RJ, Yu F, Zhang Y. 2014. Externalities, Incentives tion Technology Implementation in Small Business. Jour- and Strategic Complementarities: Understanding Herd nal of Organizational Computing and Electronic Commerce Behavior in IT Adoption. Information Systems and e-Business 2:91 — 121 Management 12:443 — 64. Reynolds S, Cotrino F, Ifedi C, Donthu N. 2021. An Exploratory Study of Executive Factors that Lead to Technology Adop- tion in Small Businesses. Journal of Small Business Strat- egy 30. 123 Ratliff L. J. and Fiez T. 2021. Adaptive Incentive Design. Univer- Tambe P, Hitt LM, Brynjolfsson E. 2012. The Extroverted Firm: Bibliography sity of Washington How External Information Practices Affect Innovation and Richter F. J. and Sinha G. 2020. Why Do Your Employees Resist Productivity. Management Science. 58(5):843 — 59. New Tech? Harvard Business Review. Tsoutsoura M. 2021. Family Firms and Management Practices. Roberts J, Shaw KL. 2022. Managers and the Management of Oxford Review of Economic Policy. 37(2):323 — 34. Organizations. NBER Working Papers 30730. UK Department for Business Innovation & Skills. 2013. Robson PJA, Bennett RJ. 2000. SME Growth: The Relationship Research on Understanding Localized Policy Interventions with Business Advice and External Collaboration. Small in Business Support and Skills. BIS Research Paper Num- Business Economics 15(3), 193 — 208. ber 156. Rosenberg N. 1982. Inside the black box: Technology and eco- Urbach N, Drews P, Ross J. 2017. Digital Business Transforma- nomics. Cambridge University Press. tion and the Changing Role of the IT Function. MIS Quar- Ruegg R, Feller I. 2003. A Toolkit for Evaluating Public R&D terly Executive 16(2) Investment. NIST GCR. Jul:03 — 875. Valbonesi P, Biagi F. 2016. Incentivizing Innovation and Adop- Rupeika-Apoga R, Bule L, Petrovska K. 2022. Digital Transfor- tion of ICT: ICT Innovation Voucher Programmes. EC Joint mation of Small and Medium Enterprises: Aspects of Pol- Research Center. icy Support. Journal of Risk and Financial Management 15(2). Verhoef PC, Thijs B, Yakov B, Abhi B, Qi Dong J, Nicolai F, Rush H, Bessant J, Hobday M. 2007. Assessing the Technolog- Haenlein M. (2021). Digital Transformation: A Multidisci- ical Capabilities of Firms: Developing a Policy Tool. R&D plinary Reflection and Research Agenda. Journal of Business Management 37(3), 221 — 36. Research 122(889 — 901). Santos-Pereira C, Durao N, Moreira F, Veloso B (2022). The Verhoogen E. 2021. Firm-Level Upgrading in Developing Coun- Importance of Digital Transformation Plans in Business. tries. NBER Working Paper No. 29461. Sustainability 14(834): 1 — 26. Valdivia, Martín, 2015. Business training plus for female entre- Saxenian A. 1994. Regional Networks: Industrial Adaptation in preneurship? Short and medium-term experimental evi- Silicon Valley and Route 128. dence from Peru. Journal of Development Economics 113(C), Schizas E, Jarvis R, Daskalakis R. 2012. SMEs and their Advis- 33 — 51 ers: Measuring Trust and Confidence. ACCA Research Wallace LG, Sheetz SD. 2014. The Adoption of Software Mea- Report 127. sures: A Technology Acceptance Model (TAM) Perspective. Schnedeider M. 2015. Digitalization of production, human capi- Information & Management 51(2), 249 — 259 tal, and organizational capital. The Impact of Digitization in Williamson O. E. and Riordan M. H. 1985. Asset specificity and Workspace 39 — 52, Springer economic organization. International Journal of Industrial Scur D, Sadun R, Van Reenen J, Lemos R, Bloom N. 2021. World Organization. 3(4): 365 — 378  Management Survey at 18: Lessons and the Way Forward. World Bank. 2019. Return on Investment of Public Support to SMEs Harvard Business School Working Paper 21 — 093. and Innovation in Poland. Washington, DC: World Bank. Shapira P, Youtie J. 2014. Impact of Technology and Innovation World Bank. 2020a. Europe 4.0 Addressing Europe’s Digital Advisory Services. Nesta Working Paper No. 13/19. Dilemma. Washington, DC: World Bank. Shapira P, Youtie J. 2016. The Impact of Technology and Inno- World Bank. 2020b. Return on Investment of Public Support to vation Advisory Services. Handbook of Innovation Policy SMEs and Innovation in Poland. Functional Analysis Technical Impact. 161 — 95. Note. Washington, DC: World Bank. Shapira P, Youtie J, and Kay L. 2011. National innovation sys- World Bank. 2022. Paths and Drivers of Productivity Growth in tems and the globalization of nanotechnology innovation. Poland: A Firm Level Perspective on Technology Adoption and The Journal of Technology Transfer 36:587 — 604. Firm Capabilities. Washington, DC: World Bank Group. Sonobe T, Higuchi Y, Nam V. H. 2017. Management skill, entrepre- Xero. 2021. Behavioral Barriers to Technology Adoption amongst neurial mindset , and enterprise survival : Evidence from ran- Small Businesses—and how to overcome them. domized experiments and repeated surveys in Vietnam. Michi- Yousafzai S, Foxall G, Pallister J. 2008. Technology Accep- gan State University. tance: A Meta-analysis of the Technology Acceptance Sorbe S, Gal P, Nicoletti G, and Timiliotis C. 2019. Digital Divi- Model. Journal of Modelling in Management 2(3). dend: Policies to Harness the Productivity Potential of Digi- Zaoui F, Souissi N. 2020. Roadmap for Digital Transformation: A tal Technologies. OECD Economic Policy Paper No 26 Literature Review. Computer Science 175(1). Srinivasan R, Lillen GL, Rangaswamy A. 2002. Technological Zolas N, Kroff Z, Brynjolfsson E, McElheran K, Beede DN, et al. Opportunism and Radical Technology Adoption: An Appli- 2020. Advanced Technologies Adoption and Use by U.S. cation to E-Business. Journal of Marketing 66(3). Firms: Evidence from the Annual Business Survey. NBER Stamoulis DS. 2022. Management and Technical IT Priorities Working Paper No. 28290. for Digital Organizations. European Journal of Business and Management Research 7(1). Strilets V, Frolov S, Datsenko V and Tymoshenko O. 2022. State support for the digitization of SMEs in European countries. Problems and Perspectives in Management. 20(4):290 — 305 Tadelis S. 2016. Reputation and Feedback Systems in Online Platform Markets. Annual Review of Economics. 8:321 — 40. 124