October 2022 A SPIKY DIGITAL BUSINESS LANDSCAPE What Can Developing Countries Do? Tingting Juni Zhu, Philip Grinsted, Hangyul Song, Malathi Velamuri A Spiky Digital Business Landscape  I A SPIKY DIGITAL BUSINESS LANDSCAPE What Can Developing Countries Do? For the first time, this report 1. What is the global footprint of today’s digital businesses? provides novel evidence of the 2. What is the biggest difference among digital businesses characteristics of 200,000 digital between developed and developing countries? businesses in 190 countries as of 3. What is special about digital business models, and the 2020 to answer three questions: implied market structure, and dynamics in developing countries? Six Key Findings “S p i k i n e s s ” The global digital business Dig·i·tal bus·i·ness [ˈdijidl ˈbɪz.nɪs] n. Digital solution providers that develop and manufacture digital technology 1 landscape is uneven favoring large markets, but a variety of digital growth pathways exists, even products or digital services; a subset of these use platform- based and/or data-intensive business models. for smaller developing countries. see map next page Digital businesses in these subsectors attracted most investments (in USD m) C o n v e rg e n c e between developed and developing cou- Developing countries 2 ntries in the business-to-consumer E-commerce* 140,791 (B2C) segment but developing Fintech* 136,675 countries still have large gaps in Mobility tech* 108,693 the more productive business-to- Travel tech 91,739 Food tech 55,882 business (B2B) segment. Tech hardware 55,079 Telecom* 53,020 Entertainment tech 51,534 Logistics tech 50,617 Business management tech 50,381 E c o n o m ie s o f s c o p e (not just scale) via expansions to multiple product Developed countries markets is the key to success – but it also Fintech* 273,605 raises concerns related to market Health tech 230,451 3 concentration. E-commerce* Business management tech 198,854 170,674 * Big data and analytics 60 percent of digital firms provide Security tech 167,135 130,374 solutions in more than one product Mobility tech * 124,717 Software and SaaS market (e.g., e-commerce, food tech, Marketing tech 112,572 107,455 and logistics tech by one firm) Telecom* 107,138 Source: FCI Digital Business Database. The numbers refer total funding data from 1970-2020 (in USD million). *Denotes that this setoctor is common across developed and developing countries. Note: Digital businesses can offer digital solutions in multiple sectors. 16 percent of digital businesses in developed countries and 18 percent in developing countries adopted p la tform - b a s ed or d ata-d riven b us ines s m od els. Digital Business Density Across Countries Top 20 High-Performing Countries by Digital Business Density After Controlling for Population and GDP 1. Estonia 11. Pakistan 2. Kenya 12. Singapore 3. India 13. Sweden 4. Israel 14. Bulgaria Digital Business Gap De ned As 5. United 15. Cayman ((actual-potential)/potential)*100% Kingdom Islands High-Performers Q5 Q4 6. United States 16. Canada Q3 7. Iceland 17. Lebanon Q2 8. Armenia 18. Vietnam Q1 No Data/Excluded 9. Cambodia 19. Myanmar Q1 10. Finland 20. South Africa Q2 Q3 Q4 Low-Performers Q5 Source: Authors’ calculations, using the FCI Digital Business Database. IBRD 46700 | OCTOBER 2022 I n v e s t o rs Entities from the US and China are the most prominent investors in digital businesses. While they are mostly focused on their 4 home markets, US investors are also prolific in other developing country markets, along with other large developing country investors (see pie chart on the right). A c q u is it io n m a rk e t s The report presents early evidence of a U-shaped pattern in the concentration of acquirers in the highest- 5 valued acquisitions as digital markets develop with more digital firms ** Other: investors from Hong Kong, Japan, and investment flows, pointing to potentially different sets of policy Malaysia, Spain, Switzerland, and the United actions for countries at different stages of digital development. Kingdom, as well as the International Finance Corporation (IFC) of the WBG. G e n d e r a n d s u s t a in a b ilit y 18 p ercent of digital firms1 have women in The digital sector remains an important management positions. While this is still 6 venue for gender inclusion and insufficient, it is more than twice the corresponding sustainability technology, with early signs share in the traditional economy (8 percent). of success in certain applications. About the Report and the Digital Business Database The report “A Spiky Digital Business Landscape” is based on the newly assembled Digital Business Database of the World Bank’s Finance, Competitiveness, and Innovation (FCI) Global Practice. The database is a firm-level database of 200,000 digital businesses in 190 countries. It was created using three different proprietary data sources – CB Insights, Pitchbook, and Briter Bridges – that use various techniques, from web-scraping to gathering firm information from entrepreneurship networks, venture capital (VC), and other investment deals. They specialize in collecting information on tech start-ups or digitalized firms that would be attractive for VC/private equity (PE) investors due to certain innovative elements in their business models, or core product offerings. Please find further details in the report “A Spiky Digital Business Landscape: What Can Developing Countries Do”, available for download at https://openknowledge.worldbank.org/. © 2022 International Bank for Reconstruction 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 conclusions expressed in this work do not neces- sarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. 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 the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encour- ages 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, including 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. TABLE OF CONTENTS Abbreviations........................................................................................... ix Acknowledgements. . ............................................................................... xii Executive Summary. . ................................................................................. 1 Chapter 1  Motivation: Demand for Evidence on the Global Digital Business Landscape.................................................................... 15 Chapter 2  Definitions, Data Sources, Information Included in the Database, and Limitations.............................................. 21 Chapter 3  Digital Business Landscape Developed Vs. Developing Countries.............................................................................. 31 Chapter 4  Convergence & Digital Growth Pathway.. .............................. 43 Chapter 5  Big Techs & the Emergence of “Digital Conglomerates”....... 57 Chapter 6  Inclusion Dimensions: Gender and Cleantech....................... 71 Chapter 7 Conclusions and Next Steps................................................... 75 Appendix. ................................................................................................ 81 Appendix A: Methodologies Used. . ........................................................ 82 Appendix B: Definitions of Digital Subsectors ....................................... 86 Appendix C: Countries in Income and Regional Groups in the Database. . ....................................................................................... 90 Appendix D: Preliminary Results: Data Regulations and Digital Business Relationship.. ............................................................... 92 Appendix E: Other Figures and Tables................................................... 95 Glossary. . .............................................................................................. 104 Bibliography.......................................................................................... 107 A Spiky Digital Business Landscape  iv LIST OF FIGURES, TABLES AND BOXES Figures Figure E 1  Flow of the Report and Key Findings.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Figure E 2  Digital Business Density Across Countries.. . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Figure E 3  Top 10 Most-Funded Subsectors and Their Main Tech Solution Types. . ............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure E 4  Distribution of Firms According to Presence in Number of Subsectors/Markets. . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure E 5  Nationality of Top 20 Investors in Digital Firms in Developed and Developing Countries.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure E 6  U-Shaped Hypothesis: Digital Business Development and Acquisition Market Concentration.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure E 7  Focus Areas of the 50 Largest Clean Tech Firms (by total funding)............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 Figure E 8  Towards A New Policy Agenda for Digital Business Development: The Need to Balance "Efficiency-Enhancing Scope and Conglomeration” vs. the Risks of “Anticompetitive Consolidation”...... 13 Figure 1.1  Market Capitalization of US ‘Big Tech’ Companies 1990-2019. . ...................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 Figure 1.2  Price-to-Earnings Ratio in Emerging Markets, and Top 10 Constituents (out of 1,418) by Market Capitalization in MSCI Emerging Markets Asia Index.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 Figure 2.1  Scope of Digital Business Used in This Report.. . . . . . . . . . . . . . . . . . . . 2 2 Figure 2.2  Creation of the FCI Database from Three Data Sources.. . . . . . . 24 Figure 2.3  Organization of the Digital Business Database.. . . . . . . . . . . . . . . . . . . 26 Figure 2.4  Staff Size Distribution, Developed vs. Developing Country Firms................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 8 Figure 2.5  Capital Intensity: Median Total Funding Raised by Digital Businesses...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Figure 3.1  Digital Business Density Across Countries.. . . . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 3.2  Top 10 Subsectors by Total Ticket Size of Exits – Developed vs Developing Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 6 Figure 3.3  Clusters Among Digital Businesses Founded After 2015 in Developed Countries.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 v  List of Figures, Tables and Boxes Figure 3.4  Distribution of Firms According to Presence in Number of Subsectors/Markets. . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 0 Figure 4.1  Founding Year of Digital Firms Receiving Funding in Applied and Deep Tech................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 6 Figure 4.2  Distribution of Exit Strategies, Developed vs. Developing Country Firms. . ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 9 Figure 4.3  Regional Digital Business Hubs and the Destination of Regional Digital Businesses ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 0 Figure 4.4  Region-Wise Shares of Platform and Data- Driven Businesses........................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Figure 5.1  Composition of the Top 20 Investors, Developed vs. Developing Countries.. .................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Figure 5.2  Nationality of the Top 20 Investors in Developed and Developing Countries . . .................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1 Figure 5.3  Distribution of Firm Sizes Among Acquiring and Target Firms in M&As Involving Digital Platforms .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4 Figure 5.4  Distribution of Antitrust Cases by Location of Headquarters............................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4 Figure 5.5  U-Shaped Hypothesis: Digital Business Development and Market Concentration. . ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Figure 5.6  Policy Matrix for Digital Business Development: The Need to Balance “Efficiency-Enhancing Scope and Conglomeration” vs. the Risks of “Anticompetitive Consolidation”.. . . . . . . . . 6 8 Figure 6.1  Share of Businesses with Women in Leadership Roles in the Sub-Saharan Africa (SSA) and Middle East and North Africa (MENA) Regions (Both Developed and Developing Countries). . . . . . . . . . . . . . 73 Figure 6.2  Operating Countries and the Focus Areas of the 50 Largest Clean Tech Firms.............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure A.1  Example of Harmonization of Data Source Industry Classification: Big Data................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3 Figure A.2  Number of Firm-Level Data from Three Data Sources. . . . . . . . . . 8 5 Figure A.3  Platform-based and Data-Driven Businesses.. . . . . . . . . . . . . . . . . . . . . 8 5 Figure A.4  Employee Size Distribution by Region.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5 Figure A.5  Top 10 Subsectors by Total Number of Exits – Developed vs Developing Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 8 Figure A.6  Top 10 Subsectors by Total Ticket Size of Exits – Developed vs Developing Countries after excluding the US, China, and India.. .......................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 8 A Spiky Digital Business Landscape  vi Figure A.7  Digital Clusters Among Developed Country Businesses Founded Before 2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Figure A.8  Digital Clusters Among Developing Country Businesses Founded Before 2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Figure A.9  Digital Clusters Among Developing Country Businesses Founded After 2015.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 0 Figure A.10  Distribution of Firms according to Presence in Number of Subsectors/Markets after excluding the US, China, and India.. ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 0 Figure A.11  Digital Firms Receiving Funding in Applied and Deep Tech by Founding Year after excluding the US, China, and India.. ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 01 Figure A.12  U-Shape Hypothesis: Acquisition Market Concentration by Countries/Income Group.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 01 Figure A.13  Top Subsectors with Women in Management Team, Entire Sub-Saharan Africa.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 02 Figure A.14  Top Subsectors with Women in Management Team, entire MENA.. ................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 02 Figure A.15  Number of Clean Tech Companies/GDP per Capita.. . . . . . . . 1 03 Tables Table 2.1  Sample Size by Income Group.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 7 Table 2.2  Sample Size by Region.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 8 Table 3.1  Top 10 Digital Subsectors by Total Funding (Developed vs. Developing). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Table 3.2  Summary: Differences in Characteristics, Developed vs. Developing Country Firms.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 6 Table 3.3  Top Funded Health Tech applications in Developed vs. Developing Countries.. ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Table 4.1  Pair-Wise Correlation Coefficients Between Population and GDP Per Capita, and the Intensity and Extensity of Platform and Data-Driven Businesses.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Table 5.1  Top 20 Investors and Number of Investees, Developed vs. Developing Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Table 5.2  Top 20 Acquirers and the Number of Digital Businesses They Have Acquired .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Table 5.3  Sectoral Distribution of Acquisitions by the Top 3 Acquirers (by Number of Acquisitions).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 vii  List of Figures, Tables and Boxes Table A.1  Ordinary Least Squares (OLS) Estimates of Principal Components of Data Governance Regulations, Interacted with Log(GDP/capita) on Digital Intensity, Longevity of Digital Businesses and Exit Rates. . ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3 Table A.2  Top 20 Country by Number of Firms, Total Funding, Investment Exit. . ............................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Table A.3  Top 10 Digital Subsectors by Number of Businesses (Developed vs. Developing). . .......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Table A.4  Top 10 Digital Subsectors by Total Funding after Excluding US, China, and India...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Boxes Box 2.1  Demand Side: The World Bank’s Technology Adoption Survey........................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Box 3.1  Estonia – a ‘Digital Republic’.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 4 Box 3.2  The Fintech, Insurance Tech, Blockchain and Cryptocurrency, Gig Economy, Property Tech, Construction Tech Mega Cluster........................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Box 4.1  Encouraging Trends in Funding for African Start-Ups. . . . . . . . . . . . . . . . 4 5 Box 4.2  KhmerOS – Localization for Inclusion and Value Creation. . . . . . . . . 4 8 Box 4.3  B2B vs B2C E-Commerce.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3 Box 4.4  Asia’s Digital Strategy Leveraging Consumers’ “Mobile-First” Approach and “Super Apps”.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 Box 5.1  Abuse of Dominance in the Digital Finance and Telecom Sector - Examples from Africa.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5 Box 5.2  Examples of Pro-Competition Actions and Principles .. . . . . . . . . . . . . 69 A Spiky Digital Business Landscape  viii LIST OF ABBREVIATIONS AI Artificial intelligence API Application programming interface B2B Business-to-business B2C Business-to-consumer BAT Baidu, Alibaba, and Tencent BRIC Brazil, Russia, India, and China BRICS Brazil, Russia, India, China, and South Africa C2C Consumer-to-consumer CIIP Competitive Industries and Innovation Program C-JET Competitiveness for Jobs & Economic Transformation CRM Customer relations management CVC Corporate venture capital DBI Digital business indicators DECAT Development Economics Vice Presidency Data Analytics and Tools Unit DFI Development finance institution EAP East Asia and Pacific ECA Europe and Central Asia EFI-TIC Equitable Growth, Finance & Institutions Vice Presidency of the World Bank, Trade, Investment, and Competitiveness Department ERP Enterprise resource planning EU European Union FAT Firm-level adoption of technology FCI Finance, Competitiveness and Innovation Global Practice of the World Bank FDI Foreign direct investment ix  List of Abbreviations FOB Free on board GBF General business functions GPT General purpose technology HHI Herfindahl-Hirschman Index ICT Information and communication technologies IFC International Finance Corporation IMF International Monetary Fund IoT Internet of things IPO Initial public offering IPR Intellectual property rights LAC Latin America & the Caribbean M&A Mergers and acquisitions MENA Middle East and North Africa MFN Most favored nation ML Machine Learning MNE Multinational enterprise MSCI Morgan Stanley Capital International MSME Micro, small and medium sized enterprise NACE Statistical Classification of Economic Activities in the European Com- munity OECD Organisation for Economic Co-operation and Development OLS Ordinary least squares P2P Peer-to-peer PC Principal component PCA Principal component analysis PE Private equity SA South Asia SaaS Software as a service SAR Special administrative region SDGs Sustainable development goals A Spiky Digital Business Landscape  x SEO Search engine optimization SME Small and medium sized enterprise SSA Sub-Saharan Africa SSBF Sector-specific business functions VC Venture capital VPU Vice Presidency Unit WBG World Bank Group Assumptions made throughout the report Rational investors: In this report, it is assumed that investors make rational deci- sions to maximize returns from their investment in tech companies (for example, they do not primarily engage in Ponzi schemes, or hostile M&As). Therefore, the deal sizes and flows used in the analyses largely reflected digital business perfor- mance and potentials. Market readiness: Investment deals (both counts and volume) in digital technolo- gy firms and their exit rates (e.g., IPOs, M&As, and buyouts) are used as a proxy to signal whether a digital technology and/or a sector (e.g., fintech, SaaS) are “market ready” in terms of having a viable product and business model, and whether there is a potential market for it. Accordingly, deal flows are assumed to reflect the supply and latent demand of digital technologies.  Market sizing: The bigger the size of the deal and the higher the exit rate, the more likely digital tech will “disrupt” the traditional sectors it serves and achieve scale. (For example, e-commerce vs. brick-and-mortar retail, fintech vs. traditional banks.) Maturity: Exit rates (for example, the percentage of firms in IPOs, M&A and buy- outs) are used as a proxy to signal that a digital technology/sector has matured in terms of improving the product and testing the business model. SME feasibility: Looking at market-ready technologies also has implications for SME tech adoption. The assumption is that only digital technologies with a reasonable “cost-benefit” ratio are likely to be adopted by SMEs. So the first-order priority is to identify those digital technologies that are most likely to be adopted by budget-constrained SMEs and are ready for mass adoption (although other constraints, such as capability failures, also play an important role). xi  List of Abbreviations ACKNOWLEDGEMENTS This report was prepared by a World Bank team led by Tingting Juni Zhu (Senior Economist). It includes Philip Grinsted (Private Sector Specialist), Hangyul Song (Data Analyst), and Malathi Velamuri (Senior Economist), working under the guidance of Martha Martinez Licetti (Practice Manager, Markets, Competition and Technology Global Unit), Mona Haddad (Director, EFI-TIC) and the following advisors: Vivien Foster (Chief Economist, Infrastructure VPU), Mary C. Hallward- Driemeier (Senior Economic Advisor, EFI VPU) at the World Bank Group; and Dr. Avi Goldfarb at the University of Toronto and Dr. Erik Van Der Marel at the European Centre for International Political Economy. The peer reviewers are Andrea Barone (Senior Economist), Smita Kuriakose (Senior Economist), and Gaurav Nayyar (Lead Economist). We are also grateful for valuable comments and input from Tania Priscilla Begazo Gomez, Rong Chen, Marcio Cruz, Seidu Dauda, Elena Gasol Ramos, Sara Nyman, Jan Orlowski, Francis Ralambotsiferana Ratsimbazafy and Dennis Sanchez Navarro. Research support has benefitted from World Bank DEC Data Analytics and Tools Unit: Trevor Monroe (Senior Operations Officer) and Aivin Vicquierra Solatorio (Data Scientist). The team also wishes to thank the Competitive Industries and Innovation Program (CIIP) Trust Fund and the Competitiveness for Jobs & Economic Transformation (C-JET) Trust Fund for providing the financial support that made this report possible. A Spiky Digital Business Landscape  xii EXECUTIVE SUMMARY 1  Executive Summary Digital technologies hold the promise of bridging wealth gaps through innovation-driven growth, but the “winners-take-most” dynamic of digital business models calls into question the net growth effect and the global footprint of this sector. Digital transforma- tion is driven by a set of digital technologies that have led to a rapid and steep decline in the costs of data storage, computation, and transmission. These technologies hold promise for bridging the wealth gap between nations by allowing developing countries to catch up with generations of previous technologies. At the same time, characteristics inherent to these technologies have the potential to result in a “winner-takes-most” dynamic, by creating market entry barriers and leading to high levels of concentration and potential market dominance (Crémer, de Montjoye and Schweitzer 2019). The fact that the digital economy holds both the promise of rapid growth and the threat of increased concentration (with the potential exercise of market power) implies that gaining an understanding of the digital business footprint across the world is essential. Unfortunately, comprehensive data on this rapidly growing sector is not easily available, making it challenging to assess how these opposing forces are playing out, especially in developing countries. For the first time, this report provides novel evidence of the characteristics of digital business and markets in 190 countries. The report defines digital businesses as digital solution providers that develop and manufacture digital technology products or digital services; a subset of these can also use platform-based and/or data-intensive network effect business models. The report draws on the World Bank’s newly assembled firm-level database of 200,000 digital businesses in 190 countries, to provide unique evidence on the current global digital business landscape. Our analysis is based on data collected from three proprietary data sources - CB Insights, Pitchbook, and Briter Bridges - covering companies founded in 1970-2020, and has been cross-checked with national statistics whenever possible.1 In this report, we present some stylized facts regarding the global digital business landscape across developed and developing countries , as well as insights into the various types of digital technologies based on the segmentation of users –for example, business-to-business (B2B) vs. business-to-consumer (B2C), and deep tech vs. applied tech solutions. FIGURE E 1  Flow of the Report and Key Findings This database is largely a supply-side assessment of digital business ecosystems including firms at all stages of the life cycle. 1  It complements other World Bank initiatives that look at the demand side – for example, the Firm Technology Adoption Survey, or firms at early-stage digital entrepreneurship. A Spiky Digital Business Landscape  2 THREE KEY QUESTIONS & NEW INSIGHTS FROM THIS REPORT SUMMARY OF THE REPORT IN ONE PAGE This report, for the first time, pro- vides novel evidence on the char- acteristics of digital businesses and markets in 190 countries. It uses the World Bank’s newly assembled firm-level database of 200,000 digital solution firms founded between 1970-2020 and sheds light on three questions raised by developing country governments. 3  Executive Summary 1. What is the global footprint of today’s digital businesses, given two opposing forces? (The proliferation of digital services versus market concentration due to “winner-take-most” dynamics in digital business models.) The digital business landscape is “spiky,” but growth pathways do exist. 1. Countries with large populations and market size dominate, though there are some strong outliers among smaller developing countries, implying that there are a variety of pathways to digital growth if a country is willing to adopt digital market policies and strategies that proactively nurture the digital sector. 2. Geographic expansion patterns of digital businesses imply that digital single- market policies are particularly important in order for firms in smaller countries to achieve scale by accessing larger regional markets and data networks. To ensure equity in digital single markets, however, and to safeguard policies where ecosystems are still undeveloped, targeted support programs are also needed. B2B 2. What is the biggest di erence among digital businesses between developed and developing countries? Limited B2B Digital Solutions 1. While digital business entry, maturity, and funding in consumer- facing applied technology (for example, e-commerce and fintech) are catching up in developing countries, B2B solutions are still limited, likely due to a combination of supply and demand factors. However, globally, B2B markets are 2-8 times larger than B2C markets. 2. There is an untapped opportunity for a technology catch- up in developing countries. Local digital businesses can develop applications that are tailored to local needs, taking into account the endowed skills, cultural preferences, and local infrastructure. 3. What is special about digital business models, and the implied market structure, and dynamics in developing countries? Scope (Not Just Scale) Economies 1. Over 60 percent of database firms provide solutions in more than one product market, even at the early stages of the firm’s development cycle (e-commerce, agtech, food tech, logistics tech, all o ered by one firm), exhibiting a multiproduct market structure. 2. This feature is facilitated by the transfer of intangible assets (data, capabilities), and the adoption of standards across interfaces and systems, leading to the creation of “product ecosystems”; that is, lines of products and services that are linked through their shared functionalities and data. 3. On the whole, digital businesses in developing countries can no longer count on one product market to scale (especially when the market size is small); they need to develop complementary products and services early on that will be able to achieve scope economies and lock in users. However, this can lead to the emergence of “digital conglomerates,” which has competition policy implications such as how to define a market, and how to treat product bundling and negative pricing. 4. Large digital conglomerates (“big techs”) are particularly active as acquirers of smaller digital businesses in developing countries due to limited investment exit pathways; this calls for regulators to monitor M&A deal flows and revisit merger notification thresholds. KEY FINDING 1 The global digital business landscape is uneven, dominated by a few countries with large market sizes. At the same time, some smaller developing economies are among the global leaders in digital business density, performing better than would be predicted based on their GDP and population; this suggests that there are a variety of feasible digital growth pathways. FIGURE E 2  Digital Business Density Across Countries Digital Business Gap De ned As ((actual-potential)/potential)*100% High-Performers Q5 Q4 Q3 Q2 Q1 No Data/Excluded Q1 Q2 Q3 Q4 Low-Performers Q5 IBRD 46700 | OCTOBER 2022 Top 20 High-Performing Countries by Digital Business Density After Controlling for Population and GDP 1. Estonia 6. United States 11. Pakistan 16. Canada 2. Kenya 7. Iceland 12, Singapore 17. Lebanon 3. India 8. Armenia 13. Sweden 18. Vietnam 4. Israel 9. Cambodia 14. Bulgaria 19. Myanmar 5. United Kingdom 10. Finland 15. Cayman Islands 20. South Africa Source: Authors’ calculations, using the FCI Digital Business Database. Note: a robustness check was conducted by substituting the number of digital businesses with a) total investments received (to proxy for firm size), and b) number of firms reaching the IPO/M&A stage (to proxy for quality). The results are reported in Appendix E. This report finds a “spiky” digital business landscape that is dominated by a few large firms in a few developed and emerging economies; but it also provides evidence that some small developing economies are strong outliers in digital business density, after controlling for population size and GDP. The number of digital businesses in a country is highly correlated with population and market size (proxied by GDP). The US, the UK, China, and India are thus the domi- nant players; but many smaller emerging markets like Estonia, Kenya, and Armenia also emerge among the top-ranked countries after controlling for population and market size. These strong “digital outliers” point to the existence of a variety of pathways for attaining high levels of digital business density. 5  Executive Summary KEY FINDING 2 There is evidence of convergence in the B2C segment between developed and developing countries but there is a limited supply of B2B solutions, pointing to an opportunity for a technology catch-up in developing countries. Over time, developing countries have been catching up in applied and consumer-facing digital applications, but there is evidence of a relatively limited supply and market size for B2B digital solutions in developing countries. This is also the difference that is starkest between developed and developing countries. The distribution of the top digital subsectors by number of digital businesses and by total investments received have overlaps between developed and developing countries; these include many consumer-facing applied technologies such as fintech, e-commerce, and mobility tech. 2 However, there appears to be a relatively limited supply and demand for more complex B2B digital solutions in developing countries. Moreover, even within the same B2B sector, firms in developed countries are generally active in the use of more sophisticated applications relative to their developing country counterparts. This pattern has implications for policy actions to stimulate B2B digital applications in developing countries - for example by encouraging local digital firms to develop B2B solutions that are tailored to local needs and that fit with local user skills, language, and infrastructure endowments, in order to increase uptake. Given the productivity-enhancing potential of B2B solutions for offline firms, this market segment should not be ignored, and should not remain underdeveloped. FIGURE E 3  Top 10 Most-Funded Subsectors and Their Main Tech Solution Types DEVELOPED COUNTRIES DEVELOPING COUNTRIES Fintech B2B B2C E-commerce B2B B2C Health tech B2B B2C Fintech B2B B2C E-commerce B2B B2C Mobility tech B2C Business management tech B2B Travel tech B2C Big data and analytics B2B Food tech B2C Security tech B2B Tech hardware B2B B2C Mobility tech B2C Telecom B2B B2C Software and SaaS B2B Entertainment tech B2C Marketing tech B2B Logistics tech B2B Telecom B2B B2C Business management tech B2B Source: Authors’ calculations using the FCI Digital Business Database. See Appendix B for definitions of the 44 digital business subsectors used throughout this report. 2  A Spiky Digital Business Landscape  6 KEY FINDING 3 Digital firms across the world are active in multiple product markets, highlighting the distinct feature of digital business models in facilitating the creation of product ecosystems due to economies of scope in product development. Firms in developing countries are using platform-based or data-driven business models to realize scope economies and create value, especially in e-commerce and fintech. 60 percent of digital firms provide solutions in more than one product market, highlighting the importance of scope economies in digital business models in both developed and developing countries. Over 25 percent of the firms in the database are active in three or more product markets in both developed and devel- oping countries (Figure E 4). For example, many e-commerce firms also provide logistics tech and agtech or food tech services. This pattern is a unique feature of digital businesses: they are not implementing scale-up strategies by focusing on increasing users in one product market alone, but are diversifying into complemen- tary products and services to build network effects and increase efficiency. FIGURE E 4  Distribution of Firms According to Presence in Number of Subsectors/Markets 50% % of Digital Businesses 40% 40% 40% 34% 33% 30% 20% 16% 17% 6% 7% 10% 4% 4% 0% 1 Subsector 2 Subsectors 3 Subsectors 4 Subsectors 5 Subsectors or more Developed (n=140K) Developing (n=38K) Source: Authors’ calculations using the FCI Digital Business Database. Scope economies also reveal unique digital subsector clustering patterns. This report finds that certain digital services tend to cluster: for example, artificial intelligence (AI) with big data analytics, business management tech, software and SaaS, and web services; or clean tech with mining tech and utilities tech (Figure 3.3). This clustering helps to explain the similarity of firm presence in multiple subsectors between developed and developing countries (Figure E 4). More specifically, this pattern of digital subsectors can be explained by a number of the features of digital businesses: (1) within-firm digital spillovers arising from the transfer of intangible assets and adjacent skills (data, ideas, capabilities, etc.) in multisided business models3, which allows firms to capture scope economies, 3  A multisided business model brings together distinct user groups whose presence and actions create, deliver, and capture value for each other (e.g., economies of scale). For example, an online marketplace brings together buyers, sellers, and advertisers – whose revenues depend on each other’s. 7  Executive Summary increase efficiency, and gain a competitive edge relatively quickly by leveraging intangible assets; (2) strategic investments by firms across sectors, as a means of “locking” consumers into the firms’ core product or service by bundling it with other services and products, enabling market “tipping.” (That is, allowing firms to nudge the market towards their own platforms/ecosystems); (3) The diffusion of certain general- purpose technologies in recent years (cloud services, AI) that can power several adjacent digital subsectors/solutions at the same time. This report finds that in developing countries, platform-based and data-driven business models are important conduits for scope economies and value creation, especially in e-commerce and fintech. This report estimates that 16 percent of digital businesses in developed countries and 18 percent in developing countries are adopting platform-based or data-driven business models. Plat- form-based firms are defined as those facilitating interactions for a large number of participants. A platform business does not own the means of production as its core services, but rather creates and facilitates the means of connection. The role of the platform business is to provide a governance structure and a set of standards and protocols that facilitates interactions at scale so that network effects can be unleashed (Deloitte 2020, Still 2017, Evans 2013). Data-driven firms are those that systematically and methodically collect or aggregate large data sets and that use advanced analytics (such as AI, big data, and blockchain) to create value, leveraging data as a key element of their business model (Hartmann, et al. 2014). E-commerce and fintech are the top sectors to leverage these platform-based or data-driven business models. KEY FINDING 4 While entities from the US and China are the most prominent investors in digital businesses, they are still mainly focused on their domestic markets. For most developing countries, big tech firms from other large emerging markets are active in the highest-valued acquisitions. US and Chinese investors are the most prominent investors in digital busi- nesses, predominantly investing in their respective domestic firms. The US is the predominant source of investment in digital businesses across the world, while Chinese investors are prominent in developing countries, as Figure E 5 reveals. These investments cover all stages of the risk capital investment cycle – seed, venture capital (VC), private equity (PE), M&A, IPOs, and so on. However, it should be noted that US investors still predominantly invest in US firms (88 percent of all their investment deals), while Chinese investors invest largely in Chinese firms (97 percent of all investment deals), even though their footprint is more global than that of investors from other countries. A Spiky Digital Business Landscape  8 FIGURE E 5  Nationality of Top 20 Investors in Digital Firms in Developed and Developing Countries Nationality of top 20 Nationality of top 20 Nationality of top 20 investors investing into investors investing into out-of-country digital businesses in digital businesses in investors investing into Developed Countries Developing Countries digital businesses in Developing Countries EU Spain 10% 5% UK 5% Other** 32% USA China * 40% 55% USA 68% USA 85% Source: Authors’ calculations using the FCI Digital Business Database. Note: *Chinese investors mainly invested in Chinese firms, therefore Chart 3 ranks the top 20 out-of-country investors. ** “Other” here includes investors from Hong Kong, Japan, Malaysia, Spain, Switzerland, and the United Kingdom as well as the International Finance Corporation (IFC) of the World Bank Group (WBG), each country representing a similar percentage share. The concentration of a few acquirers in the highest-valued acquisition deals in developing countries is driven predominantly by big tech firms from large emerging markets, and not by firms from Silicon Valley only. The success of several US tech companies – Microsoft, Google, and Amazon, to name a few - in expanding their global footprint is undisputed, as is that of several Chinese tech giants like the Alibaba group and Tencent. This report presents additional evidence that the concentration of acquirers in mega acquisition deals in developing countries is driven predominantly by big tech firms from large emerging markets like China, India, South Africa, Indonesia, and Brazil, and less so from Silicon Val- ley-based big techs.4 The predominant share of these acquisition deals is unlikely to cross the traditional antitrust review thresholds, posing additional challenges for developing country policy makers in monitoring potentially distorting behaviors. Nationality of top 20 acquirers in developing country digital firms: China (21 percent), India (21 percent), Brazil (11 4  percent), South Africa (10 percent), followed by Argentina, France, Indonesia, Mauritius, Russia, UK, and US, all with similar weights. 9  Executive Summary KEY FINDING 5 The report presents early evidence of a U-shaped pattern in the concentration of top acquirers in the highest-valued acquisitions as digital markets develop, pointing to potentially different sets of policy actions for countries at different stages of digital development. There is early evidence of a U-shaped pattern in the concentration of top acquirers in M&A deals as digital markets grow, with implications for policy at both ends of the distribution. Mergers and acquisitions (M&A) are the dominant exit strategy for digital businesses across the globe. However, this is also where potential market distortions happen. Figure E 6 plots a proxy for concentration in the M&A market for different regions based on the top 50 acquirers in each region, ranked by the dollar volume of total M&A deals they were involved in during 2000-2020. The proxy measure for concentration is the Herfindahl–Hirschman Index (HHI),5 calculated in terms of each acquirer’s share out of the total value of acquisitions (in dollar terms) of all 50 top acquirers. The HHI is plotted against the number of digital businesses in each region during that period, which is a measure of development of the digital market. Figure E 6 shows how influential a few acquirers are in markets that are in the early stages of digital development (for example in Latin America and the Caribbean (LAC), and in Sub-Saharan Africa (SSA). This is partly a consequence of the fact that only a few acquirers are bidding on the largest digital businesses in these markets; and partly a consequence of the local corporate and acquisition market structures. As the digital business sector develops, M&A activities increase with the involvement of a broader set of market players, reducing concentration (for example, in East Asia and the Pacific (EAP) developing countries). However, in some digitally mature markets, like North Amer- ica and Singapore, acquirer concentration has been increasing, especially since 2015, as some big techs are looking to consolidate their market leadership position via vertical and horizontal acquisitions.6 While more data - especially country-level panel data - is needed to further test this U-shaped hypothesis, it is consistent with the literature about entry for buyout, killer acquisitions, and aqui-hires (Rasmusen 1988, Cunningham, Ederer and Ma 2020, Mermelstein, et al. 2020). This U-shaped pattern suggests that digital market monitoring and disciplinary actions may vary according to the stage of development, and thus require considering how the benefits of consolidation – economies of scope that can drive efficiency and pro- ductivity increases – compare with the potential costs of increased concentration and lower economic welfare. The Herfindahl-Hirschman Index (HHI) is calculated as the sum of the squared shares of the M&A deal sizes of 5  the top 50 acquirers: For each acquirer in the top 50 (by total deal size), the M&A volume is divided by the total M&A volume of the top 50 acquirers, multiplied by 100, squared, and added together. The HHI accordingly ranges from >0 (low concentration) to 10,000 (high concentration). The relationship depicted is also robust to the inclusion of the top 100 acquirers, indicating that market 6  concentration is driven by a small number of mega deals. Singling out China and India doesn’t affect the HHI of all developing countries very much (<100 in HHI value difference), meaning that the rest of the developing countries - if they have M&A deal flows - would follow a similar U-shaped pattern. A Spiky Digital Business Landscape  10 FIGURE E 6  U-Shaped Hypothesis: Digital Business Development and Acquisition Market Concentration (HHI of top 50 acquirer, by M&A deal size) 1500 Digital acquisitions market concentration 1000 500 0 20000 40000 60000 80000 100000 Digital Business Development (Number of digital business) Source: Authors’ calculations using the FCI Digital Business Database. Note: The Sub-Saharan Africa (SSA), Middle East and North Africa (MENA), and Europe and Central Asia (ECA) regions are not shown separately due to the low number of acquirers with deal information in these countries. Also, using the number of M&A deals (as opposed to the number of digital businesses) to measure digital business maturity and development produces a similar U-shaped curve. n in parentheses is the number of M&A deals. KEY FINDING 6 The digital sector remains an important venue for gender inclusion and climate change action, with early signs of success using certain applications and in certain regions. There is suggestive evidence that digital businesses are becoming more gen- der-inclusive and are contributing to the environmental sustainability agenda in recent years. In the Sub-Saharan Africa (SSA) and Middle East and North Africa (MENA) regions, where there is gender-disaggregated data on digital businesses, this report finds that the share of digital firms that have women in management positions is over two times the corresponding share in the traditional economy (18 percent versus 7.5 percent). This finding is consistent with the literature (OECD 2018, Aly 2020) and highlights the digital sector as a key venue for advancing gender inclusion. We also found evidence that the clean tech firms that have received the most risk capital in recent years are engaged in renewable energy solutions and other circular economy sectors such as sustainable food production and transport, recycling, energy storage, and green packaging, suggesting that some clean tech solutions are becoming commercially viable. The intersection of climate change and digital development will likely require more research, including on the impacts of market-ready clean tech solutions, as countries scale up their deployment. 11  Executive Summary FIGURE E 7  Focus Areas of the 50 Largest Clean Tech Firms (by total funding) Alternative fuels/ energy Sustainable food and agriculture (e.g., animal feed and crops) Green building/ energy efficiency Energy storage Consulting and advisory (e.g., energy supply management, asset investments) Recycling (e.g., management of excess and returned inventory) Sustainable transport (e.g., zero-emission vehicles) Other (e.g., sustainable medical technology, air pollution measurement, environmental data collection, emissions trading, green packaging) Source: Authors’ summary using the FCI Digital Business Database. Given the global trends identified above, this report contributes to five main policy debates in nurturing digital economy growth and identified areas for further research. First, since the global landscape of digital business is “spiky” with a few regional digi- tal hubs/large countries now dominating, the importance of digital trade and digital single market policies will likely increase, particularly for helping smaller developing economies achieve scale. To give countries a strong incentive to participate, such arrangements will require cross-country harmonized policy frameworks for digital infrastructure and data, digital taxation, digital financial services, competition policies and online consumer and supplier protection. The operation of digital businesses across borders would also need changes in complementary factors such as customs clearance (for e-commerce), movement of goods (for regional logistics), foreign direct investment (to fund startups, telecom and data infrastructure). Second, given the productivity-enhancing potential of the B2B digital market seg- ment, a lot is at stake for this market segment to be ignored and remain under-devel- oped in developing countries. While fostering this segment requires initiatives from both the supply and demand sides, governments have a big role on the supply side, including human capital investments, protecting intellectual property rights, and coordinating co-invention networks. On the demand side, they can also incentivize local digital firms to develop B2B solutions that are tailored to local needs and fit with user skills, language and culture, and infrastructure endowments to increase uptake. Third, given that “scope economies” is a defining characteristic of digital business- es, and that firms can no longer reliably achieve scale and growth by specializing in one product market alone (especially when the digital market size is limited), there is a need for government programs and policies that encourage intangible asset transfer (data, IP rights) across firms, sectors, and borders in order to enable scope economies. This also includes regulations that facilitate the entry of digital business models into the traditional service sectors (e.g., logistics, finance), industry data access, re-use, interoperability policies, and standards adoption to facilitate product ecosystem development. A Spiky Digital Business Landscape  12 Fourth, while digital businesses attract investors of all types – e.g., from traditional venture capital/private equity investors to government financing – big techs play a prominent role in investing and acquiring digital businesses in developing countries compared with developed countries. This suggests that a diverse source of funding and exit options is critical in order to ensure the healthy growth of digital businesses and nurture high-quality exits. Related to this point, the observed digital M&A patterns also call for revisiting the traditional practices that govern how M&As are reviewed for anti-competitive practices in digital markets, including considering lowering the thresh- olds that trigger these merger reviews in order to detect potential killer acquisitions. Finally, there is also a need to consider different sets of policy actions for growing and safeguarding the digital market for countries at different stages of digital devel- opment. Digital business models have a tendency to move toward conglomeration as businesses mature. While regulations should not constrain firms in the early stages of digital business development from achieving scope and scale, since this is required for efficiency, and for the creation of better products and services, there is a need for governments to actively monitor investment and M&A flows, and cross-border data flows in order to achieve market discipline and maintain contestability. FIGURE E 8  Towards A New Policy Agenda for Digital Business Development: The Need to Balance "Efficiency-Enhancing Scope and Conglomeration” vs. the Risks of “Anticompetitive Consolidation” ENCOURAGE ENTRY AND ENSURE A LEVEL PLAYING SCALING FOR EFFICIENCY FIELD FOR CONTESTABILITY ■ Remove regulatory entry barriers ■ Update competition policies for new that prohibit digital business models, types of market dominance and exclusion; especially the entry of B2B services ensuring healthy digital-analog competition (logistics tech, travel tech, clean tech) ■ Accessing to input data via investing in ■ Build scope economies and network trusted and secure data exchange to e ects: Open API/standards adoption, promote innovative use of data by SMEs, co-creation programs, joining digital consider mandating data access if a firm single-market initiatives holds a bottleneck position (gatekeeper) ■ Attract a diverse source of funding and exit options for start-ups ■ Ensure equity and safeguard policies: Online consumer and supplier protection, data protection, intellectual property protection ■ Preventing anticompetitive mergers 13  Executive Summary A Spiky Digital Business Landscape  14 CHAPTER 1  MOTIVATION: DEMAND FOR EVIDENCE ON THE GLOBAL DIGITAL BUSINESS LANDSCAPE 15  Chapter 1  Motivation: Demand for Evidence on the Global Digital Business Landscape The digital revolution is transforming the world, creating immense oppor- tunities for developing countries to narrow the income gap with developed countries­­- but also presenting new challenges. The digital revolution is well underway, characterized by a set of frontier technologies, including AI, robotics, and quantum computing, and driven by a rapid and steep decline in the costs of data storage, computation, and transmission. In this increasingly digitalized world, intangible assets such as data, software, and patents have become key sources of value creation and differentiation. The COVID-19 pandemic has further accelerated this shift toward intangibles through a dramatic increase in telecommuting, online shopping, digital entertainment, and online services, among other areas. The digital economy has opened up many potential growth opportunities through which developing countries can seek to narrow the income gap with developed countries. At the same time, policy makers are grappling with the challenge of transforming the laws and regulations governing trade, taxation, labor, social security, and other spheres, which are increasingly inadequate for a digital world, and where “ecosys- tems” like Apple, Amazon, and Tencent have become the organizational form of doing business (Petit and Teece 2020). The characteristics of the digital economy create a tendency toward conglom- eration, by erecting barriers to market entry and concentrating market power. Data and the ability to create value through data become factors of production, which are important characteristics of digital businesses.7 These assets allow (global) digital firms to achieve scale without mass (Ocampo 2019). That is, they can have a strong presence in a market without being physically established in that market. In digital markets where network effects are sufficiently strong, users are drawn towards the network with the highest number of other users. These features erect barriers to market entry and make certain digital markets, such as digital platforms, prone to market tipping that favors one, or just a few, major ecosystems. This in turn creates the conditions for a “winner-take-most” economy, leading to a concentration of power and wealth accumulation for a small number of global “big tech” firms and individuals (Sturgeon 2021). Figure 1.1 shows the market capitalization of the five biggest US tech firms, and the average capitalization of the 200 largest publicly traded US firms, attesting to the rapid growth of technology firms, especially after 2010. Similar trends are also visible in China and in other emerging markets, as Figure 1.2 reveals.8 Such concentration of market power, if left unchecked, could exacerbate inequality in developed countries, and impede devel- oping countries from using the opportunities of the digital economy to achieve high-income status. Digital products and services typically involve a modular design because they are composed mainly of hardware 7  and/or software components that can be shared across product lines. This feature generates substantial economies of scope in product development, leading firms to expand into multiple product markets and create product ecosystems. Another feature that incentivizes firms to create product ecosystems is consumption synergies to lock in users, allowing consumers to enjoy benefits from the consumption of products or services from the same ecosystem (Bourreau, 2020). Baidu, Alibaba, and Tencent (BAT) are China’s dominant tech giants. Baidu controls China’s search engine 8  market; Alibaba its e-commerce and online advertising; and Tencent its messaging and social media; each of these companies also has online streaming platforms. See https://itif.org/publications/2020/11/23/chinese- competitiveness-international-digital-economy A Spiky Digital Business Landscape  16 FIGURE 1.1  Market Capitalization of US ‘Big Tech’ Companies 1990-2019 1400 1400 $ (billions) $ (billions) 1200 1200 1000 1000 800 800 600 600 400 400 200 200 0 0 1990 1995 2000 2005 2010 2015 2020 2025 Source: Birch, Kean, and D. T. Cochrane. 2021. “Big Tech: Four Emerging Forms of Digital Rentiership.” Science as Culture 1-15. FIGURE 1.2  Price-to-Earnings Ratio in Emerging Markets, and Top 10 Constituents (out of 1,418) by Market Capitalization in MSCI Emerging Markets Asia Index – as of August 31, 2021 80 60 20 2012 2014 2016 2018 2020 FLOAT ADJ MKT CAP INDEX TOP 10 CONSTITUENTS COUNTRY (USD BILLIONS) WT. (%) SECTOR Taiwan Semiconductor Mfg TW 545.77 8.53 Info Tech Tencent Holdings Li (Cn) CN 356.18 5.57 Comm Srvcs Alibaba Grp Hldg (Hk) CN 322.42 5.04 Cons Discr Samsung Electronics Co KR 315.93 4.94 Info Tech Meituan B CN 115.48 1.80 Cons Discr Reliance Industries IN 88.24 1.38 Energy Infosys IN 79.66 1.24 Info Tech China Construction Bk H CN 69.49 1.09 Financials Jd.com ADR CN 67.90 1.06 Cons Discr Housing Dev Finance Corp IN 65.66 1.03 Financials TOTAL 2,026.72 31.67 Source: https://www.firstsentierinvestors.com.au/au/en/adviser/insights/latest-insights/china-tech-con- centration-risk-emerging-markets.html; https://www.msci.com/documents/10199/17e9365e-fbf6-407e-9f48- 808f7b75a5bf. 17  Chapter 1  Motivation: Demand for Evidence on the Global Digital Business Landscape However, there is no available data on digital businesses in developing coun- tries with which to assess the new development trends and the scope of the problems discussed above. There are numerous examples of the impact of digital technology on productivity in developing countries.9 In emerging economies in par- ticular, the development of the digital economy is enabling sectors to catch up with generations of technological development, and move directly to digital solutions such as mobile banking, instead of investing in vast networks of bank branches.10 Yet comprehensive data on this rapidly growing sector, which is upending decades of relatively stable global economic patterns, is not easily accessible. Furthermore, to what extent market concentration is due to conglomeration is also a developing country problem that will require data to substantiate, since most of the big tech companies are located in only a few developed and emerging economies. This report uses a new database that offers unique evidence for analyzing the global digital landscape, comprised of digital businesses of various sizes across the world. Our analysis is based on data collected from multiple sources and covers firms founded in the period of 1970-2000. This report presents some stylized facts regarding the global digital business landscape and its changes over time. Specifically, it presents descriptive evidence regarding (1) the existence of a digital growth pathway in developing countries that has the potential of closing the wealth gap with developed countries, and describes how this pathway differs from the one adopted by developed countries; (2) the digital divide between developed and developing countries from both the supply-side and demand-side perspec- tives; and (3) market concentration in both developed and developing countries, and the tendency toward conglomeration. This report presents some stylized facts that inform the debate surrounding the regulatory approaches that will be required in order to confront the immense challenges presented by the digital economy. The economics of digital markets are not yet well understood, and current policies are not well adapted to digital business models (Goldfarb, Greenstein and Tucker 2015). Several monopo- listic platforms provide their primary product for free, or even at negative prices in order to cross-subsidize business models that are enabled by multisided platforms. Digitalization is making it more difficult to determine the location of economic activity, especially when intangible assets are an important part of value creation. (This is often referred to as scale without mass.) Conflicts of interest in other unregulated digital markets11 – the market for digital advertising, for example - have allowed certain companies to make enormous profits. Another potential regulatory loophole pertains to killer acquisitions whereby an incumbent acquires an innova- tive start-up solely to preempt future competition. These acquisitions occur Examples include mobile payments (Kenya), digital land registration (India), and e-commerce (China). 9   10  https://oxfordbusinessgroup.com/overview/bridging-divide-ever-expanding-digital-economy-creating- widespread-opportunities-0 Alphabet Inc. (the parent company of Google) owns AdX, the biggest online trading exchange for advertisements. 11  Google owns DoubleClick, one of the biggest suppliers of online advertising space. This creates obvious conflicts of interest; Google is both the owner of the exchange and the biggest supplier of advertisement slots to the exchange. There is evidence showing that Google prioritizes its own supply of advertisement slots and limits access to competing suppliers on its exchange. It also withholds inventory from other exchanges, making them less viable. These practices have significantly curtailed competition in the exchange market, prompting a recent antitrust case against the company by the State of Texas. See https://www.nytimes.com/2021/06/21/opinion/ google-monopoly-regulation-antitrust.html?referringSource=articleShare A Spiky Digital Business Landscape  18 disproportionately just below the thresholds for antitrust scrutiny (Cunningham, Ederer and Ma 2020). As policy makers begin to grasp the enormity of the disruptive effects of digital technologies on market competition, experts contend that these challenges will require an agile approach that can adapt to complex and rapidly changing technologies and will also be principle-based and proportional to risks (regulatory “sandboxes",12 for example) in order to create an enabling environ- ment for market contestability, and to manage risks without hampering innovation (PPC 2019). The rest of this report is organized as follows: • Section 2. Definitions, Data Sources, Contributions, and Potential Limita- tions provides the working definition for “digital businesses” that is used in this report as well as other key definitions, and describes the three data sources that were used to create the database. It also discusses potential limitations of the study due to the scope and definition of digital businesses used in the report, as well as data gaps. • Section 3. Digital Business Landscape, Developed vs. Developing Countries describes the characteristics of the global digital business landscape, spanning firms founded over the 1970-2020 period. It summarizes the intensity (number) of digital businesses across countries; the sectoral distribution of digital business- es; sources of funding; patterns of investor exit; and sectoral clustering patterns of digital businesses. It also highlights differences in digital dynamism between developed and developing countries as of mid-2020. • Section 4. Convergence & Digital Growth Pathway investigates whether there have been signs of convergence between developed and developing countries over the years in terms of the size of the digital business sector; subsectoral composition; rates of funding; and investor exit. In particular, it is concerned with the following questions: Are there any regions, or specific countries that stand out in the digital catch-up story? Are there any specific patterns in data-driven or platform-based digital business models across countries (for example by income level)? ​ • Section 5. Big Techs seeks evidence, if there is any, of possible digital market capture by big tech firms globally, that may be leading to market concentration. Specifically, it addresses these questions: Who are the top acquirers of digital firms? And is there evidence of a pattern of hostile acquisitions by internation- al big techs?​ Regulatory sandboxes allow firms to test new products on a small pilot scale before subjecting them to the full 12  regulatory regime. For example, Singapore allows interested parties to test energy generation and distribution technologies in a live environment, but with limits on duration and scale. In Malawi, a low-regulation “drone corridor” was created in which international groups were invited to test their drone operations. And the UK allows live testing of new financial services (PPC 2019). 19  Chapter 1  Motivation: Demand for Evidence on the Global Digital Business Landscape • Section 6. Gender and Green examines the inclusion dimension of digital businesses by turning the spotlight on the gender-related and green (clean tech)​ dimensions. Are digital businesses more inclusive from a gender perspective? Specifically, what fraction of digital businesses have at least one woman on the management team? In what sectors are such businesses prevalent? What share of digital businesses are involved in “green” solutions? How are these firms distributed across the globe? • Section 7. Key Messages presents conclusions, proposes steps for improving the database, and suggests some topics for further study. A Spiky Digital Business Landscape  20 CHAPTER 2  DEFINITIONS, DATA SOURCES, CONTRIBUTIONS, AND LIMITATIONS 21  Chapter 2  Definitions, Data Sources, Contributions, and Limitations Digital Business Definitions This report is focused on digital businesses: that is, providers of digital solutions that develop and manufacture digital technology products, or digital services in the core digital (IT/ICT) sector, or in the narrow scope of digital economy. (See Figure 2.1 for the difference between a narrow vs broad scope of the digital economy). The core digital sector is defined as economic activity that comes from producers of digital content and ICT goods and services (OECD 2020a); the narrow definition of digital economy includes all applications of digital technologies and the production of those technologies, including the platform economy, digital services, and part of the sharing and gig economies that have developed as a result of digital technologies (Bukht and Heeks 2017). This working definition is also aligned with OECD’s defini- tion of the scope of the digital economy, which allows for benchmarking. Depending on the specific user segments, or the levels of sophistication of the various solutions, digital businesses can be further analyzed by comparing B2B vs. B2C, or deep tech vs. applied tech throughout this report. This report does not focus on the broader scope of the digital economy, which includes value creation by “digitalized” traditional (or analog) businesses. This is largely because digitalization is a spectrum, and digital technologies are embedded in a growing number of traditional business operations and business models. This makes it difficult to isolate and distinguish the effects of digital technologies. In this report, digital businesses can be understood to be core digital solution providers for whom digital technology is inscribed in their organizations’ “DNA” – they usually represent 3-10 percent of national GDP (Highfill and Surfield 2022, OECD 2020a). Digital businesses can be largely divided into two distinct categories according to their business cycle: digital start-ups, and established digital businesses (including large platform-based and data-intense firms) that have already reached the exit stage of investments. FIGURE 2.1  Scope of Digital Business Used in This Report Source: Adapted from Bukht and Heeks (2017) and consistent with OECD’s (2020a) tiered digital economy definition framework to allow for benchmarking. A Spiky Digital Business Landscape  22 BOX 2.1 DEMAND SIDE: THE WORLD BANK’S TECHNOLOGY ADOPTION SURVEY It is important to evaluate digitalization opportunities in countries with both supply- and demand-side perspectives. While this digital business database is largely measuring the supply side, the World Bank’s Firm-level Adoption of Technology (FAT) Survey offers a complementary product that measures the demand side. Technology is the key driver of productivity differences across countries and firms. Despite recent progress, existing measures still fall short of providing a comprehensive characterization of the adoption and use of technology, including digital tech, by firms. From a technological standpoint, firms largely remain “black boxes” . First of all, the number of technologies covered in most surveys is rather limited, when compared to how many technologies are involved in the management and production processes of a firm, and they are centered around the use of general purpose technologies (GPTs). Second, the focus is often on the use of advanced digital technologies, which makes it impossible to understand how production takes place in companies that do not use such advanced technologies, therefore most firms in developing countries. Third, since the unit of analysis in these surveys is the firm, the existing studies were not designed to study which business functions benefit from each technology and, more importantly, what the differences are in technological sophistication within firms. To overcome these limitations, the World Bank has developed a new approach to measuring adoption of technology that shifts the unit of analysis from the firm to the business function level. The Firm-level Adoption of Technology survey was designed with the assistance of sector and technology experts who helped to identify key business functions and the technologies used to conduct asks in each of the selected business functions. The survey covers all major sectors in the economy (agriculture, manufacturing, and services), and measures technologies that are applied to general business functions (GBF) and that are common to all companies regardless of the sector where they operate, as well as sector-specific business functions (SSBF). Source: Cirera, et al. 2020. 23  Chapter 2  Definitions, Data Sources, Contributions, and Limitations Data Sources: How the FCI Digital Business Database Was Built This report is based on the newly assembled Digital Business Database of the World Bank’s Finance, Competitiveness, and Innovation (FCI) Global Practice. The database is a firm-level database of 200,000 digital businesses in 190 countries. Figure 2.2 shows how the database was created using three different proprietary data sources - CB Insights, Pitchbook, and Briter Bridges - that use various techniques, from web-scraping to gathering firm information from entrepreneurship networks, venture capital (VC), and other investment deals. They specialize in collecting information on tech start-ups or digitalized firms that would be attractive for VC/private equity (PE) investors due to certain innovative elements in their business models, or core product offerings. To the extent possible, the team also compares these three data sources against the national economic census to verify firm information and assess data representatives. FIGURE 2.2  Creation of the FCI Database from Three Data Sources ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ Covers 190+ countries across all WB regions and 44 digital subsectors* Covers 200K firm-level data Note: Firm-level data is merged from these three sources using a common identifier variable (such as company website information) to avoid duplicates. If a firm has funding data from multiple data sources, the most recent information is used for the last funding round analysis in this report. The total funding information of a firm sums up all of the funding rounds across data sources; or if the data sources only report total funding information that is inconsistent, the larger total funding amount is selected. The FCI Digital Business Database provides a conservative estimate of digital business challenges in developing countries. Comparing this data with a rep- resentative sample of all knowledge-intense and/or digital businesses in the 2019 A Spiky Digital Business Landscape  24 Romania Business Registry Census13 shows that the characteristic of this database is: (1) They are more likely to be investment-ready, profitable, employ more people, and have higher growth rates; (2) Their owners are more educated (mostly with a tertiary degree), more experienced, and more likely have studied or worked abroad. Therefore, the results based on the analysis of this group of investment-ready digital firms likely establish a lower bound (or a conservative estimate) of what is required in order to address the challenges posed by big tech firms or digital convergence, since the capacity of all digital businesses is likely lower, and additional support is required. For example, if an investment-ready e-commerce start-up that is seeded by venture capital faces challenges in building scope and scale economies, especially when facing competition from international big techs who can leverage their global data to build competitive products and provide subsidies to acquire market shares, it is likely that the challenges will be even more acute for a non-investment-ready e-commerce firm that might not be captured by this database. There are limitations in the existing commercial sources of data that cover the digital sector and start-ups. These sources have limitations that prevent them from being used as readily available digital business databases. First, existing data sources do not accurately differentiate digital businesses from traditional business- es. A manual check of the raw firm-level data acquired from the three commercial data sources we used revealed that over 50 percent of the firms were not digital solution providers but rather traditional businesses, digitalized businesses, or holding companies that are of interest to VC and PE investors. Even for databases that focus on start-ups, over 35 percent of the businesses are not digital solution providers. Additionally, the definition of the same digital subsector (for example, e-commerce) differs from one data source to another, making it unreliable to use subsector labels to merge multiple sources and conduct analysis. Lastly, the commercial sources do not identify network-effect business models such as plat- form-based or data-driven models – a unique feature of the digital business sector. Given these limitations, the FCI Digital Business Database uses three concep- tual and methodological innovations. First, it introduces a common dictionary the defines digital businesses and digital subsectors, and harmonizes funding and exit types to allow cross-country comparison when merging different data sources. This database uses a list of keywords to differentiate digital solution providers from digitalized or traditional businesses – and further validated in World Bank country pilot projects when other data sources are available to assess accuracy. Digital sub- sectors are then divided into 44 categories with a distinct definition (see Appendix B). Second, it defines and measures platform-based and data-driven businesses to allow for identification and analysis of the new emerging business models that are tending to exhibit network effects. Platform-based and data-driven businesses are identified also using a list of keywords. Each keyword should yield an accuracy rate of at least 80 percent when distinguishing platform-based or data-driven busi- The FCI Digital Business Database is best at capturing investment-ready tech companies that operate in core 13  digital sectors. For example, a sample comparison with the 2019 Romanian Business Registry Census data showed that the firms in the FCI Digital Business Database are highly concentrated in core digital sectors categorized as NACE Industry 2-digit code 62 (Computer Programming, Consultancy and Related Activities); 63 (Information Service Activities); and 58 (Publishing Activities), see more from Cruz et al. 2022) ). Since the data sources collect company-level data relevant to the investor network, the database captures a limited amount of state-owned enterprises (SOEs), or informal firms and student projects. 25  Chapter 2  Definitions, Data Sources, Contributions, and Limitations nesses from general digital businesses after manually checking their websites and functionalities.14 The manual labeling and checking process of this list of keywords sets up the basis for using machine learning, or other more advanced data analytics in identifying platform or data business models in future updates of the database. Third, it also constructs an investor-centric sample that allows identification of serial risk-capital investors, potentially hostile M&As, and traces investor networks across borders. Details about how the FCI Digital Business Database uses these three innovations are explained in Appendix A. Information Included in the Database The FCI Digital Business Database includes firm-level information that allows cross-country comparison and informs World Bank digital business operations. This cross-sectional information as of 2020 is organized into four modules: firm identification, global landscape business model identification, and funding and exits. The firm-level information provides insights into the firm’s offerings and activities as well as their geographical reach, and their investor interest. Figure 2.3 shows the key variables that is used to derive analysis through- out this report. Other variables include number of employees, balance sheet information, social media presence, and gender information, with higher degrees of missing values and hence not used as core information to derive global analyses. As the dataset to be continuously updated in the future, it strives to build a panel structure and include more variables. FIGURE 2.3  Organization of the Digital Business Database Note: Coverage of the variables and the share of missing values vary by country. Examples of general digital businesses that are not necessarily data-driven or platform-based include, electronic 14  contract signing service, online course/training sites, online health advice, digital game developer, agricultural input management software. A Spiky Digital Business Landscape  26 Limitations Estimates of the size of the digital economy, especially in countries with low levels of internet penetration and coverage by journalists, are conservative. Given the use of a narrow scope to define digital business and the digital economy, the true size of this economy is likely bigger than what is captured through the database. Data sources collect data through web scraping and self-reporting (although the data providers also check the information), which means that some digital solution firms that are in “stealth mode,” or are “student projects” who have not registered their profiles may not be covered in the database, or may be covered with only limited information. While the data sources scrape the web using 24 languages, countries with lower internet penetration and under-reported among businesses and journalists could be underrepresented in the data sources. This can potentially result in bias toward reporting a higher number and funding of digital businesses in certain countries, therefore, we do not quote and compare absolute numbers across countries to derive findings and policy debates. We conduct most of the analyses in this report by developed vs. developing countries to discern major trends and differences. We also conduct robustness checks by excluding the “Big Three” (US, China and India) when appropriate to see whether some of the global patterns still hold given these three countries have a big number of digital businesses represented in this database – see Appendix E. Summary Statistics This report defines developed countries as high-income countries and developing countries as upper middle to lower-income countries, based on the calculations of the World Bank Atlas method. Table 2.1 gives the breakdown of countries and the number of digital businesses in each income category. The detailed list of countries for each income group can be found in Appendix C. TABLE 2.1  Sample Size by Income Group INCOME GNI PER NUMBER OF NUMBER OF GROUP CAPITA DIGITAL BUSINESS COUNTRIES Over Developed High Income 156,376 72 $12,536 Upper Middle $4,046 - 25,663 50 Income $12,535 Lower Middle $1,036 - Developing 11,825 44 Income $4,045 Below Low Income 444 26 $1,035 Source: Authors’ calculation using the FCI Digital Business Database. Note: Income group calculations by according to the World Atlas method as of June 2020. The regions in the FCI Digital Business Database are based on the World Bank’s country classification system as of June 2020 and are shown in Table 2.2. The detailed list of countries in each regional group can be found in Appendix C. 27  Chapter 2  Definitions, Data Sources, Contributions, and Limitations TABLE 2.2  Sample Size by Region REGIONAL NUMBER OF NUMBER OF GROUP DIGITAL BUSINESSES COUNTRIES East Asia and 18,159 15 Pacific (EAP) Europe & Central 2,586 19 Asia (ECA) Latin America 3,318 25 & Caribbean (LAC) Middle East & 2,773 13 North Africa (MENA) South Asia (SA) 6,643 8 Sub-Saharan 4,453 40 Africa (SSA) High Income Europe 47,016 36 High Income Asia 15,633 12 North America 87,861 3 High Income LAC 651 12 High Income MENA 5,103 7 High Income SSA 112 2 Source: Authors’ calculation using the FCI Digital Business Database. Figure 2.4 shows the staff size distribution of digital firms in developed and devel- oping countries from the database (Figure A.4 further gives the regional breakdown for EAP, MENA, and SSA where employment data is available). These figures reveal that medium-sized businesses are the most prevalent in the database. FIGURE 2.4  Staff Size Distribution, Developed vs. Developing Country Firms Sta size distribution 40% 37% 34% 30% 29% 29% 22% 19% 18% 20% 12% 10% 0% 1-9 persons 10-49 persons 50-299 persons 300 or more employed employed employed persons employed Developed (n=1799) Developing (n=1779) Source: Authors’ calculations using the FCI Digital Business Database. A Spiky Digital Business Landscape  28 Figure 2.5 shows the median “ticket size” of an early-stage investment in digital firms across the globe to measure the capital intensity needed to start a digital business. On average developed country digital firms receive more funding than developing country firms; this is especially true after excluding digital firms in China and India. FIGURE 2.5  Capital Intensity: Median Total Funding Raised by Digital Businesses All Developed 2.60(n=69,241) All Developing 1.81 (n=14,344) All Developing Excl. China and India 0.45(n=5753) East Asia & Pacific 4.73(n=7,236) South Asia 1 (n=2,622) Latin America & Caribbean 0.6 (n=1,541) Europe & Central Asia 0.28(n=1,319) Sub-Saharan Africa 0.32(n=1,109) Middle East & North Africa 0.1 (n=517) 0 2 4 6 Median Total Funding Per Firm (Million USD) Source: Authors’ calculations using the FCI Digital Business Database. 29  Chapter 2  Definitions, Data Sources, Contributions, and Limitations A Spiky Digital Business Landscape  30 CHAPTER 3 DIGITAL BUSINESS LANDSCAPE: DEVELOPED VS. DEVELOPING COUNTRIES 31  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries CHAPTER SUMMARY ◆ The number of digital businesses in a country is strongly associated with population and market size (proxied by GDP). Thus, large high-income countries like the United States and the United Kingdom, as well as populous middle-income economies like India, Vietnam, and South Africa stand out with large numbers of digital businesses, confirming that market size is particularly important for businesses that rely on network effects. ◆ After controlling for population and market size, a diverse group of countries that have a more thriving digital business landscape than what would be expected (that is, they show a positive digital business gap) stand out. These outperforming countries include Estonia, Kenya, Israel, Iceland, and Armenia. This demonstrates that there are a variety of digital business pathways available to countries with limited domestic market size. ◆ 60 percent of digital businesses provide solutions in more than one product market; for example, e-commerce firms can also expand to logistics tech, agtech, and food tech. This highlights the importance of scope economies in digital business models and the need to build network effects with a product ecosystem perspective. ◆ Clustering patterns of digital subsectors reveal the dynamics of strong factors that propel the digital economy: the diffusion of general purpose technology (GPT), economies of scale and scope facilitated by intangible asset transfer (e.g., data), and network effects; as well as firm strategies for responding to the challenges of maintaining competitiveness and consolidating their market positions through “tipping”, by providing digital services in multiple adjacent verticals in order to draw more users toward the network. Digital Business Density The economic geography of the digital economy is uneven, and is dominated by countries with large populations and market size. The number of digital business- es operating in a country is strongly associated with its population and market size (proxied by GDP). These include large high-income countries like the US and UK, and populous middle-income countries like India, Vietnam, and South Africa. This appears to be driven by overall purchasing power rather than population alone. However, some smaller countries are strong outliers in digital business density, relative to what their market size and population would predict; this suggests the existence of a variety of digital growth pathways. Figure 3.1 shows the number of digital businesses operating in a country as of Q2 2020 relative to the country’s potential, given its population and market size, referred to here as the digital business gap (see methodological details under the Figure). The countries shaded in green (high performers) have more digital businesses than expected given their population and market size, whereas red-shaded countries (low performers) have less than their potential. The leaders include some of the large economies already mentioned above, but also countries of smaller sizes from different parts A Spiky Digital Business Landscape  32 of the world - for example, Estonia, Kenya, Armenia, and Cambodia, indicating that digital business growth is not necessarily limited by domestic market size.15 FIGURE 3.1  Digital Business Density Across Countries Digital Business Gap De ned As ((actual-potential)/potential)*100% High-Performers Q5 Q4 Q3 Q2 Q1 No Data/Excluded Q1 Q2 Q3 Q4 Low-Performers Q5 IBRD 46700 | OCTOBER 2022 Top 20 High-Performing Countries by Digital Business Density After Controlling for Population and GDP 1. Estonia 6. United States 11. Pakistan 16. Canada 2. Kenya 7. Iceland 12, Singapore 17. Lebanon 3. India 8. Armenia 13. Sweden 18. Vietnam 4. Israel 9. Cambodia 14. Bulgaria 19. Myanmar 5. United Kingdom 10. Finland 15. Cayman Islands 20. South Africa Source: FCI Digital Business Database (using only CB Insights data for this chart, which covers all regions); GDP (US current) and population data from the World Development Indicator Databank. Note: While the world map plots all countries that have at least one digital business, the top 20 table includes countries with at least 30 digital firms, in order to exclude outliers. Gray means that the database does not register any digital businesses in the country: this could be either due to weak digital business development or weak data collection capacity. The digital business gap is derived by regressing the number of digital businesses against GDP and population (all in log) and estimating the potential number of digital businesses (i.e., fitted values of the regres- sion). The gap expresses how much higher or lower the actual number of digital businesses is relative to the country’s potential, expressed as a percentage of the potential (i.e., gap = actual - potential)/potential). Robustness checks with internet access, GDP/per capita, and inclusion of square terms for GDP and population lead to similar results of country ranking, mainly because many of these indicators are highly correlated. The project team also substitutes the number of digital businesses with (1) total investment in digital businesses, (2) the number of digital businesses reaching the IPO/M&A stage to proxy the firm size and quality (not just quantity); these results are reported in Table A.2 in Appendix E- there are substantial overlaps in top 20 countries across these three indicators. While the top digital sectors by number of firms include both B2B and B2C solutions across the world, B2B sectors attract more funding in developed countries. The distribution of the top digital subsectors by number of digital businesses is similar between developed and developing countries, with the two Note that these patterns are based on a cross-sectional analysis of data on a large database of firms, and hence 15  provide a snapshot of the digital business landscape across the world. At this stage, the analysis cannot answer important questions such as what digital market policies and strategies have accounted for countries’ relative success, what different policy options imply for growth and productivity etc. As more data becomes available - balance-sheet panel data on firms - it will be possible to undertake further analyses and answer critical questions on alternative digital growth pathways. 33  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries BOX 3.1 ESTONIA – A ‘DIGITAL REPUBLIC’ With a population of only 1.3 million, Estonia’s success in creating a digital society provides a roadmap for smaller developing countries to foster the digital economy by digitizing public services, embracing international integration, and creating an environment for a digital business landscape to flourish. When Estonia gained independence from the Soviet Union in 1991, the country embarked on a series of fast-track reforms to modernize the economy. From the start, it took a digital approach. It launched a project called Tiigrihüpe (Tiger Leap) in 1997, investing heavily in development and the expansion of internet networks and computer literacy. Within a year of its launch, 97 percent of Estonian schools had internet access and by 2000, Estonia was the first country to pass legislation declaring access to the internet a basic human right. Free wi-fi hotspots started being built in 2001, and now cover almost all populated areas of the country. The government also understood that, in order to create a knowledge-based society, information needs to be shared efficiently while maintaining privacy. In 2001, Estonia created an anti-silo data management system called X-Road through which public and private organizations can share data securely while maintaining data privacy through cryptography. Initially developed by Estonia, the project is currently a joint collaboration between Estonia and Finland. A large number of Estonian government and financial institutions using X-Road came under a cyber-attack in 2007, exposing the vulnerability of centralized data management systems. Estonia decided to use a distributed technology that is resistant to cyber- attack, and in 2012, became the first country to use blockchain technology for governance. Citizens, not the government, own their personal data in Estonia. Other important government initiatives such as Digital ID, visa for digital nomads, e-Residency have fostered trust in digital uptake and create further business opportunities for digital startups. Today, the country is home to more tech unicorns, private companies valued at more than US$1 billion, per capita than any other small country in the world. Skype, the video chatting service that was bought by Microsoft, was launched in Estonia in 2003. Its recent unicorns include payments firm TransferWise (now Wise) and Uber competitor Taxify. Today, Estonia is considered one of the world’s most digitalized country. Source: https://www.pwc.com/gx/en/services/legal/tech/assets/estonia-the-digital-repub- lic-secured-by-blockchain.pdf; https://theconversation.com/estonia-is-a-digital-republic- what-that-means-and-why-it-may-be-everyones-future-145485. A Spiky Digital Business Landscape  34 groups having nine out of the top ten subsectors in common (see Table A.3 in Appendix E). These include sectors offering B2B (business management tech and marketing tech, for example), B2C (like digital media), and hybrid solutions (like e-commerce, and fintech). In terms of total funding received (Table 3. 1), the ranking of the top subsectors shows more divergence, with developed countries attracting more funding in B2B sectors (such as health tech, big data and analytics, security tech, and software and SaaS) while businesses in developing countries are getting more funding in predominantly B2C/hybrid sectors like travel tech, food tech, and entertainment tech) Robustness checks by excluding big markets such as US, China and India were also reported in Appendix E, and they do not yield major differences. TABLE 3.1  Top 10 Digital Subsectors by Total Funding (Developed vs. Developing) DEVELOPED COUNTRIES DEVELOPING COUNTRIES Company- Company- Top 10 Total Funding Top 10 Total Funding Subsector Subsector Subsectors  (Million USD) Subsectors (Million USD) pair N pair N Fintech* 273,605 9053 E-commerce* 140,791 2720 Health tech 230,451 12206 Fintech* 136,675 2187 E-commerce* 198,854 8535 Mobility tech* 108,693 1114 Business man- 170,674 8349 Travel tech 91,739 927 agement tech* Big data and 167,135 5997 Food tech 55,882 852 analytics Security tech 130,374 6327 Tech hardware 55,079 794 Mobility tech* 124,717 3410 Telecom* 53,020 402 Software and Entertainment 112,572 7325 51,534 1333 SaaS tech Marketing tech 107,455 7870 Logistics tech 50,617 1023 Business man- Telecom* 107,138 2791 50,381 1276 agement tech* Source: Authors’ calculations using the FCI Digital Business Database. Note: Company subsector pairing is used in the analysis because some digital businesses can offer digital solutions in multiple sectors, and the total funding does not differentiate which subsector it supports. The top subsectors by number of digital businesses are also reported in Appendix E as total funding is conditional on capital market development.; *Denotes that this sector is common across developed and developing countries. Digital Business Maturity Digital businesses in developed countries show greater market maturity, as indicated by their longevity, sources of financing, types of exits, and deal sizes of exits. A significantly higher share of digital businesses in developed countries has been in existence since the early 1970s, while most of the businesses in developing countries were founded after 2012. For this reason, businesses in developed countries have been receiving financing for longer and are more likely to reach the exit stage with a bigger exit ticket size. Table 3.2 illustrates this divide. The higher exit rates and valuation of developed country firms reflect their size, growth potential, and the depth of the capital market. In recent years, a significantly higher share of developing 35  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries country businesses also started receiving funding,16 though much of this comes from the early stages of the funding lifecycle, for example pre-seed and early-stage venture capital. TABLE 3.2  Summary: Differences in Characteristics, Developed vs. Developing Country Firms DEVELOPED DEVELOPING COUNTRIES COUNTRIES DIFFERENCE Average age of firms 14 9 Significant at 1% Average size of latest 16.7 10.5 Significant at 1% funding (USD million) % Debt financing 9.3% 0.8% Significant at 1% % Pre-seed/seed funding 36.7% 49.5% Significant at 1% Most frequent exit option M&A M&A Exit rate 29% 14% Significant at 1% Average size of exit 508 173 Significant at 5% (USD million) Source: Authors’ calculations using the FCI Digital Business Database. Sectors attracting the highest-valued exit deals differ between developed and developing countries, although fintech is top-ranked in both. Figure 3. 2 lists the rankings of the subsectors that witnessed the highest-value exits in developed and developing countries. Sectors attracting the highest valuations differ between the two groups - only five subsectors are common in the top ten rankings. However, in both groups of countries, fintech is the “unicorn” sector, attracting the highest valued deals on average. The subsectoral rankings in terms of the frequency of exits reveal more similarity between the two groups of countries (see Figure A.5 in Appendix E). FIGURE 3.2  Top 10 Subsectors by Total Ticket Size of Exits – Developed vs Developing Countries Top 10 Subsectors Top 10 Subsectors Exit Ticket Size in Developed Countries Exit Ticket Size in Developing Countries fintech 24%, $23B fintech 55%, $ 45B e-commerce 21%, $20B insurance tech 43% , $ 35B logistics tech 19%, $18B telecom 21% , $ 17B business management tech 17%, $16B tech hardware 9% , $ 7B web services 16%, $16B social network 7% , $ 6B big data and analytics 12%, $12B e-commerce 6% , $ 5B tech hardware 11%, $11B entertainment tech 5% , $ 4B edtech 10%, $10B mobility tech 5% , $ 4B social network 10%, $9B health tech 5% , $ 4B security tech 9%, $9B logistics tech 3% , $ 2B 0% 20% 40% 0% 50% 100% % of Total Exit Ticket Size ($97B) % of Total Exit Ticket Size ($81B) Source: Authors’ calculations using the FCI Digital Business Database. Note: Subsector company pairing is used in the analysis because some digital businesses offer digital solutions in multiple sectors. Those digital businesses are counted in each of their subsectors. Exit information is based on digital businesses headquartered in those developed and developing countries that have funding information. Exit types include IPO, Mergers and Acquisitions, Majority Buyout, Management Buyout, and Other Exits (which includes Sec- ondary Transaction, Stock Distribution, Asset Sale, and Dividend Recapitalization). When the USA are excluded from the developed group, the result does not change. When China and India are excluded from the developing country group, telecom, fintech, tech hardware come to the top while insurance tech drops from the top 10 subsectors. In 2020, the trend reversed back, presumably a consequence of the COVID-19 impact, with funding declining overall. 16  A Spiky Digital Business Landscape  36 Firms in developed countries are generally active in more sophisticated appli- cations relative to their developing country counterparts. A comparison of some of the key applications that digital businesses are engaged in provides another illustration of the maturity of firms in developed countries. For example, Table 3.3 highlights the key applications that developed and developing country firms are engaged in within the health tech digital sector, and underscores how far along the value chain developed country firms are, relative to their developing country counterparts. Developed country firms are active in more value-creation (B2B) applications using AI applications, for example, relative to developing countries where the applications are still in basic services, such as e-pharmacy and making virtual doctor appointments. TABLE 3.3  Top Funded Health Tech applications in Developed vs. Developing Countries HEALTHTECH IN HEALTHTECH IN DEVELOPED COUNTRIES DEVELOPING COUNTRIES Drug development E-pharmacy Health data & AI applications Virtual doctor’s appointment Health insurance solutions Health tracker Medical product distribution logistics Surgical robots Robotic and AI-based healthcare training Medicine delivery through drones Source: Authors’ calculations using the FCI Digital Business Database. Note: “Top funded” is defined as the highest total funding for digital businesses operating in these countries in this sector. Many of the advanced applications in developing countries (such as surgical robots) are from Chinese firms. Subsector clustering patterns point to the role of within-firm spillovers in generating economies of scope, a salient feature of digital technologies. Clusters capture the underlying links, complementariness, and potential spillovers of technology, skills, and information that cut across firms and industries. Figure 3.3 depicts the clustering of digital solutions among firms founded after 2015 in devel- oped countries, and Figure A.7- A 9 in Appendix E depict these patterns before 2015 in Developed Countries; and before and after 2015 in Developing Countries.17 Across the world, certain digital solutions appear to cluster together – for instance, fintech, insurance tech, blockchain and cryptocurrency, gig economy with property tech, construction tech (see Mega Cluster 4); e-commerce, logistics tech and food tech with agtech (see Mega Cluster 5). These clustering patterns can be under- stood largely as the outcome of “digital spillovers” via three channels: 1) Within the firm (internal channels - learning by doing, especially expansion to “adjacent” sectors that require similar skillsets and resources to what the company currently has, generating economies of scope); 2) Among competitors (horizontal channels – innovation by one company is emulated by others in the sector to maintain com- petitiveness);18 and 3) throughout its supply chain (vertical channels – efficiency The year 2015 is considered as the tipping point when cloud computing, a GPT, went mainstream in terms of 17  mass adoption. See https://www.business2community.com/cloud-computing/employing-cloud-2015-01135838 and https://www.computerweekly.com/news/1280095090/Cloud-market-to-reach-25bn-by-2015?amp=1 Horizontal spillovers arise when information held by one company is transferred to others in the same sector, either 18  through the movement of staff, the publication and sharing of knowledge, or simply by replication (Huawei 2017). 37  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries gains that are passed down the supply chain and among complementary services) (Huawei 2017). Traditional businesses, especially those that were integrated with global value chains, have benefitted a lot from horizontal and vertical spillovers, providing developing countries with a growth pathway through international trade and globalization. The internal channel, however, is much more salient for digital business models. This feature offers growth opportunities for developing countries that are very different from the traditional pathway, but that also presents challenges. For example, economies of scope favor the creation of an ecosystem, which provides an incentive for businesses to keep adding complementary services in order to build their network effects, thus erecting barriers to market entry and creating a momentum towards market concentration. Differences in the clustering patterns between developed and developing countries point toward underlying differences in market demand for digital services. The clustering patterns depicted in Figure 3.3 and Figure A.7- A 9 in Appendix E reveal several differences between developed and developing countries. For example, before 2015 the number of clusters was smaller in developing countries; but it has increased to the same number as in developed countries in recent years. There are also some nuanced differences noted, such as the fact that edtech seems to be more closely linked to entertainment and media in developing countries than in developed markets. Similarly, artificial intelligence (AI) appears to be more closely linked to big data analytics, business management tech, software and SaaS, and web services in developed countries, whereas it is more versatile in developing countries, signaling that the underlying market demand might be driving the clustering patterns and differences. Similarly, while e-commerce clusters with agtech, food tech, and logistics tech in both groups of countries, there are notable differences in the solu- tions it provides. For instance, a firm-level analysis on a subsample of firms involved in both e-commerce and food tech (those founded after 2015) reveals that while businesses in both groups of countries are involved in functions along vertical supply chain channels such as food delivery and grocery-to-buyer marketplace functions, developed country firms also focus on several B2B applications such as automation of food production, payment services to restaurants, the recruitment marketplace specialized for the food industry, and smart grocery management. These applications are not yet prevalent in developing countries.19 The analysis is available upon request. 19  A Spiky Digital Business Landscape  38 FIGURE 3.3  Clusters Among Digital Businesses Founded After 2015* in Developed Countries Small Cluster Mega Cluster (more than 2 small clusters linked closely) Source: Research support by World Bank DEC Analytics and Tools Unit (DECAT) using Digital Business Database. Note: Uniform Manifold Approximation and Projection was used for this analysis; this is a dimension reduction technique in data science to visualize sparse multidimensional data. This analysis shows neighbors = 4 results to balance the local versus global structure of clusters. *2015 is considered an inflection point in digitalization because of the mass availability of cloud-enabled digital solutions that started around 2010-2015. Clustering patterns over time can also depict the organic transformation of various industries due to the diffusion of GPTs, and a regrouping of sectors based on synergies in services. One source of digital spillovers is the speed of diffusion in GPTs, which involves the process of user experimentation and discov- ery (“coinvention”).20 For example, before 2015 quantum tech had been a separate cluster in developed countries, but it is now associated with subsectors like biotech, health tech, and nanotech in developed and developing countries – pointing to quantum tech becoming a GPT that is being applied across different use cases (see Figure A.7 and Figure A.8 in Appendix E). Another example is big data analytics: the potential for this solution was significantly enhanced by advances in ML and AI, while developments in cloud computing, with its pay-as-you-go payment models, enabled the diffusion of big data analytics across enterprises of various sizes in multiple sectors (IFC 2020). Other clustering patterns emerge due to increased spe- cialization of services (AI, big data analytics, business management tech, software and SaaS, and web services, for example) or synergies in applications across different sectors (fintech, insurance tech, blockchain and cryptocurrency, and gig economy, for example) – including digital solutions for traditional sectors. (Box 3.2 Simcoe (Simcoe 2015) describes the process that enables a GPT to diffuse to several application sectors, with 20  varying needs and requirements. 39  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries describes the confluence of factors over the last few years that are likely to explain the post-2015 fintech cluster shown in Figure 3.3 and Figure A.9). These changes in turn spur the demand for specific and adjacent skills and capabilities, inducing a supply-side response. Clustering patterns also arise due to the behavior of firms in response to competitive pressures brought on by the nature of digital technologies and the market. As shown in Figure 3.4, in both developed and developing countries, over 60 percent of firms are active in more than one subsector or product market, and over 25 percent operate in three or more subsectors; here the percentage distribu- tion is similar between developed and developing countries. Firms make strategic investments across sectors, as a means of “locking” consumers into the firms’ core product or service by bundling it to other services and products. Such strategies allow firms to “tip” the market toward their own platform/ecosystem by increasing customers’ dependence on the ecosystem and its interconnected web of platforms. Users may be locked into a company’s products or services if, for example, their data is at risk of being lost if they transfer to another company’s products or services. For example, Apple recently entered the credit card business by launching the Apple Card, which is essentially an analog extension of Apple Pay. This service allows customers to connect their debit and credit cards to their iPhone and pay for any purchase – like any other credit card. Apple’s objective is not to compete with other credit card companies but to keep its users from switching to Android phones since the card is tied to its own ecosystem; thus the convenience of Apple Card helps lock in users. Other big tech firms use similar strategies to prevent customers from “multi-homing,” - connecting to other platforms - and to consolidate their market position. FIGURE 3.4  Distribution of Firms According to Presence in Number of Subsectors/Markets 50% 40% 40% % of Digital Business 40% 34% 33% 30% 20% 16% 17% 6% 7% 10% 4% 4% 0% 1 Subsector 2 Subsectors 3 Subsectors 4 Subsectors 5 Subsectors or more Developed (n=140K) Developing (n=38K) Source: Authors’ calculations using the FCI Digital Business Database. Note: This finding is similar when excluding USA, China and India (Figure A.10). A Spiky Digital Business Landscape  40 BOX 3.2 THE FINTECH, INSURANCE TECH, BLOCKCHAIN AND CRYPTOCURRENCY, GIG ECONOMY, PROPERTY TECH, CONSTRUCTION TECH MEGA CLUSTER The cluster analysis revealed that many firms tend to operate simultaneously in a range of subsectors like fintech, insurance tech, blockchain and cryptocurrency, gig economy, property tech, and construction tech – making this a “mega cluster” of digital businesses. This leads to the question of why. The global financial crisis of 2007–08 eroded trust in the banking system, prompting several technological firms to operate outside the regulatory framework through peer-to-peer (P2P) networks, using a blockchain protocol. Over 2,000 platforms were developed in China alone (Fernandez-Vazquez, et al. 2019). In 2018, the Fintech unicorn industry was valued at $85.8 billion, with seven of the ten fastest-growing firms achieving their unicorn status - private companies with a valuation of US$1 billion - within 12 months of their company’s inception (Pompella and Matousek 2021). Gig economy workers are an attractive segment for fintech firms. Gig work is occupying an increasingly bigger share of the global workforce, facilitated by the growth of the platform economy. However, earnings from gig work tend to be unpredictable, making financial planning and budgeting challenging. For this reason, this segment is generally underserved by banks, making them a growing opportunity for fintech firms. Fintech companies are changing how people bank and manage transactions to level the playing field and increase financial inclusion. Gig employers are themselves finding smarter ways to combine employee data - such as historic hours worked and money earned - with new fintech innovations to help their gig workers smooth out their income volatility. These employers are giving gig workers easier access to the money they have earned. For example, Uber gives its drivers the option of instant access to their earnings in exchange for a small fee. Similarly, Lyft offers an express pay service to its workers. Some fintech firms, like the US digital banking platform Oxygen, specialize in serving the needs of the freelance economy. Their app allows users to manage everyday expenses on their Oxygen Visa debit card and to access credit, helping to provide freedom, flexibility, and predictability. Similarly, the Indian app Bon gives access to working capital to workers of the gig economy like taxi drivers, delivery personnel, contractors, and the self-employed. The company’s payments card can be used at several merchant locations around the country. The rise of fintech, changing consumer behavior, and advanced technologies are disrupting the insurance industry as well. Insurance tech and technology start-ups are offering new services such as risk-free underwriting, on-the-spot purchasing, activation, and claims processing (Deloitte 2020). Property tech has also evolved with developments in fintech. Services around housing - real estate finance, insurance, and contract signing – are changing rapidly. Yave, in Mexico, for example, offers online mortgage lending services with digital tools to complete the entire mortgage process. The construction tech sector, which is focused on the “built world” involves architects, engineers, construction firms, and facilities managers, and underpins the property tech sector. Fintech solutions such as accounts payable automation solutions, budget-friendly equipment financing, and insurance are also creating new possibilities in the construction industry (Tian, et al. 2020). 41  Chapter 3  Digital Business Landscape Developed Vs. Developing Countries A Spiky Digital Business Landscape  42 CHAPTER 4 CONVERGENCE & DIGITAL GROWTH PATHWAY 43  Chapter 4  Convergence & Digital Growth Pathway CHAPTER SUMMARY ◆ There are early signs of convergence in the digital business landscape, but this is mainly driven by China and India, and a few large digital hubs in their respective regions. ◆ There is evidence of a gap in the B2B market segment and “deep tech” sectors between developed and developing countries. Fund flows to consumer-facing applied technology sectors such as e-commerce and fintech have been catching up in developing countries since 2010, but the number and size of fund flows to B2B solutions are still small. Developing countries also lag behind in funding for deep tech sectors such as IoT and quantum tech. Developing countries can consider both supply-side and demand-side measures to minimize the B2B and deep tech gap. ◆ There is some evidence of spatial clustering: digital businesses appear to be using regional hubs to achieve scale and to internationalize, and platform and data-driven business models are important conduits for building network effects through these spatial networks. ◆ Fintech and e-commerce - plus the digital services that build on these two application ecosystems - appear to foretell the growth of the digital economy, as the growing number of digital firms and fund flows show. Early signs of convergence between developing and developed countries appear to be mainly driven by China and India, who dominate the digital business landscape among developing countries. The post-2012 developments in the digital landscape of developing countries, as described in Section 3, were mainly driven by China in the East Asia and Pacific (EAP) region, and by India in the South Asia region. These two large emerging markets make up a big e share of the business intensity, funding received, and successful exits in their regions. When comparing patterns of digital businesses with and without China and India, data show that the main effect of these two countries is to bring more sectoral diversity to their respective regions – likely reflecting the breadth and diversity of digital subsectors that thriving digital markets of such large sizes can support. However, there are also encouraging signs of the digital economy taking off rapidly in several big economies in other regions. (See Box 4.1 for a description of start-ups in Africa that are attracting investor interest). Newly established digital businesses in developing countries – led by China and India – have received significantly more funding in “applied” digital sectors relative to those in developed countries. Sectors such as e-commerce, travel tech, and fintech – which may be classified as “applied” and consumer-facing digital subsectors – have attracted more funding in developing countries in recent years, as Figure 4.1 highlights. This could be a consequence of these sectors being more mature and already established in developed countries, while still growing rapidly with more start-ups coming to the market in developing countries. A Spiky Digital Business Landscape  44 BOX 4.1 ENCOURAGING TRENDS IN FUNDING FOR AFRICAN START-UPS Access to funding remains a significant barrier for digital entrepreneurs in emerging markets. One reason for the paucity of funding is the lack of scale and predictability in local markets, which discourages large institutional investors who do not have the capacity or the risk tolerance for small and volatile deals. However, recent trends are encouraging; development finance institutions (DFIs) have started investing in venture capital firms located in developing countries, which are better suited to distributing funds in these markets than foreign institutional investors. Africa is also witnessing considerable interest from investors (PPC 2019). In 2020, there were at least 370 active investors in Africa, an increase of 43 percent over the previous year, when there were 261. The investors include institutional investors, VC firms, family offices, and angels active in Africa; this signals increasing confidence and interest in African start-ups across all stages of the start-up lifecycle. In terms of the number of deals completed by investors in 2020, the majority made between one and four investments. Only seven made more than 10 investments over the course of the year. The most active investor was Kepple Africa Ventures, which backed 36 start- ups. Accelerator-linked investments were also quite prevalent, with Flat6Labs, Y Combinator, Founders Factory Africa, 500 Startups, and MEST Africa among the most prolific investors on the continent. While investors are increasingly placing bets on start-ups in less developed start-up ecosystems in Africa, those investing at bigger ticket sizes are continuing to focus on the “big four” – Kenya, Nigeria, South Africa, and Egypt. The total amount of investment that went into start-ups from these four countries was $626 million, accounting for 89 percent of the overall funding in Africa (Disrupt Africa 2020). The continuing building up of an investment-ready pipeline of digital start-ups is likely to be key. 45  Chapter 4  Convergence & Digital Growth Pathway FIGURE 4.1  Founding Year of Digital Firms Receiving Funding in Applied and Deep Tech Applied Tech (e-commerce, fintech & travel tech) % of funded businesses 60% 50% 40% 34% 32% 23% 24% 20% 12% 10% 9% 4% 4% 0% 1971-1999 2000-2004 2005-2009 2010-2014 2015-2020 Founding year Developed (n=8775) Developing (n=2436) Deep Tech (AI/ML, internet of things, quantum tech) 100% % of funded firms 80% 67% 60% 54% 40% 28% 24% 20% 7% 5% 2% 6% 3% 5% 0% 1971-1999 2000-2004 2005-2009 2010-2014 2015-2020 Founding year Developed (n=3517) Developing (n=728) Source: Authors’ calculations using the FCI Digital Business Database. Note: The results are similar when excluding USA, China, and India. There is evidence of a continuous digital divide in “deep tech” . Figure 4.1 illustrates that more and more deep tech firms (such as firms involved in quantum tech and the internet of things (IoT)) were founded and funded since 2015 (this is also when cloud computing technologies became mainstreamed). However, in terms of the absolute number of deep tech firms being funded, there is still a substantial difference: it is about three to four times more in developed countries. Furthermore, the fund flows to deep tech firms in developing countries mainly consisted of Chinese firms (54 percent), and Indian firms (15 percent); most other developing countries account for only 1-2 percent of these fund flows. Deep tech technologies are often disruptive and transformative, and early movers can estab- lish a powerful edge over the competition. They require digital skills, advanced digital infrastructure, and intellectual property rights protection, conditions where developed countries have the advantage. They also require significant investment in basic research and promising technologies, even if clear applications are not evident yet.21 While deep tech might not be priorities for all countries, if a country wants to make “bold bets” on deep tech, in addition to making heavy investments in human capital and digital infrastructure, policy makers also need to consider other longer-term objectives, demonstrating the patience required for technology investments to pay off, facilitating basic research, co-creation of knowledge, and strengthening investor confidence. 21  https://www.bcg.com/en-us/publications/2021/how-european-corporations-becoming-deep-tech-investors A Spiky Digital Business Landscape  46 Supporting local digital businesses to develop affordable and proven B2B solutions that are tailored to local needs and user skills can be a good catch- up strategy for developing countries to close the B2B gap with developed countries. Table 3.1 and Figure 4.1 reveal a consistent gap in both the supply of and demand for B2B digital applications in developing countries. This is likely due to several reasons. Corporate customers in several B2B markets often demand services that are standardized across borders, stable at enterprise-grade, and are well-integrated with other offerings – including with legacy software. The absence of a strong regulatory environment, including a legal framework for resolving disputes in a fair and transparent manner, is often an impediment for overseas enti- ties to partner with developing country businesses for developing B2B solutions. There is also evidence of a lack of demand for complex applications in developing countries. Even in emerging major digital markets, B2B frontier tech buyers are small and fragmented, and governments are usually the biggest customers of B2B solutions (Grewal, et al., 2015). This is invariably a consequence of two factors: affordability and usability (World Bank Forthcoming). Affordability not only refers to the costs of the digital applications themselves, but also to the costs and reliability of the complementary infrastructure needed to use them (internet, electricity, roads, and logistics for some types of applications). Usability refers to the attractiveness of the B2B applications that meets the productive (or other) needs of local users while accounting for their capability and skill sets and the available local infrastructure. This is a unique comparative advantage that can be exploited by nimble local digital firms that are much more familiar with the local context and conditions than international big techs. Government can also play a role in addressing these two underlying roadblocks in order to unlock a B2B market by addressing skills gaps; avoiding the favoring of incumbents so that start-ups can enter the market, and expand; developing a regulatory framework that enforces the rule of law and that addresses socioeconomic and trust issues; and assuring gender-equal access to digital devices and services, for example. Mergers and acquisitions (M&A) are the dominant exit strategy for digital businesses, with implications for both static and dynamic efficiencies. Figure 4.2 shows the distribution of types of exits of digital firms in developed and developing countries before and after 2010. In both groups, M&As are the primary exit path, while IPOs also continue to be an important exit option.22 The importance of M&As for developing country firms attests to their potential value to investors. This has implications for competition and innovation; the literature on “killer acqui- sitions” asserts that incumbents acquire innovative start-ups in order to get rid of potential competition and to reinforce their dominance in the market. This practice impedes static efficiency – the competitive threat required to discipline the market behavior of incumbents. It also hinders dynamic efficiency - the development of disruptive technologies and innovations that are valuable to the public (Lemley and McCreary 2021). Another motivation for acquisitions by big tech firms is purchasing The IPO option is limited due to less developed financial markets in developing countries. However, Chinese 22  digital firms make extensive use of this exit option, contributing to its high share in the developing country group. 47  Chapter 4  Convergence & Digital Growth Pathway BOX 4.2 KhmerOS – LOCALIZATION FOR INCLUSION AND VALUE CREATION Localization helps build technical expertise in the local community, reduces dependence on imported software, helps narrow the linguistic digital divide, and contributes to the growth of the local ICT industry by giving rise to other innovations once confidence levels are established. New business models for developing and emerging countries can be established, which would not be possible with proprietary software. Several developing countries have used free and open-source software (FOSS) to tailor products and services to local requirements, without having to develop their own software from scratch. A key benefit of FOSS-based local open innovation is that everyone can translate the interface of a FOSS program into a local language and release it. Many localization efforts have taken place in Asia. A well-known example is Cambodia’s KhmerOS initiative, which began in 2004 to create a free and accessible open-source operating system by translating commonly used applications such as word processing, e-mail, spreadsheets and an internet browser into Khmer, the local language. This effort has created a sustainable low-cost use of local language ICT in education, government, and local society, reducing the digital divide. Project workers developed and standardized Khmer scripts and fonts, designed and manufactured keyboards and printed manuals in Khmer for the applications.  In 2005, KhmerOS prepared and printed a Khmer user guide for OpenOffice and started an ambitious training plan of government officials, computer teachers, NGO workers, and students in Phnom Penh and several provinces. Knowledge of the KhmerOS program soon spread nationwide, with schools and government operations all over the country using the new software and technology. The technology for the Khmer script keyboards and textbooks was transferred to local commercial vendors for manufacture and sale. KhmerOS has also developed a localization project for the Tetum language in East Timor, supported a similar effort in Uganda, and given direct technical support and/or advice for localization in Laos, Vietnam, Myanmar, Bhutan, Nepal, Bangladesh, Tanzania and - to a lesser degree – to other countries in Asia, Africa and Latin America. Source: http://www.bdsknowledge.org/dyn/bds/docs/816/GIZ_Manual_IT-Sector-Promo- tion.pdf, https://www.wipo.int/edocs/mdocs/copyright/en/wipo_cr_wk_ge_11/wipo_cr_wk_ ge_11_3.doc A Spiky Digital Business Landscape  48 scarce skills and talent via acqui-hires where the buyer seeks to secure the target’s talent, engineers, and personnel, rather than to develop or monetize the target’s technology, products, or services.”23 FIGURE 4.2  Distribution of Exit Strategies, Developed vs. Developing Country Firms Developed Countries Developing Countries 100% 80% 60% 40% 20% 0% 0% 20% 40% 60% 80% 15.3% (n=1219) 41.8% ( n=141) 3.6% (n=1335) 21.4% (n=899) 0.3% (n=24) 2.7% (n=9) 2.6% (n=966) 8.1% (n=339) 6.9%(n=548) 2.1% (n=7) 1.3%(n=477) 0.6% (n=24) 74.9%(n=5970) 49.6% (n=167) 90.9% (n=33814) 63.3% (n=2658) 2.7% ( n=212) 3.9% (n=13) 1.7% (n=615) 6.7% (n=282) Before 2010 (N=8K) After 2010 (N=37K) Before 2010 (N=337) After 2010 (N=4202) Source: Authors’ calculations using the FCI Digital Business Database. Note: Exit information is based on digital businesses headquartered in developed and developing countries. Exits include IPO, Mergers and Acquisitions, Majority Buyout, Management Buyout, and Other Exits, including Second- ary Transaction - Stock Distribution, Asset Sale, and Dividend Recapitalization. IPO exits in China dominated the developing country IPO deals. * The percentage difference in the exit category is significant at a 0.05 level. Across the world, there is some evidence of spatial clustering – that is, digital businesses appear to be using regional hubs to achieve scale and internationalize. Figure 4.3 highlights the top overseas destinations for regional firms– defined as those that are headquartered in the region and operate in at least two countries in the region – in Brazil, Egypt, and Nigeria. Regional firms tend to be headquartered in the biggest economies of the region – regional “hubs” – and to have operations in a notable number of adjacent countries or other digital hubs in the region (for example, Nigeria and Kenya are both regional hubs in Sub-Saharan Africa, as are Brazil and Argentina in Latin America and the Caribbean). This pattern of regional expansion suggests that geography matters even for the digital economy. This in turn offers a potential pathway for developing country firms to connect with, or set up operations in, major regional hubs to achieve scale and internationalize. Initiatives such as a single- regional digital market and digital trade policies will likely support this spatial expansion pattern. https://louislehot.com/the-state-of-the-acqui-hire-in-2021-the-good-the-bad-the-why-and-whats-next/ 23  49  Chapter 4  Convergence & Digital Growth Pathway FIGURE 4.3  Regional Digital Business Hubs and the Destination of Regional Digital Businesses ARAB REP. JORDAN KUWAIT OF EGYPT BAHRAIN QATAR UNITED ARAB SAUDI EMIRATES BURKINA FASO BENIN ARABIA COSTA RICA PANAMA CÔTE NIGERIA D'IVOIRE COLOMBIA TOGO UGANDA GHANA CAMEROON ECUADOR KENYA PERU BRAZIL CHILE SOUTH AFRICA ARGENTINA URUGUAY IBRD 46845 | OCTOBER 2022 Note: Regional hubs (Brazil, Nigeria, and Egypt as examples) are developing countries that are among top 20 coun- tries by the number of digital businesses, controlling for population and market size (Figure 3.1), and that have the greatest number of regional businesses in the region. Regional businesses are digital businesses that operate in at least two countries in the same region. The thicker the lines, the more regional businesses from regional hubs are operating in the destination country. In addition to spatial expansion, platform-based and data-driven business models are important conduits for expanding across subsectors and realizing scope economies, presenting a pathway for small developing countries to tap into the digital economy. Across the world, around 16 percent of digital businesses are estimated to be platform or data-driven businesses. Moreover, as Figure 4.4 shows, developing countries have a higher share of these businesses. Table 4.1 presents cross-country pair-wise correlation coefficients between population size and GDP per capita, and the intensive margin (the percentage of platform-based or data-driven businesses in a country) and the extensive margin (the number of sub- sectors with the presence of at least one platform-based or data-driven business).24 Interestingly, given the network effects associated with these business models, population size is not associated with the intensity of platform-based or data-driven business models. However, it is a strong predictor of the extensive margin of these businesses in both developed and developing countries. This follows from the fact that the source of value creation in these business models is competition within rapidly emerging “ecosystems,” which includes a variety of businesses from vastly different sectors (McKinsey&Company 2017).25 This reinforces the notion that one potential digital development pathway for smaller developing countries is the See Appendix A on how to identify platform and/or data-driven businesses based on the company description in 24  the database. A digital ecosystem refers to “a highly customer-centric model, where users can enjoy an end-to-end 25  experience for a wide range of products and services through a single access gateway, without leaving the ecosystem. Ecosystems will comprise diverse players who provide digitally accessed, multi-industry solutions.” (McKinsey&Company 2017) A Spiky Digital Business Landscape  50 establishment of economic links – either through a bilateral/regional cooperation agreement, or privately with major digital economies – to tap into the ecosystems created by economies of scope. To be effective, such arrangements need to address the economic and security concerns of smaller countries. This calls for inclusive policies that provide for online consumer and supplier protection, digital taxes, and competition policies, as well as a flexible approach that allows for course correction based on the impacts experienced by such initiatives in the member countries. FIGURE 4.4  Region-Wise Shares of Platform and Data-Driven Businesses % of Platform or Data Driven Businesses in Each Region 30% 23% 24% 24% n=1149 n=13893 n=27 25% 18% 18% 18% 18% n=3298 16% 18% 18% n=812 20% 16% n=6868 14% 15% n=416 n=119 15% n=501 n=1344 n=24457 n=497 15% n=2252 n=7017 10% 5% 0% All Entire Asia Entire Entire Latin Entire Middle North South Entire Regions * Pacific * Europe and America and East and America Asia Sub-Saha- Central Asia Caribbean North Africa * ran Africa Developed % Developing % Source: Authors’ calculations from the FCI Digital Business Database. Note: A firm can be both a platform-based and a data-driven business. North America only has developed countries, and South Asia only has developing countries; so the empty columns do not indicate the lack of data in these regions. * Indicates that the percentage difference between developed and developing countries in the region is significant at a 0.05 level. TABLE 4.1  Pair-Wise Correlation Coefficients Between Population and GDP Per Capita, and the Intensity and Extensity of Platform and Data-Driven Businesses CORRELATION PLATFORM & PLATFORM BUSINESS DATA BUSINESS MODEL COEFFICIENT DATA BUSINESS MODEL MODEL (PAIRWISE) Developing Developing Developing Developed Developing (excl. CHN Developed Developing (excl. CHN Developed Developing (excl. CHN & IND) & IND) & IND) Intensive margin X GDP -0.2490 -0.2272 -0.2362 -0.0062 -0.4773* -0.2491 -0.1399 0.0836 0.0665 Per Capita Platform business intense in a few countries only Intensive Margin X -0.0102 0.0963 0.1097 -0.1549 -0.1305 0.1803 0.0716 0.1596 -0.0291 Population Extensive Margin x GDP 0.0864 0.1577 0.1425 0.1946 0.2326 0.2168 0.3529* 0.1825 0.1696 Per Capita Data-driven business in developed countries operate across more subsectors on average Extensive Margin X 0.3389* 0.3424* 0.4470* 0.4124* 0.3810* 0.4937* 0.3520* 0.3773* 0.4759* Population Source: Authors’ calculations using the FCI Digital Business Database. Note: Intensity and extensity analysis only uses data from CB Insights, which covers all regions. The results in the table are also robust to the exclusion of the US and Germany from the developed country group. 51  Chapter 4  Convergence & Digital Growth Pathway In developing countries, e-commerce and fintech companies that leverage platform-based and data-driven business models foretell the growth of the digital economy.26 Fintech and e-commerce are the top sectors both in terms of number of firms and the funding received across the globe, attesting to the impor- tance of these business models in the digital pathway for developing countries. (See Box 4.3 for a description of the various business models prevalent in e-com- merce). The global value of sales through e-commerce was estimated at $27 trillion in 2019, equivalent to 30 percent of world GDP that year.27 Low barriers to market entry provide opportunities for small firms to become third-party sellers of goods and services on platforms in developing countries (UNCTAD 2019). Local platforms may also provide more convenience for consumers through shorter shipping times, tailored payment options, products more suited to local markets, and local language interfaces. Other potential benefits for the domestic real economy may be related to the development of links with local industries, logistics services, and suppliers (UNCTAD 2019). Box 4.4 describes Asia’s distinct approach to fostering the region’s digital sector, in keeping with the preference of consumers in that region for using mobile phones to access the internet. 26  Several factors explain the dominance of platforms in the digital economy: the first significant factor is direct and indirect network effects; the more users of a platform there are, the more valuable that platform becomes for everyone, even if everyone is not using the same service. A second factor is the platforms’ ability to extract, control, and analyze data; as both intermediaries and infrastructures, platforms are well-situated to record and extract all data related to online behavior and interactions among the platform users. These data are used to develop “digital intelligence” about user behavior and preferences. The data may also be sold to third parties. ​ The growth potential of digital platforms is directly related to this capacity to collect and analyze digital data. A third factor is the dynamics of path dependency; the cost of switching to another platform increases once the platform has gained traction. The significance of “platformization” stems from the fact that in the digital economy, different platforms - as opposed to supply chains, nations, or sectors, for example – have become the basis for understanding the division of value (UNCTAD 2019). Of the total e-commerce sales of $27 billion in 2017, 76 percent were sales in the top 10 countries: USA, Japan, 27  China, South Korea, UK, France, Germany, Italy, Australia, and Spain (UNCTAD 2021) A Spiky Digital Business Landscape  52 BOX 4.3 B2B VS B2C E-COMMERCE E-commerce is largely comprised of the business-to-business (B2B) and business-to-consumer (B2C) trade, along with a small volume of consumer-to-consumer (C2C) trade. Globally, B2B transactions account for over 80 percent of e-commerce sales volume (UNCTAD 2019). While B2C has more users, the volume of transactions is significantly higher in B2B. The B2B commerce market is far more complex than online retail (B2C). PC vs. Mobile: B2B e-commerce refers to transactions conducted between one business and another, for example between a wholesaler and a retailer. In this case, the items purchased will be sold to others again, and are often customized to the buyer’s special requirements. Because of the volume and complexity of B2B orders, these transactions are more often made through PCs than mobile phones. For instance, more than 80 percent of the orders in Alibaba.com, one of the biggest B2B trading platforms in the world, are made using PCs. B2C trades are those made between business and individual consumers: these transactions represent final consumption. The orders therefore tend to be much smaller, and in a “ready to ship” form. For these reasons, the predominant share of B2C orders is made via mobile phones; more than 70 percent of the orders in Taobao, a prominent B2C platform, are made through cell phones. Trust and Relationship Building: Most B2B trade represents long- term relationships, with customers making recurring purchases. Thus, from a seller’s point of view, B2B transactions require a lot more effort in order to establish client relationships, and to nurture them to garner trust. It is also very common for B2B trades to be moved off-line after successful trades and mutual trust has been established between the parties. B2C trade, however, relies on other trust-building mechanisms such as consumer reviews, chats, and aftersales services. E-commerce Policy: Cross-border B2B trades are most often characterized as FOB (free on board); the seller is responsible for complying with domestic (local) regulations, but that responsibility ends at the port. Once the goods are loaded to ship, the B2B buyers are responsible for the process at the receiving end: getting the import certification, clearing import customs, and so on. In cross- border B2C trades, by contrast, the sellers are responsible for the goods until they are delivered to the consumers. Until recently, cross-border B2C trade was loosely regulated. Existing multilateral 53  Chapter 4  Convergence & Digital Growth Pathway BOX 4.3 CONT trade rules cover some issues related to e-commerce, but they are not comprehensive. Some countries have established a de minimis level below which import duties do not apply, and typical B2C orders fall below this level. Hence some e-commerce policies enacted in recent years, such as e-customs and consumer protection, have limited effects on B2C trade. Some countries are placing more responsibility on e-commerce platforms to ensure product safety and collect taxes. For instance, the European countries recently started requiring e-commerce platforms to help enforce value-added taxes. Source: Authors’ interview with researchers at the Luohan Academy of the Alibaba Group The clustering of digital solutions indicates that improved logistics, stemming from immediate demand for e-commerce across the developing world, offers a big opportunity for firms to increase productivity across multiple sectors. An effective logistics sector is a core enabler of commerce; yet there is a persistent gap in logistics capabilities between high-income and low-income countries (WBG 2018). New business models, including e‐commerce, are contributing to the growing demand for logistics services. The previous sections of this report reveal that across the world, e-commerce solutions tend to be offered at the same time as logistics solutions. A firm-level analysis of 50 firms that are involved in both e-com- merce and logistics solutions in both developed and developing countries reveals that across the world, these firms are engaged in bolstering supply chain capabili- ties for e-commerce.28 This also suggests that there are significant opportunities for firms in developing countries to make the logistics industry more resource-efficient and responsive to customer needs with the use of digital tools. For example, digital platforms, in addition to helping businesses connect with consumers, can better match shipment demand with logistics capabilities through end‐to‐end online booking services. Such solutions would be particularly beneficial to micro, small, and medium enterprises (MSME) sellers in their internationalization efforts, since they often face multiple hurdles in arranging international shipments. Moreover, the internet of things (IoT), coupled with advanced data analysis techniques, allows for real‐time analysis of supply chain data (OECD 2020b). The analysis is available upon request. 28  A Spiky Digital Business Landscape  54 BOX 4.4 ASIA’S DIGITAL STRATEGY LEVERAGING CONSUMERS’ “MOBILE- FIRST” APPROACH AND “SUPER APPS” Asia’s digital growth has a distinctly regional flavor, stemming from the region’s consumers taking a “mobile-first” approach to the internet, instead of using personal computers or other devices, as well as the opportunities arising from inefficiencies in the off- line sectors. In Asia, mobile e-commerce spending accounts for around three quarters of total spending on e-commerce, compared with 37 percent in the European Union (EU) and 31 percent in North America. Relative to other regions, consumers in Asia spend much more time on their mobile phones, reflecting Asia’s mobile- first approach to the internet (McKinsey Global Institute 2020). This feature has led to the phenomenon of “super apps” in Asia, which refer to a closed ecosystem of several apps that are used by people every day. These super apps cover multiple aspects of people’s lives – social media, retail, healthcare, mobility - and offer a seamless, integrated, contextualized, and efficient experience (McKinsey Global Institute 2021). A prominent example is Tencent, the Chinese tech conglomerate and leader in the “super app” realm, which has taken its WeChat app – initially a messaging app – and expanded this billion-plus user app into an ecosystem of services that includes taxi rides, payments/virtual wallets, hotel reservations, games, medical consultations, and more. Several inefficiencies in off-line sectors, including fragmentation and weak logistics, also present opportunities for firms to leverage digital solutions and expand their footprint. 55  Chapter 4  Convergence & Digital Growth Pathway A Spiky Digital Business Landscape  56 CHAPTER 5 BIG TECHS & THE EMERGENCE OF “DIGITAL CONGLOMERATES” 57  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” CHAPTER SUMMARY ◆ In addition to traditional venture capital and private equity funds, big tech firms are playing a disproportionally important role as investors in digital businesses in developing countries. ◆ The state is also an active investor in digital businesses, but in developed countries the public sector invests mainly in early-stage businesses, while in developing countries the public sector tends to invest in later-stages businesses with a large ticket size, by leveraging state-owned financial institutions (mainly from China). ◆ In terms of M&A activities, there is evidence of strong interest by acquirers, especially big techs from Brazil, Russia, India, China, and South Africa (BRICS) and European countries, in the digital businesses of developing countries. ◆ While acquisitions in the same sector are most common, cross-sectoral acquisitions are also prevalent, consistent with the returns to scope that characterize digital business models, and market power consolidation. ◆ There is suggestive evidence of a u-shaped pattern in market concentration of acquirers involved in high-valued acquisitions in relation to market maturity: concentration is high in the early stages of digital development, decreases as digital businesses develop, and tends to increase again in mature markets. This potentially signals the need to sequence regulations as digital businesses develop, especially to balance the benefits of consolidation and vertical expansions versus the cost of anti-competitive behaviors. Big tech firms play a disproportionally important role as active investors in developing countries. As Figure 5.1 shows, these firms – which are mostly from the US and China – are playing a bigger role as investors in developing countries, while private equity (PE) and private venture capital (VC) is still the most important risk capital across the globe. Table 5.1 also reveals that the footprint of big techs is exten- sive – many big techs can have several hundred firms as their investees.29 A robust- ness check using deal size (in US dollars) to rank top investors also confirmed that big techs play a major role as investors in digital businesses in developing countries.30 This finding is cross-checked and confirmed with the Orbis ownership structure database at the World Bank 29  Finance, Competitiveness, and Innovation Global Practice. While funding information is only available per company and not per investor (since one firm can have many 30  investors), robustness checks were carried out using the total deal size divided by the total number of investors, to establish a lower bound estimated investment size of the top investors. Even with this conservative estimate, tech companies still make up a quarter of the top 20 investors in developing countries. A Spiky Digital Business Landscape  58 FIGURE 5.1  Composition of the Top 20 Investors, Developed vs. Developing Countries Top 20 Investors in Top 20 Investors in Companies Headquartered Companies Headquartered in Developed Countries in Developing Countries Tech company Tech Other (sovereign (incl. CVC arm) company wealth fund, public 5% (incl. CVC innovation fund) arm) 25% 20% Financial VC/PE Other 70% Financial (university VC/PE holding, SOE) 75% 5% Source: Authors’ calculations using the FCI Digital Business Database. Note: CVC means in-house corporate venture capital units. This breakdown is by the number of deals, not the deal size ($). TABLE 5.1  Top 20 Investors and Number of Investees, Developed vs. Developing Countries Top 20 investors of firms in Top 20 investors of firms in developed countries and # of investees developing countries and # of investees Country of Country of Investor # Investees investor Investor # Investees investor 1 Y Combinator 1269 United States 1 IDG Capital 432 China Plug and Play Sequoia Capital 2 1114 United States 2 307 China Accelerator China 3 Techstars 1108 United States 3 ZhenFund 256 China U.S. Department Tencent 4 916 United States 4 247 China of Defense Holdings 5 Intel Capital 764 United States Microsoft 5 231 United States 6 500 Startups 661 United States ScaleUp 7 MassChallenge 656 United States Matrix 6 221 China New Enterprise Partners China 8 593 United States 7 500 Startups 217 United States Associates National Science Google for 9 591 United States 8 Startups 202 United States Foundation 10 Horizon 2020 577 EU Accelerator 11 Accel 538 United States Shenzhen 9 201 China United Capital Group 12 3i Group 499 10 Legend Capital 184 China Kingdom 13500 Accelerator 462 United States Qiming Venture 11 183 China Kleiner Perkins Partners 14 Caufield `456 United States 12 Y Combinator 173 United States & Byers 15 Sequoia Capital 447 United States 13 GSR Ventures 155 United States Bessemer 14 Intel Capital 146 United States 16 Venture 416 United States Shunwei Partners 15 144 China 17 Bpifrance 402 France Capital Partners 18 SV Angel 398 United States 16 Wayra 143 Spain High-Tech 17 GGV Capital 135 United States 19 382 Germany Alibaba Grunderfonds 18 134 China Silicon Valley Group 20 377 United States 19 Plum Ventures 133 China Bank 20 Kickstarter 377 United States 20 Fortune Capital 132 China Private company Government or public-private partnership Source: Authors’ calculations using the FCI Digital Business Database. 59  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” In addition to big techs, states also play a role as an investor in digital busi- nesses in both developed and developing countries, but there are nuances. Historically, governments of developed countries – especially the US – have funded major technology innovations (Mazzucato 2013).31 As Figure 5.1 and Table 5.1 show, public investors (for example, through innovation funds or sovereign wealth funds) account for a larger number of the investment deals that are headquartered in developed countries; this is less so in developing countries. However, when ranking top investors in terms of the dollar volume of their investments, this observation is reversed. Public investors in developing countries rise to prominence, investing heavily in the later stages of the firm cycle with a large deal size. This is especially the case with Chinese state-owned banks.32 In contrast, public investors in devel- oped countries are not among the top 20 investors by funding volume anymore. The US and China are the prominent investors in digital businesses, though they still predominantly invest in digital businesses in their own countries, capitalizing on and driving the rapid growth of their respective digital econo- mies. The dominance of US investors in digital businesses across the world, and of Chinese investors in developing countries is undisputed, as the left two panels in Figure 5.2 reveal. These investments cover all stages of the investment cycle – seed capital, VC funding, M&A, and so on. However, most investments take place in the markets where investors have their headquarters – that is, 88 percent of investments by the top 20 US investors in developed markets are in US companies; only 12 percent are in other developed-market businesses such as those in the UK. Similarly, 97 percent of investments by the top 20 Chinese investors in developing countries are in Chinese companies, and 3 percent are in overseas firms based in other developing countries. When ranking out-of-country top investors for devel- oping country digital firms (right-most panel in Figure 5.2), US investors dominate with over two-thirds of the total share. Over time, investments in digital businesses by the US Department of Defense have been predominantly in 31  general purpose technology (GPT) and sophisticated deep tech sectors such as tech hardware (integrated photonic chips, ultrafast laser tech, silicon photonics) and software (for example, multichannel thermocouple data acquisition systems), and in applications likely to be directly used by the defense industry. Some top Chinese state investors now appear as the top investors in terms of dollar amount per deal: that is, the 32  China Development Bank, China Construction Bank, and China Integrated Circuit Industry Investment Fund. A Spiky Digital Business Landscape  60 FIGURE 5.2  Nationality of the Top 20 Investors in Developed and Developing Countries33 Nationality of top 20 Nationality of top 20 Nationality of top 20 investors investing into investors investing into out-of-country digital businesses in digital businesses in investors investing into Developed Countries Developing Countries digital businesses in Developing Countries EU Spain 10% 5% UK 5% Other ** 32% USA China * 40% 55% USA 68% USA 85% Source: Authors’ calculations using the FCI Digital Business Database. Note: The top 20 investors are those investors with the greatest number of deals in these countries: their nationalities reflect the headquarters location of the investors. * Chinese investors mainly invested in Chinese firms; therefore Chart 3 ranks the top 20 out-of-country investors only. ** “Other” here includes investors. from Hong Kong, Japan, Malaysia, Spain, Switzerland, and the United Kingdom, as well as the International Finance Corporation (IFC) of the WBG (each country has a similar percentage share). In addition to being big investors in digital businesses, M&A deal flows reveal that big tech firms from emerging markets are also prominent acquirers in developing-country acquisitions, while Silicon Valley based firms are less prominent. Table 5.2 shows the dominance of US acquirers in terms of the total number of firm acquisitions in developed countries, with Microsoft, Google, and IBM topping the list. Among acquirers of firms in developing countries, those from large emerging markets, notably the BRICs countries, and high-income European countries, stand out. In terms of the top deals by ticket size, acquirers from the US, Germany, France, Japan, and South Korea are among the top acquirers of firms in developed countries, while acquirers of firms in developing markets represent a more diverse set of countries. Thus, the aggressive flurry of acquisitions by the Silicon Valley big techs described in the news does not appear to extend to firms in developing countries. However, given the opportunities presented by digital tech- nologies to scale without mass, many established big techs are using sophisticated M&A strategies to acquire markets34 – for example, they can do so by collecting data when providing “free” services in another country, without having a physical presence. They can then engage in buyout activities to obtain control of a local start-up and monetize the data they have collected, making it even more difficult for regulators to monitor and regulate the market if the country has limited bargaining While funding information is only available per company and not per investor (since one firm can have multiple 33  investors), robustness checks were carried out using total funding divided by total number of investors to establish a lower bound estimate of the funding size of top investors. Using these investment flows (in US dollars), the top 20 investors in developed country firms are also 81 percent from the USA while the remaining amount is shared by Japan, Kuwait, UAE, and Singapore. (There are no European investors in the top 20). For developing countries, 60 percent of the top 20 investors are from China, and only 20 percent from the USA, whereas there is one investor each from Japan, Malaysia, Saudi Arabia, and Singapore in the top 20. https://www.crowell.com/NewsEvents/Publications/Articles/Antitrust-in-the-Digital-Age 34  61  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” power vis-à-vis foreign big techs. Given that M&A is a major exit strategy for digital firms, this calls for further research on whether there are any indications that digital firms in developing countries are consistently being prematurely acquired by big techs and that they are not reaching their full potential as independent firms - or that, on the contrary, they have grown to be more successful firms after being acquired, by tapping into the parent company’s networks and capabilities. TABLE 5.2  Top 20 Acquirers and the Number of Digital Businesses They Have Acquired DEVELOPED DEVELOPING Acquirer Acquirer # Rank (HQ Country) # deals Rank (HQ Country) deals 1 Microsoft (USA) 172 1 Alibaba (China) 28 Publicis Groupe 2 Google (USA) 170 2 18 (France) Dentsu Aegis Network 3 IBM (USA) 126 3 15 (UK) Tencent Holdings 4 Cisco (USA) 124 4 13 (China) 5 Accenture (Ireland) 105 4 Quikr (India) 13 6 Oracle (USA) 102 4 Bytedance (China) 13 7 Apple (USA) 83 7 Yandex (Russia) 12 8 Yahoo (USA) 79 7 Magazine Luiza (Brazil) 12 9 J2 Global (USA) 71 7 Baidu (China) 12 Vista Equity 10 69 10 Gojek (Indonesia) 11 Partners (USA) Hewlett-Packard 11 66 10 Bharti Airtel (India) 11 (USA) The Carlyle Group MercadoLibre 12 64 12 10 (USA) (Argentina) EOH Holdings (South 13 Facebook (USA) 61 12 10 Africa) 14 Amazon (USA) 60 14 Naspers (South Africa) 9 15 Autodesk (USA) 57 14 iFood (Brazil) 9 The Riverside Freshworks (United 16 56 16 8 Company (USA) States) Marlin Equity 17 54 16 Flipkart (India) 8 Partners (USA) 18 Salesforce (USA) 53 16 CarDekho (India) 8 EQT Partners 4Sight Holdings 18 53 16 8 (Sweden) (Mauritius) 19 VMware (USA) 52 20 Trimble (USA) 50 20 Thoma Bravo (USA) 50 Source: FCI Digital Business Database. Note: For developing countries there are 12 acquirers sharing the 20th rank; they were not included in this table. Target sectors of the top acquirers in both developed and developing countries reveal both horizontal (within-sector) and vertical (cross-sector) acquisitions, leading to trends of digital conglomeration. Table 5.3 lists the sectoral distri- bution of acquisitions by the top three acquirers (by number of acquisitions) in developed and developing countries. In general, the most frequent acquisitions are in the same sector as the acquirer, a strategy pursued to either achieve economies A Spiky Digital Business Landscape  62 of scale or to stifle competition and establish market dominance. At the same time, it is evident that the acquiring companies are present in several subsectors, some well outside their core sector – for example, Microsoft expanding to e-commerce and fintech, or Alibaba expanding to the business management tech, web services, and SaaS sectors – consistent with the existence of economies of scope.35 These acquisition patterns contribute to the clustering patterns of digital sectors pre- sented in Section 3. These trends pose additional challenges for regulators when evaluating potential violations on competition, since it is becoming increasingly difficult to define market boundaries.36 TABLE 5.3  Sectoral Distribution of Acquisitions by the Top 3 Acquirers (by Number of Acquisitions) TOP 3 ACQUIRERS TOP 3 ACQUIRERS BUYING BUSINESSES IN BUYING BUSINESSES IN DEVELOPED COUNTRIES DEVELOPING COUNTRIES Dentsu Publicis Microsoft Google IBM Alibaba Aegis Group Business man- Software and Software and E-commerce Marketing tech 1 agement tech Marketing tech (16) SaaS (40) SaaS (25) (14) (18) (48) Business man- Big data and Software and 2 agement tech Digital media (24) Digital media (7) Social network (6) analytics (31) SaaS (5) (35) Big data and Security tech Entertainment 3 Web services (23) Social network (5) Web services (4) analytics (23) (26) tech (5) Entertainment Marketing tech Software and Web services Business manage- 4 Digital media (3) tech (20) (21) SaaS (25) (5) ment tech (2) Entertainment Big data and Business manage- 5 Security tech (19) Fintech (23) E-commerce (2) tech (20) analytics(4) ment tech (2) Marketing tech Marketing tech 6 E-commerce (14) HR tech (14) Web services (1) Travel tech (2) (16) (4) Business man- Tech hardware Software and SaaS 7 Digital media (14) Health tech (4) E-commerce (1) agement tech (13) (13) (1) Entertainment 8 Web services (14) Security tech (13) Health tech (12) Telecom (4) Food tech (1) tech (1) Logistics tech Logistics tech Big data and 9 E-commerce (12) Fintech (10) Fintech (1) (11) (4) analytics(1) Artificial Web services Security tech 10 Fintech (10) HR tech (1) intelligence (10) (10) (3) Source: Authors’ calculations using the FCI Digital Business Database. Note: Subsector-company pairing is used in the analysis because some digital businesses can offer digital solutions in multiple sectors. Those digital businesses are counted in each of their subsectors. The evidence presented in Table 5.2 is consistent with the literature sug- gesting that a small number of big tech firms are using their dominance to entrench their market positions. Figure 5.3, based on data on completed antitrust cases across the world as of January 2020, reveals that in over 90 percent of the An analysis of the sectoral composition of firms acquired by Google, Amazon, and Facebook between 2008 35  and 2018 revealed that while all three firms acquired businesses outside of their core sector of activity, Google’s acquisitions were the most heterogenous (Argentesi, et al. 2020). Strategic partnering among global corporations is another channel for diversifying across sectors and increasing 36  revenue growth, contributing to the clustering patterns discussed in Section 3 of this report. Several deals in recent years between global digital platform corporations and multinational enterprises (MNEs) in traditional sectors signify their efforts to enter different sectors to maintain growth momentum. For example, stagnating revenue growth is presumed to be the reason behind Baidu, China’s dominant search engine, entering the self-driving vehicles market and partnering with a number of vehicle manufacturers such as Ford, Volkswagen, Toyota, and Daimler. Another example is Walmart partnering with Google to use Google Assistant, allowing Walmart to sell through a new e-commerce platform using voice computing technology. 63  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” M&As of digital platform businesses, the acquirer was a very large firm while over 50 percent of the target firms were small and medium sized. Some, though not all, of these are likely to be “killer acquisitions” and/or “acqui-hires.” FIGURE 5.3  Distribution of Firm Sizes Among Acquiring and Target Firms in M&As Involving Digital Platforms 100% 92% 80% 60% 40% 30% 27% 30% 22% 20% 14% 3% 0% 0% Small Medium Large Very Large Small Medium Large Very Large Acquiror Target Source: Nyman, Sara, and Rodrigo Barajas. 2021. While finalized antitrust cases in digital markets in recent years have dispro- portionately involved abuse of dominance by developed country firms, there are several completed and ongoing cases involving developing country firms as well. There is evidence of firms in developed countries using anticompetitive practices in overseas markets to gain market dominance. An analysis of publicly available information on 103 finalized antitrust cases across the world as of January 2020 reveals that most cases concerning the abuse of dominance and anticompetitive agreements have been filed in developing countries against firms headquartered overseas (Figure 5.4). See Box 5.1 for some examples involving the telecom and payments sectors in Africa. International cooperation to prevent such abuse is needed: for instance, there could be a coordinated effort on digital tax and data interoperability policies, and standards adoption to allow for data flows across firms, industries, and borders so that firms in developing countries have a fair chance to scale as well. FIGURE 5.4  Distribution of Antitrust Cases by Location of Headquarters All Cases Abuse of Anticompetitive Merger Dominance Agreement 10% 35% 35% 43% 45% 54% 26% 55% 65% 10% 11% 13% Domestic Foreign Foreign/Domestic Source: Nyman, Sara, and Rodrigo Barajas. 2021. A Spiky Digital Business Landscape  64 BOX 5.1 ABUSE OF DOMINANCE IN THE DIGITAL FINANCE AND TELECOM SECTOR - EXAMPLES FROM AFRICA In several developing countries, the telecommunication and banking sectors are working together to create mobile banking services for those with limited access to traditional banking. This digital finance and commerce (DFC) sector involves at least two different network industries – financial services and telecommunications, both characterized by significant fixed costs and sunk investments, and economies of scale and scope. These features make the sector more prone to market concentration and potential anti-competitive practices. In several countries, dominant mobile network operators (MNOs) in the telecom sector are now taking advantage of their market positions to undertake unilateral actions that defend/ reinforce their dominance in DFC markets - and possibly other related markets such as mobile communications - while potentially harming competitors and taking advantage of consumers. MNOs compete with banks and other mobile financial service (MFS) providers or third parties in the provision of mobile payments, but MNOs also own key communications infrastructure required to provide mobile payments. MNOs are suppliers of inputs in the form of Unstructured Supplementary Service Data (USSD), a critical piece of infrastructure used to provide MFS on nearly any phone, at low cost, and without requiring access to the user’s SIM card. USSD enables customers to send instructions to the MFS provider along with their personal identification number (PIN) for authentication, while enabling the MFS provider to send responses to clients and confirm transactions. They therefore have an incentive to limit third party access to upstream communication inputs. MNOs also compete on the downstream retail markets in DFC markets, in that they are direct competitors to banks wanting to offer mobile financial services. Thus, there are also instances of anti-competitive behavior further down the value chain with regard to APIs*, agents and customer pricing.  Dominant companies in DFC markets are also likely to leverage their dominant market power in one market (such as mobile telecommunications), through practices such as bundling or tying in ancillary markets (such as P2P transfers or mobile insurance), in order to increase their market share in the ancillary markets, thereby reinforcing network effects. Such behavior includes excessive and/ or discriminatory pricing of USSD, SMS and other communication channels. For instance, there have been complaints of excessive USSD prices and poor quality of service in Kenya, of unfair revenue share structures for USSD session fees in Uganda, and of discriminatory pricing against banks for USSD for banks in Zimbabwe. In Uganda, Ezee Money - an independent mobile money 65  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” BOX 5.1 CONT service provider - successfully sued MTN Uganda, for denying it access to its USSD gateway.  Other recent abuse of dominance issues include agent exclusivity. Airtel filed a complaint with the Competition Authority of Kenya (CAK) to force MNO Safaricom to remove the exclusive arrangements that it held with agents to allow access by rival MNOs. After hearings with the CAK and applications for judicial review at the High Court, the matter was settled by Safaricom agreeing to expunge all restrictive covenants in its M-pesa agent agreements. Kenyan-based MTO Bitpesa sued Safaricom unsuccessfully over its refusal to provide it access through an API to the M-PESA DFC platform.  * The Application Programming Interfaces (API) defines the way a developer should write a program that successfully requests services from the mobile money platform. For example, this would allow an application like Uber or Amazon to seamlessly link directly to a mobile wallet instead of a credit card or bank account. An open API allows any developer to build applications that use mobile money to facilitate payments for new types of services, in this case, using M-PESA as a payment platform. Conversely, a lack of open APIs results in corresponding inefficient duplications, with each business implementing imperfect copies of the same business logic. Source: https://www.bloomberg.com/news/articles/2022-04-08/kenyan-mobile-money- gets-boost-in-shift-to-seamless-payments;https://blog.mondato.com/dominance-competi- tion-dfc/ ITU (2019), “Competition aspects of digital financial services” (https://www.itu.int/ hub/publication/T-TUT-DFS-2019-3/)  There is suggestive evidence of a U-shaped pattern in the concentration of acquirers in M&A deals as digital markets develop, with implications for policy at both ends of the distribution. Figure 5.5 plots a proxy for concentration in acquiring markets for various regions. It is based on the top 50 acquirers in each region, ranked by the dollar volume of the deals they were involved in during 2000- 2020 against the number of digital businesses in each region during that period, which is a measure of the development of the digital business sector. The proxy measure of concentration is the Herfindahl–Hirschman Index (HHI), which is cal- culated in terms of each acquirer’s share of the total value of acquisitions (in dollar terms) of top 50 acquirers.37 Figure 5.5 reveals a U-shaped pattern; there is a high The Herfindahl-Hirschman Index (HHI) is calculated as sum of the squared shares of the M&A deal sizes of 37  the top 50 acquirers. For each acquirer in the top 50 (by their total deal size), their M&A volume is divided by the total M&A volume of the top 50 acquirers, multiplied by a hundred, squared, and added together. The HHI accordingly ranges from >0 (low concentration) to 10,000 (high concentration). A Spiky Digital Business Landscape  66 concentration in markets in the early stages of digital development (for example in Latin America and the Caribbean, and Sub-Saharan Africa). This is partly a conse- quence of very few M&A deals taking place, with only the largest digital businesses being bought by a few of the largest bidders (acquirers); and partly a consequence of the local corporate and acquisition market structures. This high concentration indicates that the least developed regions – with smaller markets and fewer players - should be particularly alert to potential anticompetitive consolidations through M&A. Even though there are fewer M&A deals occurring there than in developed regions, authorities may focus their attention and resources on assessing the merits of the relatively smaller set of deals that do occur. As the digital business sector grows, M&A activities increase with the involvement of a broader set of market players - more acquirers and acquirees, as in East Asia and the Pacific (EAP). This leads to a decrease in the HHI (Figure 5.5). However, in digitally mature regions like North America and high-income Asia, which have many more established digital businesses, acquiring market concentration has been increasing, especially since 2015, as some big techs seek to utilize M&A as an instrument to further consolidate their market leadership position via both vertical and horizontal expansions.38 While more research and data are needed to further test this hypothesis, it fits with the literature about entry for buyout, killer acquisitions, and aqui-hires (Rasmusen 1988, Cunningham, Ederer and Ma 2020, Mermelstein, et al. 2020). FIGURE 5.5  U-Shaped Hypothesis: Digital Business Development and Market Concentration (HHI of top 50 acquirer, by M&A deal size) 1500 Digital acquisitions market concentration 1000 500 0 20000 40000 60000 80000 100000 Digital Business Development (Number of digital business) Source: Authors’ calculations using the FCI Digital Business Database; number in parentheses are the number of deals by top 50 acquirers when calculating HHI. Note: The South Asia (SA), Middle East and North Africa (MENA) and Europe and Central Asia (ECA) regions are not shown separately due to only a small number of acquirers with deal information. Also, using the number of M&A deals (as opposed to the number of digital businesses) to measure digital business maturity and develop- ment produces a similar U-shaped curve. 38  The nonlinear relationship is also robust to the inclusion of the top 100 acquirers, indicating that market concentration is driven by a small number of mega deals. 67  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” The trends in market concentration highlight the need for policy makers around the world to monitor digital market concentration, while implementing policies to support entry into the market, and scaling of new businesses. Whereas economies of scale and scope are a driver of efficiency and productivity gains, especially when digital businesses are at an early stage of development, they may cause innovation-hindering effects in the longer term if left unchecked. For example, excessive consolidation in markets can lead to a greater risk of the abuse of dominance and the abuse of buyer power. Figure 5.6 shows some key policy options that can help to balance these two objectives. Authorities should pay particular attention to prospective mergers (merger reviews) when funding or acquirer markets are more concentrated (such as in less developed regions) since in such cases the potential for anticompetitive consolidation is higher. This does not mean prohibiting all mergers – it means carefully assessing whether the potential anticompetitive effects from consolidation might outweigh the potential procom- petitive effects of gaining economies of scale and scope through acquisition. At the same time, policy makers must encourage the entry and expansion of new digital businesses – that is, moving to the right along the U-shaped curve in Figure 5.5 – so that countries have more and more diverse digital firms, as a way of countering the natural market concentration at the incipient stages of development. For most developing countries, this largely means embedding pro-competition principles in digital business support, or creating industry development policies to maintain a level playing field over the medium term, and supporting digital businesses to enter the market and scale. See Box 5.2 for some policy options that can help do this. FIGURE 5.6  Policy Matrix for Digital Business Development: The Need to Balance “Efficiency-Enhancing Scope and Conglomeration” vs. the Risks of “Anticompetitive Consolidation” ENCOURAGE ENTRY AND ENSURE A LEVEL PLAYING SCALING FOR EFFICIENCY FIELD FOR CONTESTABILITY ■ Remove regulatory entry barriers ■ Update competition policies for new that prohibit digital business models, types of market dominance and exclusion; especially the entry of B2B services ensuring healthy digital-analog competition (logistics tech, travel tech, clean tech) ■ Accessing to input data via investing in ■ Build scope economies and network trusted and secure data exchange to e ects: Open API/standards adoption, promote innovative use of data by SMEs, co-creation programs, joining digital consider mandating data access if a firm single-market initiatives holds a bottleneck position (gatekeeper) ■ Attract a diverse source of funding and exit options for start-ups ■ Ensure equity and safeguard policies: Online consumer and supplier protection, data protection, intellectual property protection ■ Preventing anticompetitive mergers A Spiky Digital Business Landscape  68 BOX 5.2 EXAMPLES OF PRO-COMPETITION ACTIONS AND PRINCIPLES The following actions and principles could help countries develop their digital ecosystems through the entry and scaling of new digital businesses. This is a non-exhaustive list of examples of pro- competition principles that policy makers could follow to promote the sustainable entry into and expansion of digital businesses. The precise set of policies that might be required to boost competition will depend on the phase of development of the digital business ecosystem and the institutional setting in a given country. • Ensure that “traditional” sectoral regulations do not unduly restrict entry or create an uneven playing field for digital businesses (for example, regulation of ride hailing, accommodation, retail, finance that discriminates against digital business models, etc.). • Ensure economy-wide policies–for example regarding tax and public procurement. Do not discriminate against digital business models (for example, address policies that discriminate against services versus goods). • Direct state support to digital businesses (financial or in-kind) should be designed to be: • Well targeted to market failures, with a clear path to commercial viability/sustainability • Designed to avoid unnecessary distortions: for example, any selectivity should be merit-based, and should draw on well- defined criteria • Transparent, publicly disclosed, and well monitored • Support/ training schemes provided in partnership with incumbent digital businesses should not lock users or other digital businesses into one solution over others. • Allow for access to important input data that may be required for firms to develop and offer new services/products. • Create policies for trusted and secure data flow, data exchange, and data ecosystems, both for personal and nonpersonal data • Provide government support for data repositories, data spaces, or data exchanges (for example, regarding the setting of standards and interoperability requirements) • Provide open government data where necessary • Mandate data access where there is a clear economic case (for example, where there are very high positive externalities from data access, or where a firm holds a bottleneck position in a market) 69  Chapter 5  Big Techs & the Emergence of “Digital Conglomerates” BOX 5.2 CONT • Encourage interoperability between systems and ensure that relevant standards (for technology interoperability, data exchange, etc.) are set in a way that is fair, reasonable, and nondiscriminatory. • Facilitate industry interoperability initiatives, or consider mandating interoperability and specifying relevant standards where there is an economic rationale (for example, high externalities, natural monopoly characteristics, coordination failures); • Avoid basing standards solely on the suggestions of incumbents; • Seek the opinions of small players or potential entrants into the market when setting standards. • Encourage user switching and user choice. • Encourage or set rules on the transparency of user terms (on pricing, data collection policies, contractual terms, etc.) • Implement rules for user data portability • For digital businesses with a significant market position, discourage or consider imposing ex ante regulation that prohibits: • Exclusive contracts • Preinstallation of apps on operating systems • Tying of services provided by the digital business (for example, the provision of one service conditional on buying another service) • Where a functional competition authority exists, monitor digital markets and address anticompetitive practices by digital businesses ex post. • Combatting exclusionary conduct that can hinder the entry and expansion of newer or smaller firms is especially important, for example through refusal to deal, Most Favored Nation (MFN) clauses, exclusive contracts, self-preferencing, tying, and bundling. • Ensure a healthy mix of venture capital financing sources and exit options. • Avoid dominance of corporate/big tech VC by attracting financial VC with “smart capital” that brings in knowledge and networks. • Capital market development that enables more exit options, so M&A is not the sole solution. Source: Adapted from the World Bank Group’s Markets and Competition Policy Assessment Toolkit. A Spiky Digital Business Landscape  70 CHAPTER 6 INCLUSION DIMENSIONS: GENDER AND CLEANTECH 71  Chapter 6  Inclusion Dimensions: Gender and Cleantech CHAPTER SUMMARY ◆ In the Sub-Saharan Africa (SSA) and Middle East and North Africa (MENA) regions, where the database has gender disaggregates, the respective shares of digital businesses that have at least one woman in the core team are 19 percent and 17 percent These shares are significantly higher than the women-owned shares in the traditional sectors of these regions. Moreover, having female representation on the management team does not appear to discourage funding.​The encouragement of gender inclusion continues to be a key policy, since 19 percent means there is still plenty of room to grow, and there are more opportunities in managerial roles for women in this sector as well. ◆ Women are equally present in B2B and B2C sectors, and in both deep tech and applied tech, indicating that there is no sign of women sorting into certain digital subsectors. ◆ The clean tech sector is still relatively small compared with other verticals. Developed countries in Europe and North America have the largest number of clean tech companies per capita. However, after controlling for per capita GDP, the intensity of clean tech companies is highest in India, the US, and China​. Large clean tech firms (those with total funding of over USD 100 million) predominantly operate in the US, and are focused largely on renewable energy and other circular economy sectors such as sustainable food production and transport, recycling, energy storage, and green packaging. When using public subsidies to deploy clean tech solutions, it is important to assess their commercial viability and sustainability. The analysis of the gender dimension in this section is based solely on data from Briter Bridges, which covers digital businesses in the entire Sub-Saharan Africa and Middle East and North Africa regions.39 This is the only source that contains information on the gender composition of the management team of a digital business.40 The analysis of digital firms engaged in clean tech uses data from CB Insights, which covers all regions in the world.41 Women’s Leadership in the Digital Economy The digital sector in both the Sub-Saharan Africa and Middle East regions reveals a significantly higher share of women in leadership roles compared to the traditional economy sectors. Figure 6.1 indicates the share of digital businesses that have at least one woman in the core management team. In both regions, these shares are over two times the corresponding shares in the traditional economy, suggesting that the digital sector is more woman-friendly and inclusive. This analysis uses Briter Bridges information, which reports on the gender of the core team members. Digital 39  Businesses with Core Team Gender information = 340. The average share of female business owners are from the World Bank Gender Indicator Report, which covers 5 of the 12 Sub-Saharan African (SSA) countries with gender information in the FCI Database, and 6 of the 12 Middle East and North African countries with gender information in the FCI Database. These two regions include high-income countries such as Mauritius, Bahrain, Kuwait, Qatar, Saudi Arabia, and UAE. N of digital businesses with core (management) team gender information in the entire Middle East and Africa = 340. 40  The count of clean tech businesses shows businesses that are headquartered in the said country. Only countries with 41  five or more clean tech businesses are considered. N of countries=45, n of clean tech businesses =3,581. A Spiky Digital Business Landscape  72 FIGURE 6.1  Share of Businesses with Women in Leadership Roles in the Sub- Saharan Africa (SSA) and Middle East and North Africa (MENA) Regions (Both Developed and Developing Countries) Gender Breakdown of Digital Businesses 20% 19% 17% % of Digital Business with 15% Women in the Core Team 10% Average Share of Female 7% 7% Business Owners, Averaged 5% between 2014-2018 0% (Traditional Sectors) Entire Sub- Entire Middle East Saharan Africa and North Africa (n=148) (n=203) Source: Authors’ calculations using the FCI Digital Business Database and the World Bank Gender Indicator Report. Note: The gender information has a lot of missing values, and SSA and MENA are the only regions with this information. The other regions are excluded from this analysis due to missing values. The sectoral distribution of digital firms with women in leadership roles includes both B2B and B2C sectors in both regions. Figure A.13 and Figure A.14 in Appendix E give the distribution of subsectors among women-led businesses for the SSA and MENA regions. In terms of the number of digital firms, the top sectors include both B2B and B2C solutions and deep tech and applied tech solutions in both regions. There is no evidence of women sorting themselves into “women-ori- ented” subsectors in these regions. The Emergence of Clean Tech42 Clean tech is still a relatively small digital subsector; both income and market size are associated with the prevalence of clean tech firms across the world. This subsector is still small in terms of the number of firms, investments received, and successful exits. While developed countries in North America and Europe have the highest number of clean tech firms, the biggest concentration controlling for GDP per capita is in the US, China, and India (Figure A.15 in Appendix E). In addi- tion to these three countries, the list of the top 20 countries in this sector includes Canada (in North America), several developed countries in Western Europe, as well as a number of populous and mid-sized emerging markets from East Asia and the Pacific (EAP) and Sub-Saharan Africa (SSA). Large clean tech firms that can deploy solutions at scale are mainly focused on alternative fuels and energy sources, and circular economy sectors. The US is dominant in terms of the number of large clean tech firms – those with funding over $100 million. Figure 6.2 shows that large clean tech firms across the world are princi- pally active in renewable energy sources and circular economy sectors (sustainable This analysis uses CB Insights data which covers all regions. The count of clean tech businesses show 42  businesses that are headquartered in the said country. Only countries with 5 or more clean tech businesses reported in the dataset are included. Country N= 45. Clean tech businesses N=3581; 2019 GDP (in current US dollars) is from the World Bank Development Indicator Databank. 73  Chapter 6  Inclusion Dimensions: Gender and Cleantech food production and transport, recycling, energy storage, green packaging, etc.). This suggests that these types of clean tech solutions may be more market-ready and the technologies more ready to scale. Careful considerations are needed when deploying clean tech solutions that are still at an early stage of development, especially in relation to technical feasibility and commercial sustainability. FIGURE 6.2  Operating Countries and the Focus Areas of the 50 Largest Clean Tech Firms Large Clean Tech Firms by Operating Location FOCUS AREA #FIRMS (Firms whose total funding is over $100M) Alternative fuels / energy 13 United States 30 Sustainable food and agriculture China 6 (e.g. animal 7 United Kingdom 4 feed and crops) Canada 3 Green building / energy 7 India 3 efficiency United Arab Emirates 3 Energy storage 5 Switzerland 2 2 Consulting and advisory (e.g. Germany energy supply management, 5 South Africa 2 asset investments) Qatar 1 Recycling (e.g. management of Japan 1 4 excess and returned inventory) Singapore 1 Sustainable transport (e.g. Uganda 1 2 zero-emission vehicles) Kenya 1 Rwanda 1 Other (e.g. sustainable medical France 1 technology, air pollution 0 10 20 30 40 measurement, environmental 7 Count of Large Clean Tech Firms data collection, emissions (Total Funding over $100M) trading, green packaging) Source: Authors’ calculations using the FCI Digital Business Database. A Spiky Digital Business Landscape  74 CHAPTER 7 CONCLUSIONS AND NEXT STEPS 75  Chapter 6  Inclusion Dimensions: Gender and Cleantech The economic geography of the digital business world is uneven; it is dominat- ed by a few large economies, but some smaller countries with various income levels and in different parts of the world can are strong outliers. While there is no evidence of a clear North-South divide, developed countries and large devel- oping countries feature in the top rankings for digital business density and foreign investment in the digital sector, suggesting that population size and purchasing power are major determinants of digital business activity. However, some smaller countries, for example Kenya, Estonia, Cambodia also stand out in digital business density after controlling for their population and GDP size. This points to the exis- tence of digital growth pathways for smaller developing countries. While dominant digital subsectors across the world include both B2B and B2C applications, there is evidence of a relative lack of demand and supply for B2B solutions in particular and for deep tech solutions in developing countries. The digital economy in the developed world shows distinct characteristics of market maturity. On average, digital firms in developed countries have been in existence for several years, have received funding during all stages of the funding cycle, and have high exit rates, with notably bigger ticket sizes compared to firms in developing countries. Moreover, firms in developed countries are generally engaged in more val- ue-added (and more B2B) applications than firms in developing countries both within and across subsectors. Creating an effective regulatory framework to build investor confidence and incentivizing domestic firms to localize - build applications in local languages, suited to the local culture and digital infrastructure - is likely to boost this segment in developing countries. Regional economic arrangements are also likely to provide bigger markets and build skills through learning-by-doing and tapping into international talents. Significantly more developed country firms engaged in the deep tech sector (such as quantum tech and IoT) attract funding relative to developing country firms. A digital divide in deep tech is likely if the current trend continues, as China and to a lesser extent India receive most funding in deep tech among devel- oping countries. The lack of funding for these applications could be a consequence of market failure, as these technologies require heavy up-front investments with uncertain payoffs. While this implies the need for public funding to address the failure, “bold bets” on deep tech requires a careful consideration of supply and demand-side conditions in order to ensure a market for these solutions. There is evidence of strong “economies of scope” in both developed and devel- oping countries, related to digital spillovers arising from intangible assets. About 60 percent of digital businesses across the world offer services in more than one product market. This pattern is a consequence of re-using intangible assets and GPTs (e.g., data analytics) across multiple sectors, and the agile response from businesses looking to take advantage of these economies of scope to build network effects and lock-in users (“tipping”). Digital innovations in financial services and e-commerce, for example, are enablers for some other digital sectors and help explain the dominance of these two digital services across the world. These digital spillovers are a significant source of value creation in the new digital business models. China and India currently dominate the digital business landscape among developing countries, but several others appear poised for significant growth. While there is strong evidence of increased digital market dynamism in developing countries since 2012 - more firms entering the market, increased rates of funding, A Spiky Digital Business Landscape  76 and increased rates of exit via IPOs and M&As - these trends are currently driven by a few large emerging markets, most notably China and India. At the same time, strong investor interest in other emerging economies – South Africa, Brazil, and Russia – suggests a vast and untapped potential that is waiting to be unleashed. A regional market perspective holds promise for smaller developing countries to become members of vibrant digital single markets. A number of factors sup- port this contention: (1) As noted in Chapters 3 and 4, the digital economy is already characterized by regional “leaders” with large domestic market size, such as China in East Asia and the Pacific (EAP), India in South Asia (SA, Brazil in Latin America and the Caribbean (LAC), Egypt in the Middle East and North Africa (MENA), and Nigeria in Sub-Saharan Africa (SSA), pointing to the continued importance of geography in the digital economy; (2) A digital single market offers the scale and scope needed to attract investments that small, individual economies alone may not be able to do43; (3) Digital firms have shown the willingness to overcome barriers such as linguistic diversity in order to tap into a big but related market (in India, for example), suggesting that such structural barriers may not be binding once the potential market size crosses a certain threshold; (4) Cross-border digital services trade in recent years has grown faster than trade in goods enabled by cloud facili- ties and technologies. However, international trade barriers are limiting the sector’s potential growth.44 This is the case in the Mercosur region45, for example. Regulatory harmonization of regional trade arrangements among member states would enable online sellers to seamlessly operate across the region (Suominen 2018). Regional economic integration could also help developing countries to overcome the high-skills bias of digital business models via intangible assets transfer. Digital business models are biased towards high-skills. The economies of scope feature of digital technologies implies that intangible assets (data, capa- bilities, IP rights) need to be deployed in different adjacent sectors - e.g., AI that is moving into energy, health or transport - which in itself requires (technological) innovation and higher skill levels. Moreover, the mobility of skills towards more digital and data-driven activities is becoming increasingly important. To overcome their skills disadvantage, in addition to investing heavily in human capital, develop- ing countries can consider tapping in the skillsets and capabilities in major regional hubs to enable intangible assets transfer (IP rights, capabilities). Across the world, platform and data-driven businesses are leveraging network effects by tapping into cross-sectoral complementarities, especially in the e-commerce and fintech sectors. These trends bode well for developing countries, since platforms have the potential for significant value creation. In many developing countries, global platforms are either not present, or they are operating alongside domestic or regional platforms (Jumia in Africa, MercadoLibre in Latin America, Lazada (now owned by Alibaba in Southeast Asia), and Flipkart (now owned by Walmart in See also the forthcoming World Bank flagship report “Digitalizing Africa: Investment opportunities, economic 43  potential, and policy needs for private sector development” https://www.weforum.org/agenda/2020/06/trade-in-digital-services-is-booming-here-s-how-we-can-unleash- 44  its-full-potential/ Mercosur is the Southern Common Market (MERCOSUR in Spanish initials) comprised of state parties Argentina, 45  Brazil, Paraguay, Uruguay, and Venezuela; and associated states Bolivia, Chile, Colombia, Ecuador, Guyana, Peru, and Surinam. Mercosur is an economic bloc that integrates national economies into the international markets. (https://www.mercosur.int/en/about-mercosur/mercosur-in-brief/) 77  Chapter 6  Inclusion Dimensions: Gender and Cleantech India). Low barriers to digital platform adoption provide opportunities for small firms to also become part of a platform economy, as third-party sellers. Platform firms are also compilers and users of big data, which is being used to develop digital intelligence regarding user behavior and preferences. The data may also be sold to third parties. The nature of digital technologies is creating a strong momentum for conglomer- ation, favoring established incumbents, and concentrating market power in the hands of a few major platforms, including those headquartered in large emerging economies. Platforms have used their intermediary role to take over firms in other “adjacent” subsectors to become multisided digital platforms. One reason for this is weak intellectual property rights, for example in China. This means that companies cannot rely on property rights to build market share protection from competition, and hence must use vertical integration as the means to do so. In the US, however, profit motives drive such behavior. For example, Facebook is spending up to $1 billion on original content in the form of TV shows. Platforms are also extending their activities into nondigital industries as they become increasingly digitized: Google’s and Tencent’s ventures into self-driving cars and Alibaba’s spread into convenience stores are examples of this. These expansions are driven not by traditional horizontal or vertical merger rationales, but by economies of scope arising from data. The rise of artificial intelligence (AI) is reinforcing this trend, allowing companies to expand into new industries. For instance, companies specializing in AI are moving into industries such as energy, health care, and transportation, which are much larger than the advertising industry. Another trend that has emerged is that of strategic partnerships between multinational enterprises (MNEs) in traditional sectors and global digital platform corporations, with the aim of leveraging key advanced technology platforms (for example, AI and IoT), and horizontal digital competencies (for example, voice AI and motion control expertise) across sectors (UNCTAD 2019). In terms of M&A activities, there is evidence of strong interest by acquirers, especially big techs from the BRICS and European countries, in the digital businesses of developing countries. In developed countries, the top acquirers of high-value M&A deals are from the US and the EU. In developing countries the acquirers are less geographically concentrated, though China is the most prominent among them. Moreover, acquirers are targeting more B2C sectors in developing countries, presumably driven by market access and the rapid growth of emerging B2C applications.​Across both groups, horizontal acquisitions (acquisitions within the same subsector) are more common, through there are several cross-sectoral acqui- sitions as well, especially by big techs. There is evidence of a u-shaped pattern in the concentration of acquirers in the highest-valued acquisition deals; concentration is high in the early stages of digital development, then decreases as digital businesses develop, and tends to increase again in mature markets. This potentially signals the need to sequence regulations as digital businesses develop, balancing efficiency-en- hancing consolidation versus anticompetitive behaviors. There is evidence suggesting that the digital economy may be more gender-in- clusive, compared to the traditional economy. In the SSA and MENA regions, the share of digital firms that have women in management positions is over two times the corresponding share in the traditional economy. Another encouraging detail is that there is no evidence of women sorting into specific digital sectors; digital firms with some female leadership are present in both technology-heavy sectors like big A Spiky Digital Business Landscape  78 data and AI, as well as in more applied sectors like edtech and entertainment tech. In other words, overall the digital sector remains a key venue for gender inclusion. The intersection between climate change and digital transformation shows early signs of success in terms of market-ready clean tech solutions. Digital technologies have the potential to significantly reduce carbon emissions.46 The report finds evidence that some clean tech firms are engaged in deploying such solutions. Clean tech companies are prevalent in several high-income countries in Europe and North America, as well as in large emerging markets like China and India. These firms are predominantly focused on renewable energy solutions and other “circular economy” sectors such as sustainable food production and trans- port, recycling, energy storage, and green packaging. Areas for Further Study It is important to leverage this FCI Digital Business Database and merge it with others in order to better understand the drivers of the digital business ecosys- tem. It is possible to create a cross-country data set using the information currently available in the database and employing other data sources to gather information on several other factors for all of the countries: for example, the data infrastructure environment; the broader infrastructure environment (electricity, logistics); indicators of skills and human capital; indicators of regional integration and/or proximity to a cluster, and so on. Such a data set can then be used to analyze the digital market dynamics (digital business intensity, economies of scale and scope), taking advantage of country-subnational variations to get better insights into some of the drivers of the local digital business ecosystem. Appendix D also shows some preliminary results on the relationship between data regulations and digital business growth, leveraging data from the World Development Report: Data for Better Lives (World Bank 2021). Analyzing balance sheet data of digital firms is likely to reveal insights into the impact of GPTs on technology adoption and the demand for digital services, especially in the B2B markets. The database has basic balance sheet data for a subsample of firms over time. A deeper analysis of firm revenues and profits would offer some understanding of the demand for a firm’s product(s), as well as of the technology product development and deployment timeline, and the speed of diffusion. This report found evidence of changing digital subsector cluster patterns and a significant increase in the number of digital businesses in both developing and developed countries around 2015, which was the inflection point for cloud computing going mainstream. ​ Cloud computing and other SaaS systems such as customer relations management (CRM), with their pay-as-you-go payment model based on the usage or subscription model, are especially attractive for MSMEs, allowing them to adopt modular and cost-effective standard solutions without incurring heavy capital investments in IT infrastructure. Triangulating the balance sheet data with information on the diffusion of relevant GPTs and other develop- ments would provide useful insights relating to tech adoption and the use of digital apps. Such analyses could potentially enable a deeper understanding of the factors underlying the relatively lower demand for B2B solutions in developing countries. https://www.ey.com/en_ch/decarbonization/how-digitization-acts-as-a-driver-of-decarbonization 46   79  Chapter 6  Inclusion Dimensions: Gender and Cleantech The economics of digital business models are still not well understood, especially in the context of economies of scope and digital single market initiatives, challenging efforts to develop competition policy and introduce regulatory reform around digital market contestability. This report uses M&A deal flows to proxy digital market concentration. Going forward, it will be important to enlarge this database by tapping into other M&A-focus data sets like Thomas Reuters to understand what happens to digital firms once they are acquired by big techs over time. Do they still experience growth, or they are prematurely “killed”? Competition policies regulating platform ecosystems are complex; while Microsoft and Apple are competitors at some level and have very high market valuations, they follow very different business models, rendering competition policies that are developed without considering their business models to be obsolete.47 Similarly, privacy and data protection have welfare implications for society, but also welfare distribution implications for different user groups and platform operators, and they are not static. There is so far very little empirical evidence that can clarify trade-offs in private behavior and in policy making.48 Industry observers and experts caution against making inferences regarding market concentration based on metrics developed for the traditional economy (OECD 2021). With regard to dominant platforms, several economists contend that requiring (horizontal) data interopera- bility can address concerns about market dominance more effectively than calling for a break-up of these firms, which is likely to scuttle the efficiency benefits of consolidation (Crémer, de Montjoye and Schweitzer 2019). Thus, further research is needed to understand the dynamics of competition and welfare in digital markets including who are likely to be the winners and losers, especially those concerning gatekeeper platforms that have their complex technologies and business models intertwined, and can operate without a physical presence in a country. Finally, is there a “minimum package” needed in order for e-commerce to flourish? One key finding of this report is that e-commerce and fintech almost always signal the takeoff of growth in the digital economy, in both developed and developing countries. However, these two subsectors also tend to form mega clusters/ecosystems in order for their services to become “usable” (not merely accessible). The sectoral clustering patterns discussed in Chapter 3 of this report revealed that in both developed and developing countries, e-commerce firms tend to provide services in agtech, food tech and logistics tech; fintech firms also pro- vide services in blockchain, insurance tech, and gig job platforms. This is consistent with both anecdotal examples of successful e-commerce and fintech platforms worldwide as well as detailed research analyses by FAO and UNCTAD among others. This poses challenges for policy makers. On the one hand, it seems that an e-commerce firm with a “minimum package” of digital services is needed in order for the platform service to be “usable” and to achieve efficiency gains, especially given the lack of complementary infrastructure in developing countries; on the other hand, this poses antitrust and other ex-ante sectoral regulation challenges. For example, how “minimum” is minimum? Is it actually anticompetitive behavior in disguise? And are policy makers able to distinguish between the two? 47  https://www.investopedia.com/articles/markets/111015/apple-vs-microsoft-vs-google-how-their-business- models-compare.asp https://ec.europa.eu/jrc/sites/default/files/JRC101501.pdf 48   A Spiky Digital Business Landscape  80 APPENDIX 81  Appendix APPENDIX A: METHODOLOGIES USED Methodology Used to Create a Common Dictionary for the Database How to distinguish digital from digitalized business 1. Defining digital businesses. The data sources included digital solution busi- nesses (narrowly defined) as well as digitalized traditional businesses that are of interest to investors and businesses. Since the FCI Digital business Dataset focuses on digital solutions, an additional filtering method was needed, although there is no clean-cut definition of digital firms vs. digitalized firms. This database uses the following definition for digital vs. digitalized businesses: • Digital business: Digital solution providers that develop and manufacture digital technology products, or digital services in the narrow scope of the digital economy, or in the core digital (IT/ICT) sector. • Digitalized traditional business: Traditional businesses that use digital technologies in the inputs and outputs of their business models. Most data- source businesses founded before 1970 were in this category and are therefore excluded from the FCI Digital Business Database. 2. Selecting “digital keywords.” To differentiate digital business vs. digitalized traditional business, to the best of our ability, we used a list of digital solution-re- lated keywords in the company descriptions to filter them. Validation: A manual check of a random subsample (n=50 for each keyword) to select keywords that result in at least 80 percent of the sample being digital businesses. If 80 percent or more of the 50 firms were accurate, we kept the digital keyword. If less than 80 percent were accurate, we dropped the digital keyword. 3. Identifying digital businesses using “digital keywords.” All of the companies from the data sources have been filtered so that only the companies that include one of digital keywords in the company description were retained for the database. Applying these keyword filters resulted in at least 80 percent of the sample being digital solution firms compared with manual checks of a random subsample. • Digital keywords: 3d, 3-d, algorithm, Aggregat, analy, app, apps, application, artificial intelligence, automat, augment, blockchain, broadband, cloud, computing, crypto, cyber, data, digital, drone, e-, ICT, information technology, internet, internet of thing, IoT, lidar, machine learning, mobile, online, quantum, A Spiky Digital Business Landscape  82 real time, real-time, remote, robot, satellite, Saas, search engine, smart, social media, software, streaming, technology, telecom, virtual, wearable, web • Harmonization of data source industry classification into one of 44 subsectors: All of the firm-level data comes with at least one industry classification drawn from the sources. Many companies have more than one industry classification. A harmonized list of subsectors was needed to merge the different data sources. Example below: Big Data FIGURE A.1  Example of Harmonization of Data Source Industry Classification: Big Data • Merging digital businesses from data sources: The firm-level data has been merged from three sources using a common identifier variable (company website information) to avoid duplicates. If a firm has information from multiple data sources, the earlier founding year, the larger total funding amount, and the latest funding information were selected. For example, if a firm with data from two sources is reported to have raised funds in 2019 in one data source, and it says 2020 in the other data, the data set tags fundraising information from 2020 as the latest funding information. If a firm has two conflicting bits of information on operating status (operating or bankrupt/not operating), the bankruptcy status is selected for the database with the assumption that a data source reporting the bankruptcy of a firm is more updated than another data source reporting that the same firm is operating. 4. Accuracy test of the harmonized subsectors: Manual checks of a random subsample (n=50 for each harmonized subsector) was conducted to determine if the harmonization was accurate. If 80 percent or more of the 50 firms were accurate, we kept the harmonized subsector. If less than 80 percent were accu- rate, we dropped the harmonized subsector 5. Identify subsectors through keyword screening of firm description: To have a definition of the subsectors to the best of our ability we used a list of subsector related keywords. We conducted a manual check of a random subsample (n=50 for each keyword) to select keywords that resulted in at least 80 percent of the sample offering digital solutions in that subsector. 83  Appendix • Keywords to identify firms in the sector (Example: Big Data): big data, bigdata, daas, data acquisition, data aggregat, data analy, data as a service, data base, data center, data centre, data collection, data infrastructure, data insight, data integration, data maintenance, data management, data mining, data monitor, data operation, data prep, data process, data scien, data service, data set, data solution, data structure, data tech, data visualisation, data visualization, data-analy, data-as-a-service, database, database maintenance, database monitor, datacenter, data-center, datacentre, data-centre, data-driv- en, dataset, distributed transaction, enterprise data, historical data, information visualization, integrate data, intelligent analysis, master data, predictive analy, real time data, realtime data, real-time data, streaming data, visualize data 6. Future uses for the keywords: The next phases of data updates will use machine learning methods to identify and label subsectors. Methodology Used to Identify Platform-Based and Data-Driven Businesses 1. Defining platform-based and data-driven businesses: Identifying plat- form-based and data-driven business models among digital businesses helps better understand the role of data in digital business models and its effect on market dynamics. This database uses the following definitions. Platform-based: Firms that facilitate interactions for a large number of participants. Platform-based business does not own the means of production, but rather creates and facilitates the means of connection. The role of the platform business is to provide a gover- nance structure and a set of standards and protocols that facilitate interactions at scale so that network effects can be unleashed (Deloitte 2020). Data-driven: Firms that systematically and methodically collect or aggregate large data sets and use advanced analytics (such as artificial intelligence, big data, and blockchain) to create value, leveraging data as a key element of their business model (Hartmann, et al. 2014). Data-driven businesses can also help traditional industries upgrade through servicification to optimize production processes, increase sales, streamline decision making, and even rethink revenue models. 2. Selecting platform-based and data-driven keywords: To identify plat- form-based and data-driven businesses to the best of our ability, we used a list of business model-related keywords in company descriptions as a filter. A manual check of a random subsample (n=50 for each keyword) to select keywords that resulted in at least 80 percent of the sample having a platform-based or a data-driven business model. If 80 percent or more of the 50 firms were accurate, we kept the identifier keyword. If less than 80 percent were accurate, we dropped the identifier keyword. 3. Assigning platform-based and data-driven businesses using the keywords: Digital businesses with one of the platform-based or data-driven keywords in the company description were assigned as platform-based or data-driven businesses (or both). Applying these keyword filters resulted in at least 80 A Spiky Digital Business Landscape  84 percent of the sample being digital solution firms compared with manual checks of a random subsample. • Platform-based keywords: booking, carpool, classified, compar, crowdfund, crowdsource, hail, job board, job search, job site, job website, marketplace, peer to peer, peer-to-peer, sharing platform, platform as a service, platform-as- a-service • Data-driven keywords: algorithm, artificial intelligence, big data, data as a service, data science, data-as-a-service, database, deep learning, deep-learn- ing, geolocation, machine learning, predict, real time, real-time, AI, ML, data analy, data storage, data visualiz, predictive analy 4. Future uses of the keywords: The next phases of data updates will use machine learning methods to assign platform and data-driven business models. Amount of Firm-Level Data Drawn from Three Data Sources FIGURE A.2  Number of Firm-Level Data from Three Data Sources FIGURE A.3  Platform-based and Data-Driven Businesses Other Digital Businesses 84% 16% Platform or Data-Driven Digital Businesses 85  Appendix APPENDIX B: DEFINITIONS OF DIGITAL SUBSECTORS SUBSECTOR DEFINITION 3D printing Developing and using 3D Printing, or additive manufacturing, refers to the manufacturing process and the technology related to printing a three- dimensional object. This sector encompasses the actual printer as well as software related to 3D printing. Aerospac tech Developing and using technology to provides services, research and innovation related to spaceflight, aviation, satellite and space exploration. This subsector includes but is not limited to satellite operations management software, products enabled by satellite connection (such as real-time aerial mapping), spacecraft and aircraft development software and spatial communication technology. Agtech Developing and using digital technologies to enable the agriculture technology value chain - including but not limited to digital agriculture software and hardware (sensors, imagery, precision ag), mixed and integrated agriculture innovation, plat and crop science, animal and livestock science, post-farm agriculture value chain (agri marketplace, delivery, logistics, supply chain innovation), and agriculture waste management. Artificial Developing and using technology for machines to autonomously learn and act intelligence through data analytics. This sector will inevitably be closely related to Big data & machine and analytics since AI and ML utilizes a large quantity of data to perform its given learning functions. Big data Developing and using technology for recording, collection, distribution, and usage and analytics of large volumne of data. Big data refers to data that is too large, fast or complex that it s difficult to process using traditional methods. This sector includes firms that use data as a service, data analysis and visualization services and data collection services Biotech Developing and using biotechnology to create products that are dependent upon developing and creating new products by utilizing and manipulating biological systems and living organisms. This subsector includes firms developing database for biotech research and IoT device for biotech. Blockchain Developing and using technology to use blockchain applications and the and crypto- distributed ledger technology. This subsector includes but is not limited to firms currency using smart contracts, crowd funding, supply chain auditing, cryptocurrency, identity management, intellectual property and file storage etc. Cryptocurrency space includes companies providing services or developing technology related to the exchange, storage, facilitation of payments, and securing cryptocurrency. Business Developing and using technology to improve business operations. This subsector management includes but is not limited to operations management/optimization software, tech customer relations management (CRM), customer service tools, enterprise resource planning (ERP) products, and corporate digitization consulting. Civic tech Developing and using technology to improve and aid the relationship between civil society, governmental functions and humanitarian well-being. This subsector includes but is not limited to government management system, data analytics on political and governance process, taxation managment, civil society reporting system, and monitoring products and services A Spiky Digital Business Landscape  86 Clean tech Developing and using technology to improve the creation, distribution, usage and monitoring clean and sustainability products and services. This subsector includes but is not limited to digitally-enabled clean energy products and services, sustainable product e-Commerce, clean technology logistics technology and recycle and waste management technology Construction Developing and using technology to improve construction value chain. This tech subsector includes but is not limited to construction operation management software, construction safety IoT services, and construction logistics software Digital media Developing and using technology to improve the creation, editing, storage, access, distribution, publishing analysis and delivery of media on digital settings. This subsector includes but is not exclusive of digital journalism, social media, e-media searching and subscription platforms, and publishing logistics management products and services Drones Developing the technology, utilizing, servicing and delivering automated or remote-controlled mechanical devices and technology, including unmanned aerial vehicles, subsea vehicles and land vehicles. E-commerce Developing and using digital technology to facilitate and improve the sale of products over internet networks. BEA considers e‐commerce to include digitally‐ ordered, digitally‐delivered, or platform‐enabled transactions (BEA, Barefoot et al 2018). This subsector includes but is not limited to online marketplace, aggregator e-Commerce, e-Commerce analytics, e-Commerce transaction, e-Commerce logistics Edtech Developing and using technology to enhance teaching, learning and training process in and outside of classrooms. This subsector includes but is not limited to learning devices (tablets and interactive "smart" boards"), educational institution management systems, virtual learning products and services, remote learning products and services, and instructor and student assistance program Entertainment Developing and using technology to improve the creation, distribution, delivery, tech analysis and usage of entertainment products and services. This subsector includes but is not limited to e-sports, e-casino, movies, animation studios and gaming (hardware and software) products, music and video streaming platforms and services, arts, music algorithm software, and entertainment event online management and entertainment oriented social media Fintech Developing and using technology for financial services usually offered by traditional banks including loans, payments, wealth and investment management as well as software providers automating financial processes or addressing core business needs of financial firms Food tech Developing and using technology to improve food and beverage production, distribution, purchasing and consumption. This subsector includes but is not limited to restaurant aggregator/ review platform, food e-marketplace, food lifestyle media as well as pre-packaged food subscription firms Gig economy Developing and using technology to connect gig economy workers to gig economy opportunities including different sharing economy opportunities. This subsector includes but is not limited to freelancer/gig worker hiring platform, gig worker workflow management software, gig worker insurance platforms. Health tech Developing and using technology to improve the creation, facilitation, delivery, safety, reliability and analysis of healthcare services. This subsector includes but is not limited to telehealth, e-health platforms, pharmatech, technical medical device development, medical laboratory management, and diagnostic algorithm development. 87  Appendix HR tech Developing and using technology to improve the management, research, analysis and organization of human resource functions. This subsector includes but is not limited to human resource management software/ platform, recruitment algorithm, job posting platforms, employee performance and time tracking, employee training (remote and/or virtual) and reporting tools Insurance Developing and using technology to improve the creation, distribution, delivery, tech usage and analysis of insurance products and services Internet of Developing, producing and using Internet of Things (IoT) devices - physical things objects that are embedded with sensors that monitors, stores and sends data for a use in the physical space Legal tech Developing and using technology to improve creating, distributing, using, interpreting, organizing and assessing legal products and services. This subsector includes but is not limited to tele-legal service, legal service aggregator, algorithmic legal service and caseload management solutions Logistics tech Developing and using technology to improve the movement of goods. This subsector includes but is not limited to digital supply chain management, cargo management software, supply chain tracking and operation management software Manufactur- Developing and using technology to improve the operation and management ing tech of the manufacturing value chain. This subsector includes but is not limited to automation solutions, smart factory products and data-based production analytics tools Marketing Developing and using technology to improve the marketing value chain. This tech subsector includes but is not limited to digital marketing content creation, digital marketing consultancy, marketing data and analytics, search engine optimization (SEO) technology and customer tracking and interaction products and services Mining tech Developing and using technology to improve the mining value chain. This subsector includes but is not limited to seismic data analytics, mining operation optimization, supply chain management software and risk detection technologies Mobility tech Developing and using technology to improve the movement of people. This subsector includes but is not limited to passenger transportation (air travel, train, automobile) logistics, traffic monitoring and tracking, on-demand ride share and haul (both for motorized and non-motorized means of transportation), passenger transportation repair platforms and online maps Nanotech Developing and using nanotechnology to create products that are dependent upon the ability to manipulate materials at an atomic level, usually due to the materials exhibiting novel properties at the sub-atomic level Pet tech Developing and using technology to improve products and services regarding animal and pet care. This subsector includes but is not limited to animal care matching platforms, tele-vet care, animal product e-Commerce, animal monitoring IoT and wearables and animal care social media Property tech Developing and using technology to improve the real estate and property development value chain. This subsector includes but is not limited to property sale and renting platforms, property management software, renter verification software, and smart home applications Quantum tech Developing and using digital technology through quantum computing principals (using Qubits instead of normal computer bits of either 0 and 1). This subsector includes hardware and software components of quantum computing Reality tech Developing and using technology that provides user experience in a different reality environment. This includes both virtual and augmented reality A Spiky Digital Business Landscape  88 Robotics Developing and using technology for remote-controlled mechanical devices including machineries programmed to perform repetitive tasks and precision tasks Security tech Developing and using technology to improve safety and security products and services. This subsector includes but is not limited to cybersecurity-related products and services, security monitor and security IoTs, and wearables Social Developing and using technology to eables users to connect and communicate network with each others by posting information, comments, messages, images through a dedicated website or applications. This subsector includes social media, messaging platforms, services conducted through social media, and content sharing platforms Software and Developing and using technology to offer software as a service (SaaS) or product. SaaS This subsector includes but is not limited to digital infrastructure software, application and web design/coding, industry specific software etc Tech Producing or contributing to the process of producing physical parts of computer, hardware machinery and related devices that enable digital infrastructure and digital usage. Includes firms making or servicing internal and external hardware for devices that enable digital connectivity and software installment Telecom Developing and deploying telecommunication technology to enable digital infrastructure and digital connectivity. This subsector includes but is not limited to telecommunication service providers, telecom infrastructure developers (tech hardware related to broadband, fiber optics), internet connectivity services (internet and mobile network service) for both individual consumers and businesses Travel tech Developing and using technology to improve the travel and tourism value chain. This subsector includes but is not limited to travel booking platforms, travel review and discovery platforms, and travel security software Utilities tech Developing and using technology to improve the utility value chain including water and waste management utility. This subsector includes but is not limited to utility management software, utilities monitoring and tracking services, mobile payment for utilities, leak detection IoTs, technology-enabled toilets, sanitation IoTs, sanitation monitoring tools, and sanitation-related tele-health products and services Wearables Developing and using wearable devices with sensors that collects and analyzes data based on the user's activities. This subsector includes firms developing soft and hardware related to wearable technology Web services Developing and using technologies to connect users to access web-based application and data source via standard web protocol. This subsector includes but is not limited to hosting services, cloud services, web and application development, web application engineering and ICT connectivity solution providers 89  Appendix APPENDIX C: COUNTRIES IN INCOME AND REGIONAL GROUPS IN THE DATABASE DEVELOPED BUSINESS N COUNTRY N INCOME GNI PER COUNTRY DIGITAL GROUP CAPITA Andorra, Antigua and Barbuda, Australia, Austria, The Bahamas, Bahrain, Barbados, Belgium, Bermuda, British Virgin Islands, Brunei Darussalam, Canada, Cayman Islands, Chile, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Faroe Islands, Finland, France, French Polynesia, Germany, Gibraltar, Greece, Greenland, Guam, Hong Developed Kong SAR, China, Hungary, Iceland, Ireland, Israel, Italy, High Japan, Republic of Korea, Kuwait, Latvia, Liechtenstein, $12,536 156,376 72 Income Lithuania, Luxembourg, Macao SAR, China, Mauritius, Monaco, Netherlands, New Zealand, Norway, Oman, Palau, Panama, Poland, Portugal, Puerto Rico, Qatar, Ro- mania, San Marino, Saudi Arabia, Seychelles, Singapore, Slovak Republic, Slovenia, Spain, St. Kitts and Nevis, Sweden, Switzerland, Taiwan, China, Trinidad and Tobago, United Arab Emirates, United Kingdom, United States, Uruguay, US Virgin Islands Albania, Argentina, Armenia, Azerbaijan, Belarus, Belize, Bosnia and Herzegovina, Botswana, Brazil, Bulgaria, China, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, Fiji, Gabon, Georgia, Grenada, Guate- Upper mala, Guyana, Indonesia, Islamic Republic of Iran, Islamic $4,046 - Middle 25,663 50 Rep., Iraq, Jamaica, Jordan, Kazakhstan, Kosovo, Leba- $12,535 non, Libya, Malaysia, Maldives, Marshall Islands, Mexico, Income Montenegro, Namibia, North Macedonia, Paraguay, Peru, Russian Federation, Serbia, South Africa, St. Lucia, St. Vincent and the Grenadines, Suriname, Thailand, Turkey, Bolivian Republic of Venezuela Developing Algeria, Angola, Bangladesh, Benin, Bhutan, Bolivia, Cambodia, Cameroon, Comoros, Cote d'Ivoire, Djibouti, Arab Republic of Egypt, Arab Rep., El Salvador, Eswatini, Lower Ghana, Honduras, India, Kenya, Kyrgyz Republic, Lao $1,036 - Middle 11,825 44 PDR, Lesotho, Mauritania, Moldova, Mongolia, Morocco, $4,045 Myanmar, Nepal, Nicaragua, Nigeria, Pakistan, Papua Income New Guinea, Philippines, Senegal, Solomon Islands, Sri Lanka, Tanzania, Tunisia, Ukraine, Uzbekistan, Vanuatu, Vietnam, West Bank and Gaza, Zambia, Zimbabwe Afghanistan, Burkina Faso, Burundi, Central African Republic, Chad, Democratic Republic of Congo, Ethio- Low pia, The Gambia, Guinea, Guinea-Bissau, Haiti, Liberia, $1,035 444 26 Income Madagascar, Malawi, Mali, Mozambique, Niger, Rwanda, Sierra Leone, Somalia, Sudan, Syrian Arab Republic, Tajikistan, Togo, Uganda, Republic of Yemen. A Spiky Digital Business Landscape  90 The regions in the FCI Digital Business Database are based on the World Bank List of Economies (June 2020) and are divided as follows: BUSINESS N COUNTRY N DIGITAL REGION COUNTRY LIST Cambodia, China, Fiji, Indonesia, Lao PDR, Malaysia, Marshall East Asia and 18,159 15 Islands, Mongolia, Myanmar, Papua New Guinea, Philippines, Pacific (EAP) Solomon Islands, Thailand, Vanuatu, Vietnam Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Europe & Bulgaria, Georgia, Kazakhstan, Kosovo, Kyrgyz Republic, Moldo- Central Asia 2,586 19 va, Montenegro, North Macedonia, Russian Federation, Serbia, (ECA) Tajikistan, Turkey, Ukraine, Uzbekistan Argentina, Belize, Bolivia, Brazil, Colombia, Costa Rica, Cuba, Latin America & Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, the Caribbean 3,318 25 Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicara- (LAC) gua, Paraguay, Peru, St. Lucia, St. Vincent and the Grenadines, Suriname, Bolivian Republico of Venezuela Middle East & Algeria, Djibouti, Arab Republic of Egypt, Islamic Republic of North Africa 2,773 13 Iran, Iraq, Jordan, Lebanon, Libya, Morocco, Syrian Arab Repub- (MENA) lic, Tunisia, West Bank and Gaza, Republic of Yemen. Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Paki- South Asia (SA) 6,643 8 stan, Sri Lanka Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Comoros, Democratic Repub- lic of Congo, Cote d'Ivoire, Eswatini, Ethiopia, Gabon, Gambia, Sub-Saharan 4,453 40 Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Mada- Africa (SSA) gascar, Malawi, Mali, Mauritania, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, Somalia, South Africa, Sudan, Tanzania, Togo, Uganda, Zambia, Zimbabwe Andorra, Austria, Belgium, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Faroe Islands, Finland, France, Germany, Gibraltar, Greece, Greenland, Hungary, Iceland, Ireland, Italy, High-Income 47,016 36 Latvia, Liechtenstein, Lithuania, Luxembourg, Monaco, Nether- Europe lands, Norway, Poland, Portugal, Romania, San Marino, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United King- dom Australia, Brunei Darussalam, French Polynesia, Guam, Hong High-Income 15,633 12 Kong SAR, China, Japan, Republic of Korea, Macao SAR, China, Asia New Zealand, Palau, Singapore, Taiwan, China North America 87,861 3 Bermuda, Canada, United States Antigua and Barbuda, The Bahamas, Barbados, British Virgin High-Income 651 12 Islands, Cayman Islands, Chile, Panama, Puerto Rico, St. Kitts LAC and Nevis, Trinidad and Tobago, Uruguay, US Virgin Islands High-Income Bahrain, Israel, Kuwait, Oman, Qatar, Saudi Arabia, United Arab 5,103 7 MENA Emirates High-Income 112 2 Mauritius, Seychelles SSA 91  Appendix APPENDIX D: PRELIMINARY RESULTS: DATA REGULATIONS AND DIGITAL BUSINESS RELATIONSHIP The World Bank’s Data Regulation Diagnostic covers seven dimensions of data regulations which can be grouped into two basic types: (i) those enabling the use of e-commerce/e-transactions, and public and private intent data; and (ii) those safeguarding personal data, nonpersonal data and cross-border data flows, and governing cybersecurity.49 Each country is conferred a score for each of these seven dimensions. The principal component analysis (PCA) of these seven indicators of digital regulations reveals three dominant components, comprised of i) private intent data, personal data, cybersecurity, and cross-border data flows; ii) e-commerce transactions, public intent data and private data (which is included in both components); and iii) nonpersonal data.50 Table A.1 presents the results from a cross-country correlation analysis of these three principal components of the seven regulatory dimensions, with three indicators of digital business dynamism: digital business intensity; longevity; and the exit rates of digital businesses.51 All of the estimated relationships control for the country’s GDP, population, and Ease of Doing Business Score.52 These correlations reveal that regulations concerning enablers (PC2) are positively correlated with the digital business intensity of a country; but they do not appear to have a significant effect on their longevity and exit rates. The The definition of data regulation framework and indicators of digital regulations are from the World Development 49  Report 2021: Data for Better Lives (WBG 2021) and Mapping Data Governance Legal Frameworks around the World: Findings from the Global Data Regulations Diagnostic (Chen 2021). Safeguards promote trust in data transactions by avoiding or limiting harm arising from the misue of data. Enablers facilitate access to and reuse of data within and among stakeholder groups to ensure that the full social and economic value of data can be captured. E-commerce/e-transactions: transactions occurring online; public intent data: data collected for public purposes; private intent data: data collected by the private sector as part of routine business processes; personal data: data that identify the individual; non-personal data: data that do not contain any personally identifiable information; cross-border data: movement of data outside of domestic jurisdiction; cybersecurity: security measures to tackle cybercrimes. All three principal components have eigenvalues of at least 1, and together explain 76 percent of the variation of 50  the seven indicators (Kaiser-Meyer-Olkin value of 0.75). One reason for using a principal components approach is that the assignment of unitary weight to all dimensions 51  for the construction of a “regulatory index” is somewhat arbitrary. In addition, it is likely that there is high collinearity between some of these components, which would introduce multicollinearity in any regression analysis. The regulatory framework governing digital data has to balance the need for creating an enabling environment 52  to promote and facilitate investment in the digital industry while also achieving public policy objectives such as consumer protection, data privacy protection, and cybersecurity. These regulations are part of the broader regulatory framework, governing international trade in goods and services. The World Development Report for 2021 (WBG 2021) finds a poor regulatory environment in lower-income countries, with critical gaps in data safeguards as well as constraints on data sharing and interoperability. A Spiky Digital Business Landscape  92 association with digital business intensity also does not appear to vary by income level (that is, the interaction term of PC2 is not significant). These results are robust to the inclusion of the country’s population and the Ease of Doing Business Score.53 Regulations concerning data safeguards (PC1) are negatively associated with all three outcome variables (that is, the estimates are all negative and significant) but they have a positive association - with all the dependent variables - for richer coun- tries (that is, the interaction effect of PC1 and log(GDP/capita) is positive and signif- icant), suggesting that data regulation hurts digital businesses more in developing countries. This effect is robust to the addition of controls for population and the Ease of Doing Business Score. The finding that data regulations negatively impact digital businesses is consistent with some findings; there is evidence showing that data regulations have a detrimental impact on the profitability of, and investment in digital businesses (Goldfarb, Greenstein and Tucker 2015); (Jia, Jin and Wagman 2021). However, the evidence is not unequivocal.54 Nonpersonal data regulation, such as industry data policy (PC3) is positively associated with the longevity of digital businesses in high-income countries (the overall level effect is negative, but the interaction term is both positive and significant).55 TABLE A.1  Ordinary Least Squares (OLS) Estimates of Principal Components of Data Governance Regulations, Interacted with Log(GDP/capita) on Digital Intensity, Longevity of Digital Businesses and Exit Rates Digital business Digital business Digital business exits intensity longevity (Number of digital (Number of digital (Number of digital businesses that businesses businesses that are at reached exit stage (log)) least 5 years old (log)) (log)) Variables (1) (2) (3) (4) (5) (6) Principal component 1 (mainly personal data, cross-border, -1.729* -2.069*** -1.593* -1.896*** -1.748* -2.083*** cybersecurity, private data) – PC1 (0.894) (0.514) (0.877) (0.568) (0.956) (0.632) Principal component 2 (mainly e-Commerce, public 2.590** 1.309* 1.744 0.835 0.989 -0.274 data, private data) – PC2 (1.183) (0.696) (1.168) (0.773) (1.264) (0.856) Principal component 3 (mainly -2.497 -1.480 -3.102* -2.325* -2.816 -1.814 nonpersonal data) (1.856) (1.068) (1.808) (1.174) (1.985) (1.313) The estimates in Table A.1 are robust to other model specifications – for instance, including the square terms for 53  the principal components – as well as the exclusion of large economies like the US, China, and India, which tend to be outliers in other analyses included in this report. See, for example, (Johnson, Shriver and Goldberg 2021). 54  This finding is likely a consequence of two factors: several low-income countries do not have any regulations 55  pertaining to nonpersonal data (e.g., industry data policy); and high-income countries have the highest prevalence of digital businesses reaching the exit stage. 93  Appendix TABLE A.1 (CONT)  Ordinary Least Squares (OLS) Estimates of Principal Components of Data Governance Regulations, Interacted with Log(GDP/capita) on Digital Intensity, Longevity of Digital Businesses and Exit Rates Digital business Digital business Digital business exits intensity longevity (Number of digital (Number of digital (Number of digital businesses that businesses businesses that are at reached exit stage (log)) least 5 years old (log)) (log)) Log(GDP/ 0.219** 0.260*** 0.202* 0.240*** 0.228** 0.269*** population)*PC1 (0.105) (0.0601) (0.102) (0.0664) (0.112) (0.0740) Log(GDP/ -0.214 -0.134 -0.129 -0.0841 -0.0318 0.0467 population)*PC2 (0.140) (0.0811) (0.137) (0.0900) (0.149) (0.0998) Log(GDP/ 0.280 0.176 0.338* 0.258** 0.320 0.218 population)*PC3 (0.202) (0.116) (0.196) (0.127) (0.216) (0.143) Log(GDP/population) 0.668*** -0.265* 0.488*** -0.271* 0.376* -0.543*** (0.182) (0.139) (0.181) (0.157) (0.195) (0.171) Log(Population) 0.962*** 0.875*** 0.948*** (0.0857) (0.0946) (0.105) Ease of doing 0.0507*** 0.0319** 0.0501*** business score 2020 (0.0134) (0.0152) (0.0165) Constant -1.440 -21.45*** -0.993 -18.99*** -1.219 -20.94*** (1.534) (1.930) (1.514) (2.142) (1.641) (2.374) Observations 74 74 71 71 74 74 R-squared 0.575 0.865 0.474 0.787 0.467 0.775 Source: World Bank Digital Business Indicators (DBI), FCI Digital Business Database. Note: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. This analysis only uses data from CB Insights, which covers all regions. Country N=79. This evidence is preliminary; and the lack of panel data as well as the limited sample size call for a more thorough investigation of this very important relation- ship between data regulations and digital market dynamism. A Spiky Digital Business Landscape  94 APPENDIX E: OTHER FIGURES AND TABLES FIGURE A.4  Employee Size Distribution by Region Entire Asia Pacific 40% 36% 35% 34% 30% 27% 24% 26% 20% 12% 10% 7% 0% 1-9 persons 10-49 persons 50-299 persons 300 or more persons Developed (N=301) Developing (N=735) Entire Middle East and North Africa 40% 36% 33% 33% 30% 29% 29% 20% 20% 16% 10% 6% 0% 1-9 persons 10-49 persons 50-299 persons 300 or more persons Developed (N=367) Developing (N=180) Entire Sub-Saharan Africa 40% 37% 33% 32% 30% 25% 26% 21% 20% 16% 10% 10% 0% 1-9 persons 10-49 persons 50-299 persons 300 or more persons Developed (N=19) Developing (N=689) Source: Authors’ calculations using the FCI Digital Business Database. Note: Staff size indicates the number of the company’s employees in all of the their operating locations. Staff size information has a lot of missing values, and this figure only shows digital companies in SSA, MENA, and some EAP countries. The staff size brackets are defined by the IFC’s MSME definition. 95  Appendix TABLE A.2  Top 20 Country by Number of Firms, Total Funding, Investment Exit BUSINESS TOTAL RANK DENSITY FUNDING # OF EXIT % OF EXIT 1 Estonia Kenya India New Zealand 2 Kenya India United Kingdom United States 3 India Colombia United States Brazil 4 Israel Cayman Islands Sweden Germany 5 United Kingdom Ghana New Zealand Sweden 6 United States Singapore Estonia Netherlands 7 Iceland Israel Kenya Belgium 8 Armenia Myanmar Finland United Kingdom 9 Cambodia Tanzania Cambodia Argentina 10 Finland Cambodia Iceland Canada 11 Pakistan Estonia Canada France 12 Singapore Nigeria Myanmar Costa Rica 13 Sweden China South Africa Philippines 14 Bulgaria Iceland Netherlands India 15 Cayman Islands Uruguay Denmark Norway 16 Canada Luxembourg Bulgaria Denmark 17 Lebanon South Africa Singapore Peru 18 Vietnam Indonesia France Australia 19 Myanmar Vietnam Latvia Czech Republic 20 South Africa Malaysia Vietnam South Africa RED: Countries that show up GREEN: Countries that show up among top 20 in 4 categories as top 20 in at least 2 categories Note: All rankings are normalized using the same methodology as digital business density gap. A Spiky Digital Business Landscape  96 TABLE A.3  Top 10 Digital Subsectors by Number of Businesses (Developed vs. Developing) DEVELOPED COUNTRIES DEVELOPING COUNTRIES (N=280,119) (N=69,346) 1 Health tech* 1 E-commerce 2 Business management tech 2 Fintech 3 Fintech 3 Health tech* 4 Marketing tech 4 Business management tech 5 E-commerce 5 Marketing tech 6 Software and SaaS 6 Entertainment tech 7 Security tech 7 Edtech 8 Big data and analytics 8 Big data and analytics 9 Entertainment tech 9 Software and SaaS 10 Digital media 10 Digital media RED: Countries that show up GREEN: Countries that show up among top 20 in 4 categories as top 20 in at least 2 categories Source: Authors’ calculations using the FCI Digital Business Database Note: *One reason health tech ranks particularly high is because 70 percent of health tech firms offer solutions in two or more subsectors, for example in IT, CRM, and other digital services tailored for the heath care industry (i.e. health tech firms do not necessarily only engage in medical services). Subsector-company pair is used in the analysis because some digital businesses offer digital solutions in multiple sectors. Those digital businesses are counted in each of their subsectors. TABLE A.4  Top 10 Digital Subsectors by Total Funding after Excluding US, China, and India DEVELOPED COUNTRIES DEVELOPING COUNTRIES (EXCL. US) (EXCL. CHINA AND INDIA) Company- Company- Top 10 Total Funding Top 10 Total Funding Subsector Subsector Subsectors (Million USD) Subsectors (Million USD) pair N pair N Fintech* 90242.89 7934 Fintech* 93,156 1247 Mobility tech* 64444.2 3691 E-commerce* 52,791 1276 Security tech* 52416.53 5220 Travel tech* 42,788 394 E-commerce* 52248.5 7489 Mobility tech* 34,263 405 Social Travel tech* 44792.53 2657 31,995 337 network* Big data and 42048.72 4964 Security tech* 28,504 390 analytics Social 40707.22 2209 Telecom* 26,916 250 network* Health tech 40242.79 8548 Insurance tech 25,550 176 Business man- Software and 31807.95 7526 17,787 405 agement tech SaaS Entertainment Telecom* 30057.24 2656 17,000 418 tech Note: *subsectors that are top 10 in both income groups; Blue boxes are the new top 10 subsectors when com- pared with Table 3.1 before the US, China, and India are excluded from the sample. 97  Appendix FIGURE A.5  Top 10 Subsectors by Total Number of Exits – Developed vs Developing Countries Top 10 Subsectors Top 10 Subsectors Count of Exited Digital Count of Exited Digital Business in Developed countries Business in Developing countries business management tech 15% , N=6873 e-commerce 15%, N=884 health tech 14% , N=6503 fintech 13%, N=753 fintech 13% , N=5784 marketing tech 12%, N=670 marketing tech 12% , N=5646 business management tech 11%, N=632 software and SaaS 11% , N=5116 entertainment tech 9%, N=521 e-commerce 11% , N=5088 tech hardware 9%, N=504 security tech 9%, N=4220 digital media 8%, N=448 big data and analytics 9%, N=4052 software and SaaS 7%, N=434 digital media 8%, N=3754 big data and analytics 7%, N=429 tech hardware 8%, N=3597 telecom 7%, N=407 0% 5% 10% 15% 20% 0% 5% 10% 15% 20% Number of Exited Digital Business Number of Exited Digital Business in Developed countries (N=45,369) in Developing countries (N=5795) Source: Authors’ calculations using the FCI Digital Business Database. Some digital businesses can offer digital solutions in multiple sectors, and their funding information is counted multiple times in each subsector, with the assumption that the investments in multi-subsector firms trickles down across subsectors. A robustness check was done for multi-subsector firms by using total funding divided by number of subsectors a firm operates in, with similar ranking of subsectors. Note: Exit information is based on digital businesses headquartered in developed and developing countries that have funding information. Exits include IPOs, Mergers and Acquisitions, Majority Buyouts, Management Buyouts, and Other Exits including Secondary Transaction - Stock Distribution, Asset Sale. Dividend Recapitalization. FIGURE A.6  Top 10 Subsectors by Total Ticket Size of Exits – Developed vs Developing Countries after excluding the US, China, and India Top 10 Subsectors Top 10 Subsectors Exit Ticket Size in Developed countries Exit Ticket Size in Developing countries (excl. USA) (excl. China and India) fintech 40.4, 46% telecom 17.6, 31% e-commerce 38.8, 44% fintech 10.8, 19% logistics tech 16.7, 19% tech hardware 7.6, 13% web services 11.9 , 14% social network 5.9, 10% big data and analytics 11.2 , 13% e-commerce 5.2, 9% social network 9.1 , 10% entertainment tech 4.3, 7% mobility tech 6.2, 7% health tech 3.8, 7% software and SaaS 5.7, 7% logistics tech 2.6, 4% entertainment tech 5.7, 6% digital media 2.3, 4% digital media 4.8, 5% prop tech 2.1, 4% 0% 20% 40% 60% 0% 20% 40% % of exit value ($87B) % of exit value ($57B) Source: Authors’ calculations using the FCI Digital Business Database. A Spiky Digital Business Landscape  98 FIGURE A.7  Digital Clusters Among Developed Country Businesses Founded Before 2015 Small Cluster Mega Cluster (more than 2 small clusters linked closely) FIGURE A.8  Digital Clusters Among Developing Country Businesses Founded Before 2015 Small Cluster Mega Cluster (more than 2 small clusters linked closely) 99  Appendix FIGURE A.9  Digital Clusters Among Developing Country Businesses Founded After 2015 Small Cluster Mega Cluster (more than 2 small clusters linked closely) Source: Research support by WB DEC Analytics and Tools Unit (DECAT) using the FCI Digital Business Database. Note: Uniform Manifold Approximation and Projection was used for this analysis. This is a dimension reduction technique used in data science to visualize sparse multidimensional data. This analysis shows neighbors = 4 results to balance local versus global structure of clusters. *2015 is considered an inflection point in digitalization because of the mass availability of cloud-enabled digital solutions that started around 2010-2015. FIGURE A.10  Distribution of Firms according to Presence in Number of Subsectors/Markets after excluding the US, China, and India 50% 42% 42% % of Digital Business 40% 33% 32% 30% 20% 16% 16% 6% 7% 10% 3% 4% 0% 1 Subsector 2 Subsectors 3 Subsectors 4 Subsectors 5 Subsectors or more Developed excl. USA (N=68K) Developing excl. China and India (N=17K) Source: Authors’ calculations using the FCI Digital Business Database. A Spiky Digital Business Landscape  100 FIGURE A.11  Digital Firms Receiving Funding in Applied and Deep Tech by Founding Year after excluding the US, China, and India Applied Tech (e-commerce, fintech & travel tech) % of funded business 60% 51% 35% 37% 40% 28% 20% 17% 11% 7% 7% 4% 3% 0% 1971-1999 2000-2004 2005-2009 2010-2014 2015-2020 Founding year Developed (Excl. USA) (N=4012) Developing (Excl. CHN, IND) (N=1461) Deep Tech (AI/ML, internet of things, quantum tech) 100% % of funded firms 80% 66% 58% 60% 40% 27% 26% 20% 7% 4% 4% 5% 1% 4% 0% 1971-1999 2000-2004 2005-2009 2010-2014 2015-2020 Founding year Developed (Excl. USA) (N=1727) Developing (Excl. CHN, IND) (N=243) Source: Authors’ calculations using the FCI Digital Business Database. FIGURE A.12  U-Shape Hypothesis: Acquisition Market Concentration by Countries/Income Group (HHI of top 50 acquirer, by M&A deal size) 250000 Digital acquisitions market concentration 20000 15000 10000 5000 0 0 20000 40000 60000 GDP per Capita (US Current, 2020) Source: Authors’ calculations using the FCI Digital Business Database. Note: Scandinavia shows up as an outlier due to few major acquirers (Nokia – Finland, Elisa Automate (Now Elisa Polystar) – Finland, Nordic Capital – Denmark). Individual country observations that have at least 100 M&A deals by top 50 acquirers are separated out to show their HHI. If a country does not have enough sample size of M&A deals, we group it to the specific income group in the region that it belongs. Count of M&A deals by top 50 acquirers: Non-Europe high income countries: 72; Rest of high income Europe: 59; Scandinavia: 63; Upper middle income: 57; Lower middle income: 53; Canada: 69; China: 80; France: 60; Germany: 59; United Kingdom: 66; United States: 746. 101  Appendix FIGURE A.13  Top Subsectors with Women in Management Team, Entire Sub- Saharan Africa Top Subsectors in Entire SSA: Gender Breakdown health tech n=13, 35% n=54, 65% agtech n=7, 27% n=30, 73% big data and analytics n=6, 26% n=26, 74% artificial intelligence n=13, 26% n=58, 74% entertainment tech n=13, 23% n=60, 77% edtech n=52, 23% n=254, 77% HR tech n=13, 19% n=65, 81% travel tech n=6, 19% n=31, 81% clean tech n=38, 19% n=203, 81% food tech n=10, 18% n=54, 82% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Digital Businesses with At Least One Woman in the Core Team Proportion of Digital Businesses with No Women in the Core Team FIGURE A.14  Top Subsectors with Women in Management Team, entire MENA Top Subsectors in Entire MENA: Gender Breakdown social network n=9, 35% n=37, 65% security tech n=7, 32% n=29, 68% health tech n=16, 26% n=69, 74% edtech n=11, 21% n=48, 79% logistics tech n=44, 20% n=192, 80% big data and analytics n=13, 20% n=58, 80% artificial intelligence n=19, 19% n=90, 81% food tech n=12, 19% n=59, 81% software and SaaS n=8, 19% n=40, 81% fintech n=43, 19% n=271, 81% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Digital Businesses with At Least One Woman in the Core Team Proportion of Digital Businesses with No Women in the Core Team Source: Authors’ calculations using the FCI Digital Business Database, and the World Bank Gender Indicator Report. Note: Other regions are excluded from this analysis due to missing values. Subsector-company pairing is used in the analysis because some digital businesses offer digital solutions in multiple sectors. A Spiky Digital Business Landscape  102 FIGURE A.15  Number of Clean Tech Companies/GDP per Capita TOP 20 COUNTRIES (BY RANKING) Clean tech rms per GDP per Capita *Denotes developed countries 0.000 0.033 India (n=72) USA (n=1554) * China (n=168) Kenya (n=16) United Kingdom (n=360)* Uganda (n=5) South Africa (n=25) Canada (n=178)* Egypt (n=10) France (n=133) * Nigeria (n=7) Germany (n=145)* Vietnam (n=7) Brazil (n=21) Ghana (n=5) IBRD 46844 | Spain (n=58)* OCTOBER 2022 Japan (n=63)* IBRD 46844 | OCTOBER 2022 Sweden (n=77)* Malaysia (n=16) Thailand (n=10) Source: Authors’ calculations using the FCI Digital Business Database. 103  Appendix GLOSSARY Key Definitions Acqui-hire: An acquisition where the buyer’s intention is to secure the target company’s talent and personnel rather than to develop or monetize the target’s technology, product or services. Applied tech: An approach to problem solving that relies on applying technology to create practical solutions. E-commerce, fintech, mobility tech are examples of digital technology being applied to specific areas to provide better solutions. B2B (business-to-business): A business process that enables the transaction of goods and services between businesses. B2C (business-to-consumer): A business process that enables transactions between businesses and end-user consumers of the products or services. Big tech: Refers to major large technology companies that have dominant influence in the market. Clean tech: Refers to digital businesses providing solutions in to enhance environmental impact and/or reduce environmental harm. Cloud computing: Refers to practice of using a network of remote servers hosted on the internet (the cloud) to access computing power, storage, databases without owning the physical computing infrastructure. Crowdfunding: The practice of funding a project or venture by raising money from a large number of people, typically via the Internet. Crowdsourcing: A method of information gathering where a large group (“the crowd”) participates in gathering information, come up with innovative ideas and pitch in on solving problems together. Data-driven business: A firm that systematically and methodically collects or aggregates large data sets and leverages advanced analytics (such as artificial intelligence, big data, and blockchain) to create value, and that leverages data as a key element of their business model. Data-driven businesses can also help traditional industries upgrade through services designed to optimize production processes, increase sales, streamline decision making, and even rethink revenue models. Deep technology: An approach to problem solving that relies on using emerging technology to create useful applications. Deep tech often uses or develops emerging technology embedded in science and advanced engineering to create high-value B2B or B2C solutions. A Spiky Digital Business Landscape  104 Digital business gap: Difference between the actual and potential number of digital businesses in a country, where the latter is estimated by regressing the actual number of digital businesses in the country on its GDP and population (all measured in logs). Digital conglomeration: The process of forming a conglomerate enabled by digital business models. Digital single market: Refers to the European Commission’s initiative to transform the European single market into one characterized by ensuring the free movement of people, services and capital and allowing individuals and businesses to seamlessly access and engage in online activities irrespective of their nationality or place of residence. Exit: Initial public offerings (IPO), mergers and acquisitions (M&A), majority buyout (including management and leveraged buyouts), and other exits, including stock distribution secondary transactions, asset sales, and dividend recapitalization. Investment deals: Investment deals range from the very early to the mature stages, and include grants (pitch competition prizes, foundation grants, etc.); pre-seed/ seed funding (for the initial stages of company development, for example through angel investors, incubator/accelerators, and crowdfunding); venture capital (early stage: Series A-B, late stage: Series C-Z); private equity; debt financing; mezzanine (a hybrid of debt and equity financing where lenders can convert their investment to equity in case of default); and other forms of capitalization (capital to cover operational and developmental work, for example, or bonds, corporate asset purchases, joint ventures, corporate licensing, secondary transactions, etc.). Killer acquisitions: An acquisition where a big incumbent firm acquires a startup solely to pre-empt future competition from the target company. P2P (peer-to-peer): A business process that enables interactions between two individuals (crowdfunding, social network platforms, and peer-to-peer exchanges). Platform-based business: A firm that facilitates interactions that include a large number of participants. The platform business model does not own the means of production, but rather creates and facilitates the means of connection. The role of the platform-based business is to provide a governance structure and a set of standards and protocols that facilitate interactions at scale so that network effects can be unleashed. Platform-based businesses benefit from networking effects (the larger the crowd, the more value added to the platform).  Regulatory sandboxes: A regulatory sandbox is a regulatory approach that allows live, time-bound testing of innovations under a regulator’s oversight. Novel financial products, technologies, and business models can be tested under a set of rules, supervision requirements, and appropriate safeguards. Scale without mass: Outcome enabled by digital technologies whereby highly digitalized businesses are able to locate various stages of their production processes across different jurisdictions, accessing a large number of customers around the globe without any significant physical presence (such as property or employees), thus achieving operational local scale without local mass. 105  Glossary Super-app: Refers to an application that forms a comprehensive ecosystem of multiple inter-related services on one platform (e.g., payment, e-commerce, communication). Ticket size: The amount invested in a company during an investment deal (refers to the size or the amount of the investment deal). Tipping: The tendency of a digital platform/ecosystem to pull away from its rivals in popularity once it has gained a critical mass of users. This phenomenon is enabled by positive network effects. Unicorns: Refers to firms with a market value of over $1 billion. Web scraping: Refers to the process of using computer programming software to methodically and automatically collect information from across the internet. A Spiky Digital Business Landscape  106 BIBLIOGRAPHY Aly, Heidi. 2020. “Digital transformation, development and productivity in devel- oping countries: is artificial intelligence a curse or a blessing?” Review of Economics and Political Science 2631-3561. Argentesi, E., Paolo Buccirossi, Emilio Calvano, Tomaso Duso, Alessia Marrazzo, and Salvatore Nava. 2020. Tech-over: Mergers and merger policy in digital markets. March. https://voxeu.org/article/mergers-and-merger-policy-digital-markets. Birch, Kean, and D. T. Cochrane. 2021. “Big Tech: Four Emerging Forms of Digital Rentiership.” Science as Culture 1-15. Bukht, Rumana, and Richard Heeks. 2017. “Defining, conceptualising and measuring the digital economy.” Development Informatics working paper 68. Chen, Rong. 2021. “Mapping Data Governance Legal Frameworks around the World: Findings from the Global Data Regulation Diagnostic.” Policy Research Working Papers. The World Bank. doi: https://doi.org/10.1596/1813-9450-9615. Cirera, Xavier, Diego Comin, Marcio Cruz, and Kyung Min Lee. 2020. “Technology Within and Across Firms.” Policy Research Working Paper No 9476. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/34795. Crémer, Jacques, Yves-Alexandre de Montjoye, and Heike Schweitzer. 2019. Compe- tition Policy for the digital era. Brussels: European Commission. Cunningham, Colleen, Florian Ederer, and Song Ma. 2020. Killer Acquisitions. Vol. 129. Journal of Political Economy, 04 19. 649-792. doi: https://dx.doi.org/10.2139/ ssrn.3241707. Deloitte. 2020. “Fintech: On the brink of further disruption.” Netherlands. Disrupt Africa. 2020. African Tech Startups FUNDING REPORT 2020. Disrupt Africa. Evans, and David S, Richard Schmalensee. 2013. “The antitrst analysis of multisided platform businesses (No.w18783).” National Bureau of Economic Research. Fernandez-Vazquez, Simon, Rafael Rosillo, David De La Fuente, and Paolo Priore. 2019. “Blockchain in FinTech: A Mapping Study.” Sustainability 11 (22). Goldfarb, Avi, Shane M. Greenstein, and Catherine E. Tucker. 2015. “Introduction to “Economic Analysis of the Digital Economy” .” Economic Analysis of the Digital Economy. Grewal, R., G.L. Lilien, S. Bharadwaj, P. Jindal, U. Kayande, R.F. Lusch, M. Mantrala, et al. 2015. “Business-to-business buying: Challenges and opportunities. .” Customer needs and Solutions 2 (3): 193-208. 107  Glossary Hartmann, Philipp Max, Mohamed Zaki, Niels Feldmann, and Andy Neely. 2014. “Big Data for Big Business?: A Taxonomy of Data-driven Business Models used by Start-up Firms.” Cambridge Service Alliance. University of Cambridge. http:// www.nsuchaud.fr/wp-content/uploads/2014/08/Big-Data-for-Big-Business-A- Taxonomy-of-Data-driven-Business-Models-used-by-Start-up-Firm.pdf. Highfill, Tina, and Christopher Surfield. 2022. New and Revised Statistics of the U.S. Digital Economy, 2005-2020. Bureau of Economic Analysis. https://www.bea. gov/system/files/2022-05/New%20and%20Revised%20Statistics%20of%20 the%20U.S.%20Digital%20Economy%202005-2020.pdf. Huawei. 2017. Digital Spillover Measuring the true impact of the digital economy. Oxford Economics. IFC. 2020. Computing platforms for big data analytics and artificial intelligence. IFC. Jia, Jian, Ginger Zhe Jin, and Liad Wagman. 2021. “The short-run effects of the general data protection regulation on technology venture investment.” Market- ing Science 661-684. Johnson, Garrett, Scott Shriver, and Samuel Goldberg. 2021. “Privacy & market concentration: intended & unintended consequences of the GDPR.” SSRN Working Paper 3477686. Lemley, Mark A, and Andrew McCreary. 2021. “Exit strategy.” B.U.L. Review. Mazzucato, Marianna. 2013. The Entrepreneurial State: Debunking public vs. private sector myths. London: Anthem Press. McKinsey Global Institute. 2020. How Asia can boost growth through technological leapfrogging. McKinsey&Company. McKinsey Global Institute. 2021. The trailblazing consumers in Asia propelling growth. McKinsey&Company. McKinsey&Company. 2017. “Competing in a world of sectors without borders .” McKinsey Quarterly, July 12. Mermelstein, B., V. Nocke, M. A. Satterthwaite, and M. D. Whinston. 2020. “Internal versus external growth in industries with scale economies: A computational model of optimal merger policy.” Vol. 128(1). Journal of Political Economy. 301- 341. Nyman, Sara, and Rodrigo Barajas. 2021. Antitrust and Digital Platforms: An analysis of global patterns and approaches by competition authorities. Wold Bank Group. Ocampo, Andrés Felipe Ramírez. 2019. “Scale Without Mass: Permanent Estab- lishments in the Digital Economy.” IBDT Actual. https://www.ibdt.org.br/ RDTIA/n-5-2019/scale-without-mass-permanent-establishments-in-the-digi- tal-economy/. OECD. 2020a. A Roadmap toward a Common Framework for Measuring the Digital Economy. Saudi Arabia: OECD. https://www.oecd.org/digital/ieconomy/road- map-toward-a-common-framework-for-measuring-the-digital-economy.pdf. OECD. 2018. Bridging the Digital Gender Divide: Include, Upskill, innovate. OECD. A Spiky Digital Business Landscape  108 OECD. 2021. Ex Ante Regulation and Competition in Digital Markets. OECD Compe- tition Committee Discussion Paper. , https://www.oecd.org/daf/competition/ ex-ante-regulation-andcompetition-in-digital-markets.htm. OECD. 2020b. Latin American Economic Outlook 2020: Digital Transformation for Building Back Better. OECD. Petit, Nicolas, and David J. Teece. 2020. Taking Ecosystems Competition Seriously in the Digital Economy. OECD. Pompella, Maurizio, and Roman Matousek. 2021. Fintech and Blockchain: Contem- porary Issues, New Paradigms, and Disruption. Switzerland: Palgrave Macmillan. PPC. 2019. The Digital Roadmap: how developing countries can get ahead. Pathways for Prosperity Commission. Oxford, UK. Rasmusen, Eric. 1988. “Entry for Buyout.” Vol. 36(3). The Journal of Industrial Eco- nomics. 281-299. doi: https://doi.org/10.2307/2098468. Simcoe, Timothy. 2015. “Modularity and the Evolution of the Internet.” In Economic analysis of the digital economy, 21-47. University of Chicago Press. Still, Kaisa E. 2017. “Business Model Innovation of Startups Developing Multisided Digital Platforms.” IEEE. Sturgeon, Timothy J. 2021. “ Upgrading strategies for the digital economy .” Global Strategy Journal 11 (1): 34-57. Suominen, Kati. 2018. Fueling Digital Trade in Mercosur: A Regulatory Roadmap. Inter-American Development Bank. Tian, Diana, Justin Crist Lee, Daniel Fetner, and Ryan Freedman. 2020. “The Inter- section of Construction Tech and FinTech.” Alpaca. UNCTAD. 2019. Digital Economy Report: Value Creation and Capture - Implications for Developing Countries. UNCTAD. UNCTAD. 2021. “Estimates of Global E-Commerce 2019 abd Preliminary Assess- ment of COVID-19 Impact on Online Retail 2020.” Geneva. WBG. 2018. Connecting to Compete: Trade Logistics in the Global Economy. World Bank Group. WBG. 2021. Data for Better Lives. World Bank Group. World Bank. Forthcoming. “Technology Transformation for Jobs in Africa: How Digital Development Support Good Jobs for All.” 109  Glossary A Spiky Digital Business Landscape  110