91561 2 Do mLabs Make a Difference? A HOLISTIC OUTCOME ASSESSMENT OF infoDev’S MOBILE ENTREPRENEURSHIP ENABLERS 3 Copyright The report “Do mLabs Make a Difference? A Holistic Outcome Assessment of infoDev’s Mobile Entrepreneurship Enablers” is available at http:// www.infodev.org/mobile/mLaboutcomes Copyright Rights and Permissions © 2014 International Bank for Reconstruction and This work is available under the Creative Development / The World Bank Commons Attribution 3.0 Unported license (CC BY 3.0) Mailing Address: MSN I9-900 1818 H St. NW, http://creativecommons.org/licenses/by/3.0. Washington D.C., 20433 USA Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, Telephone: (+1) 202- 458-4070 and adapt this work, including for commercial purposes, under the following conditions: Website: www.infoDev.org Attribution—Please cite the work as follows: Email: info@infodev.org infoDev. 2014. “Do mLabs Make a Difference? A Holistic Outcome Assessment of infoDev’s Twitter: @infoDev Mobile Entrepreneurship Enablers.” Washington, DC: The World Bank. Website: Facebook: /infoDevWBG http://www.infodev.org/mobile/mLaboutcomes License: Creative Commons Attribution CC BY Some rights reserved. 3.0 This work is a product of the staff of The World Translations—If you create a translation of this Bank with external contributions. Note that work, please add the following disclaimer along The World Bank does not necessarily own each with the attribution. component of the content included in the work. This translation was not created by The World The World Bank therefore does not warrant that Bank and should not be considered an official the use of the content contained in the work will World Bank translation. The World Bank shall not infringe on the rights of third parties. The risk not be liable for any content or error in this of claims resulting from such infringement rests translation. solely with you. All queries on rights and licenses should be The findings, interpretations, and conclusions addressed to infoDev, The World Bank, 1818 H expressed in this work do not necessarily Street NW, MSN 5K–508, 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 of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. 4 Acknowledgments This Outcome Assessment was produced by We would also like to thank the comments and the Center of Partnerships for Development observations from the team of peer reviewers: (CAD) and commissioned by infoDev, a global Jean-Louis Racine (Europe and Central Asia partnership program with the World Bank. The Financial and Private Sector Development report features an analysis of the holistic effects Department, World Bank), Janne Sykkö (Ministry of mobile application labs (mLabs), which were for Foreign Affairs of Finland), Lany Sommers implemented under the Creating Sustainable (Business Development Middle East and Africa Businesses in the Knowledge Economy (CSBKE) Region, Blackberry), and Omar Cissé (CTIC Dakar). program, funded by the government of Finland in Additional comments from the infoDev team and partnership with Nokia. consultants came from Helen Akanisi, Catherine Amelink, Brett Dickstein, and Zoe Lu. This report would not have been possible without the valuable contribution of all mLab managers, Finally, we thank Carollyne Hutter for copy-editing mobile application developers, entrepreneurs, and haiku media for designing the report. incubators, accelerators, institutional partners, government agencies, mobile operators, About infoDev and CSBKE mobile device manufacturers, journalists, infoDev, a global trust fund program in the and universities that have generously made Financial and Private Sector Development Network themselves available for interviews in Armenia, of the World Bank Group, supports growth- South Africa, and Kenya. In particular, the CAD oriented entrepreneurs through innovative venture Team would like to give a special thanks to John enablers. It assists entrepreneurs to secure Kieti, manager of mLab East Africa, Derrick appropriate early-stage financing; convening Kotze, CEO of mLab Southern Africa, and Mariam entrepreneurs, investors, policy makers, mentors, Davtyan, director of mLab ECA, for the warm and other stakeholders for dialogue and action. welcome and active support during the exercise infoDev also produces cutting-edge knowledge and for their invaluable help, patience, and products, closely linked to our work on the ground. guidance. Among other initiatives, infoDev has implemented the Creating Sustainable Businesses in the This assessment allowed the team to meet with Knowledge Economy (CSBKE) program, a trust a number of entrepreneurs with a passionate fund established by the government of Finland dedication to make a difference every day. We with the objective to increase the growth of small, thank them very much for their availability innovative, and technology-based business, and inspirational insights. We would also like primarily in the information and communications to express our gratitude to the wide array technology (ICT) and agribusiness sectors. The of organizations that kindly agreed to share CSBKE program was designed as a public-private their views on the role of mLabs within the partnership among infoDev, Finland, and the Nokia entrepreneurial ecosystems in their countries. Corporation and covers the period from March We hope the report provides them with useful 2010 to June 2014. feedback and insights. About CAD The CAD team is also very grateful for CAD (Centre of Partnerships for Development) contributions from Nicolas Friederici (Oxford is a network of international experts specialized Internet Institute) and to Loren Nadres and Toni in international development, local economic Eliasz (infoDev) for their intense and committed development, and public-private partnerships, with reviews. Ellen Olafsen (infoDev) was the task a focus on small and medium enterprises (SMEs) team leader for the project. in developing countries, entrepreneurship, Base of the Pyramid, and monitoring and evaluation tools and methods. TABLE OF CONTENTS 1. EXECUTIVE SUMMARY 8 2. SCOPE AND OBJECTIVES 10 3. ASSESSMENT METHODOLOGY 12 APPROACH AND RATIONALE 12 DATA COLLECTION 12 SAMPLING 12 LIMITATIONS 13 4. MLABS IN THE CONTEXT OF ENTREPRENEURIAL ECOSYSTEMS 14 MLAB SOUTHERN AFRICA AND THE SOUTH AFRICAN ENTREPRENEURIAL ECOSYSTEM 15 MLAB ECA AND THE ARMENIAN ENTREPRENEURIAL ECOSYSTEM 18 MLAB EAST AFRICA AND THE KENYAN ENTREPRENEURIAL ECOSYSTEM 22 5. ASSESSMENT OF MLABS’ EFFECTS 26 5.1 MLABS’ EFFECTS ON STARTUP CREATION AND DEVELOPMENT 26 5.1.1 High-Level Findings on mLabs’ Contribution to Startup Creation and Development 27 5.1.2 Startup Creation 28 5.1.3 Revenue Generation by Supported Startups 30 5.1.4 Job Creation by Supported Startups 31 5.1.5 External Investment Attracted by Supported Startups 32 5.2 DEVELOPMENT IMPACT THROUGH THE USAGE OF MOBILE APPLICATIONS 34 5.2.1 Identification Process for Mobile Applications with Development Impact 35 5.2.2 Characterization of the Sample of Identified Applications 36 5.2.3 Assessment and Illustration of Development Impact through Application Usage 38 5.3 MLABS’ EFFECTS ON ENTREPRENEURIAL ECOSYSTEMS 53 5.3.1 Strengthening the Talent Pool and Focusing on Early Stage Innovators 56 5.3.3 Inspiring and Stimulating Innovation in the Ecosystem 61 5.3.3 Inspiring and Stimulating Innovation in the Ecosystem 63 6 5.3.4 Connecting Local, Regional, and International Ecosystems 67 6. SUMMARY AND CONCLUSIONS 69 6.1 MLABS’ EFFECTS ON STARTUP CREATION AND DEVELOPMENT 69 6.2 DEVELOPMENT IMPACT THROUGH THE USAGE OF MOBILE APPLICATIONS 70 6.3 EFFECTS OF MLABS ON ENTREPRENEURIAL ECOSYSTEMS 71 6.4 OVERALL ASSESSMENT 71 ANNEXES 74 ANNEX A: CHARTS ON MLAB-SUPPORTED STARTUP CREATION AND DEVELOPMENT 74 ANNEX B: LIST OF INTERVIEW PARTICIPANTS 79 ANNEX C: RESEARCH PROJECT TIMELINE 84 Box 1: The Entrepreneurial Ecosystem 11 Box 2: GoMetro 41 Box 3: Eneza Education 45 Box 4: Creating Incentives and Channels for Startup Creation and Development through Events and Competitions 57 Box 5: Community Building: A Key Explanatory Factor Behind Several Effects 65 Figure 1: mLab Southern Africa Ecosystem 16 Figure 2: mLab ECA Ecosystem 20 Figure 3: Ecosystem mLab East Africa 24 Figure 4: Numbers of Startups Created by mLabs 28 Figure 5: Revenue Generation by mLab-supported Startups and Entrepreneurs 30 Figure 6: Jobs Created by mLab-supported Startups and Entrepreneurs 31 Figure 7: External Investments Attracted by mLab Startups 32 Figure 8: Mobile Applications by mLab Service Type 35 Figure 9: Overview of the Selected Applications by Country and by Sector 37 Figure 10: Operational Life of the 29 Selected Applications 37 Figure 11: Overview of mLabs’ Contributions to Entrepreneurial Ecosystems 54 Figure 12: An Example of Multiplier Effects in the Creation and Transfer of Knowledge 59 Table 1: The Three Research Questions for the Assessment 11 Table 2: Sample Groups 13 Table 3: mLab Southern Africa’s Role in the South African Entrepreneurial Ecosystem 15 Table 4: mLab ECA’s Role in the Armenian Entrepreneurial Ecosystem 21 Table 5: mLab East Africa’s Role in the Kenyan Entrepreneurial Ecosystem 25 Table 6: Data on Training and Coaching Delivery 58 Table 7: Number of New Mobile Applications brought to Market 63 7 1. EXECUTIVE SUMMARY This report is an assessment of the overall effects of infoDev’s mobile application labs (mLabs). mLabs are specialized mobile business incubation and acceleration facilities offering physical workspace, mentoring and coaching, devices for mobile application testing, training, and startup competitions. mLabs were designed as pilot projects under the Creating Sustainable Businesses in the Knowledge Economy (CSBKE)1 program and have since become part of infoDev’s Mobile Innovation Program.2 The Oucome Assessment compares mLabs’ results The geographical scope of the assessment is South against the original program objectives. At the Africa for mLab Southern Africa; Armenia for mLab outset, the goal for mLabs was to derive and ECA (Eastern Europe, South Caucasus, and Central test new approaches that would “1) increase the Asia); and Kenya for mLab East Africa. mLab East competitiveness of innovative enterprises in the Asia in Vietnam was not part of this assessment as mobile industry, and 2) ensure that locally relevant the mLab was not operational long enough to justify applications are created to meet growing developing a comprehensive effect analysis. The assessment country user demands.” Within the three-year project thus covers three of the four operational mLabs. implementation period, the pilots were expected to “generate 8-10 mobile applications; increase the Data collection encompassed 116 semistructured commercialization rate of applications that have the interviews with stakeholders from 82 different potential for significant development impact; and organizations, covering most stakeholder groups of increase the scale and competitiveness of innovative analyzed entrepreneurial ecosystems. In addition mobile applications enterprises leading to greater to primary qualitative data, secondary quantitative reach to disadvantaged populations.” The design data from monitoring records and statistical sources and testing of mLabs started in 2010. were considered. Overall, the analysis followed a qualitative in-depth assessment methodology and In accordance to the original objectives, the a contribution analysis approach, employing a “best assessment examines mLabs’ effects on startup effort” qualitative counterfactual. creation and development (as the best available proxy measures of increased competitiveness in mobile The assessment adds to an earlier evaluation application enterprises) and the development impact of mLabs’ business models 3. The present report generated through the use of applications developed therefore covers only the information pertaining by mLab-supported developers and entrepreneurs, to the Outcome Assessment. It will not elaborate including the number of relevant applications. The on findings and background information already assessment also includes an analysis of mLabs’ described in the Business Model Evaluation, effects on entrepreneurial ecosystems, since such as the following: information on budgets anecdotal evidence previous to the assessment of the mLab program, the rationale and original pointed towards potentially substantial effects that conception of the program, future directions, were not originally envisioned under the CSBKE detailed case studies of each mLab including mLab program. East Asia, lessons learned, challenges and failures, 1 · The CSBKE program was launched in 2010 with the objective to derive and test new approaches to advancing innovation and entrepreneurship in developing countries. 2 · See http://www.infodev.org/workprogram. 3 · Available at http://www.infodev.org/mobilebusinessmodels. 8 4 · Mobile startup and entrepreneur case studies are available at http://www.infodev.org/mobilecases. operational models, financial sustainability, and incipient stage of roll-out, but examples with potential more. The report provides excerpts of more detailed development impact include applications for reforestation case studies from noteworthy mobile application and water quality management. startups that were published separately 4. mLabs’ Contribution to Entrepreneurial Ecosystems mLabs’ Effects on Startup Creation and Development mLabs have functioned as active ecosystem builders: Reported results together with qualitative data from their business models and value propositions interviews suggest that mLab-supported startups have focused not only on startup creation, but also have had tangible economic success after less than on elements that have expanded the breadth of three years of operation. There are reasonable entrepreneurial ecosystems. mLabs have created new indications that mLabs contributed to this success, spaces for exchanges, brought in new actors, generated though the evidence is not entirely conclusive and new linkages between stakeholders, opened channels the net economic effect cannot be quantified with and pipelines for startups to emerge that did not exist precision. mLabs supported and incentivized the previously, and stimulated new innovation processes in creation of close to 70 new startups through their the ecosystem. Pilot experiences of all three analyzed activities. The newly created and other client startups mLabs have also shown that the regional and cross- increased revenues over the duration of the support national nature of mLabs enables a high potential to period, created over 180 high-quality jobs, and raised connect local, regional, and international ecosystems. more than $2.2 million in investments and seed funding as of September 2013 (including grants). It was assessed that mLabs had a positive impact on the ecosystem based on the report analysis. Development Impact Through the Use of Mobile Applications Notably, the analysis was able to identify unexpected and indirect contributions, including for client and For the purpose of this study, ‘development impact’ beneficiary groups that are only indirectly connected was defined as a mobile application’s contribution to with the mLabs. While this assessment does not the Millennium Development Goals (MDGs), addressing present precise quantifications, the evidence overall the needs of users at the Base of the Pyramid (BoP), or suggests that the direct net economic effect of mLabs otherwise pointing towards transformative potential to through the creation and development of mobile improve conditions for people and businesses. By this application startups is only one part, and potentially definition, 41 of the total 292 identified mobile applications not even the largest part, of mLabs’ contribution. had actual or potential development impact, of which 29 were analyzed in detail. In particular, mLabs have proven a viable mechanism to enhance entrepreneurial ecosystems. This group of 29 applications includes a few significant This effect is particularly noteworthy as it had not development impact success stories, several applications been an explicit goal in the original mLab program that generating satisfactory effects, and a majority of design. The effect also implies promise for mLab applications still in pilot stages. The number of mobile replications in countries and regions with incipient applications generated greatly exceeded the original and latent, yet promising, mobile entrepreneurship target of 8 to 10 applications per mLab, and in most cases, ecosystems. Regarding development impact the applications represented locally relevant solutions. through the use of mobile applications, the original (and in hindsight, modest) goals for the number of In Kenya, development impact through application applications produced per mLab were surpassed. usage included improved access for impoverished rural Yet, the number of applications with significant communities to fundamental services such as health, development impact that have scaled has so far education, and financial services. In South Africa, three been low. The coming years will determine whether applications are generating a positive impact on mobility the many applications currently undergoing testing and employment opportunities, and two others are raising will in fact be successful in the market, which would awareness on crucial problems such as gender violence add significantly to the overall development impact and child abuse. In Armenia, applications are at a more through application usage. 9 2. SCOPE AND OBJECTIVES This report aims to broadly answer the question: Do mLabs make a positive difference for their clients and stakeholders? To do so, any other notable effects. Originally, the objective of the the document aims to assess mLabs’ holistic mLab program was to derive and test a new operational effects, including effects that had originally approaches that would “1) increase the competitiveness been envisioned as well as those that became of innovative enterprises in the mobile industry, and 2) apparent later. ensure that locally relevant applications are created to meet growing developing country user demands.” Within mLab services are geared towards three broad goals: the project implementation period (three years), the pilots (1) startup creation and support; (2) skills development; were expected to “generate 8-10 mobile applications; and (3) community building5. mLabs were designed as increase the commercialization rate of applications that pilot projects under the Creating Sustainable Businesses have the potential for significant development impact; in the Knowledge Economy (CSBKE)6 program from and increase the scale and competitiveness of innovative infoDev and have since become part of infoDev’s Mobile m-applications enterprises leading to greater reach to Innovation Program7. disadvantaged populations.” infoDev took the traditional concept of business The two first types of effects that the assessment examines incubation and integrated it with elements drawn from thus follow logically from the original objectives of the areas, such as startup acceleration, app economies, mLab program: the report describes (1) mLabs’ effects on and tech and startup communities. The program moved startup creation and development (as the best available from traditional incubators toward more flexible mobile proxy measures of increased competitiveness in mobile entrepreneurship enablers that would be placed at application enterprises)10 and (2) the development impact the heart of mobile innovation and entrepreneurship generated through the use of applications developed by ecosystems. mLab-supported developers and entrepreneurs, including the number of relevant applications. In addition, during Today, mLabs form a part of infoDev’s Mobile Innovation early consultations with infoDev it became clear that a Program, a work stream that supports growth- third type of effect could substantially contribute to mLabs’ oriented mobile app businesses in emerging and overall effect, even if it had not originally been envisioned frontier markets8. The program does this by enabling under the CSBKE program. Namely, the assessment entrepreneurship in the mobile industry, through revealed that (3) mLabs could have notable effects on venture incubation and acceleration, building mobile the host cities’ and nations’ entrepreneurial ecosystems innovation communities of entrepreneurs, investors, (see box 1 for a definition). The three research questions partners, and mentors, and by researching the app that resulted from initial discussions are summarized in economy of local markets. table 1. Broader goals of CSBKE, such as improvements in the knowledge economy or reaching disadvantaged At the core, this report was an effort to test whether mLabs populations with mobile innovations, are covered implicitly had achieved their original goals and if they resulted in under these three types of effects. 5 · Source: The Business Models of mLabs and mHubs – An Evaluation of infoDev´s Mobile Innovation Support Pilots, available at http://www.infodev.org/mobilebusinessmodels. 6 · CSBKE is a public private partnership program implemented by infoDev with support from the government of Finland, in partnership with Nokia. 7 · See http://www.infodev.org/workprogram. 8 · infoDev’s complete Work Program is at http://www.infodev.org/workprogram, and more information on the Mobile Innovation Program can be found at http:// www.infodev.org/mobile 9 · Source: CSBKE project documentation. 10 · Although this is not stated explicitly in the excerpt from CSBKE program documentation, the goal was to improve mobile application enterprise competitiveness in the early stages 10 of innovation, including up from the idea stage (“From Mind to Market”). This is where infoDev and others identified a support gap. Also see the rational for mobile entrepreneurship enablers in the Business Model Evaluation, available at http://www.infodev.org/mobilebusinessmodels. BOX 1: THE ENTREPRENEURIAL ECOSYSTEM WHAT IS AN ENTREPRENEURIAL ECOSYSTEM? In this report, the term entrepreneurial ecosystem refers to all organizations, individuals, and institutions with an influence on growth-oriented technology entrepreneurs that focus on early- stage mobile application innovations, as well as the roles of these actors and the interactions between them. SSMENT CH QUESTIONS FOR THE ASSE TABLE 1: THE THREE RESEAR EFFECTS ON STARTUP DEVELOPMENT IMPACT ECOSYSTEM CREATION AND THROUGH THE USAGE OF IMPACT DEVELOPMENT MOBILE APPLICATIONS What has been What is the development What have the contribution of impact from the usage been the mLabs to the creation of mobile applications effects of and success of and software developed mLabs on mobile application by entrepreneurs and entrepreneurial enterprises? innovators supported by ecosystems? mLabs? The assessment is complementary to the earlier description of the rationale and original conception evaluation The Business Models of mLabs and mHubs of the program, an extensive discussion of lessons – An Evaluation of infoDev’s Mobile Innovation Support learned, reflections on future directions for mLabs Pilots11. The report analyzed the business models that and infoDev, detailed case studies, more information mLabs and mHubs12 implemented, putting the focus on the operational models of mLabs and mHubs on how they generated value for stakeholders and (including mLab East Asia and four mHubs not income for themselves. This Outcome Assessment covered here), and descriptions of challenges and enhances the analysis of economic effects conducted failures, as well as additional contextual information. in the Business Model Evaluation, and adds an analysis of the development impact through the use of The remainder of the report is structured as follows: mobile applications, as well as effects and influence Chapter 2 describes the assessment methodology of mLabs on local entrepreneurial ecosystems. and chapter 3 gives outlines of mLabs’ intervention context, mainly mapping the three analyzed At the same time, this report will not cover findings entrepreneurial ecosystems. Chapter 4 includes the and comprehensive background information on the main findings of the assessment with one section mLab program, which were already described in for each examined effect type, namely the economic the Business Model Evaluation. Indeed, it is advised results of mLabs; the development impact generated that readers of this report first review sections of through the use of applications developed by mLab- interest in the Business Model Evaluation to be able supported developers and entrepreneurs; and mLabs’ to put the analysis in the present assessment into impact on entrepreneurial ecosystems. Chapter 5 context. Namely, the Business Model Evaluation summarizes the findings and includes concluding includes budget information and financial plans, a remarks. 11 · In the following, we will refer to this report simply as the Business Model Evaluation. It is available at http://www.infodev.org/mobilebusinessmodels. 12 · mHubs, or Mobile Social Networking Hubs, are another type of mobile entrepreneurship enabler implemented under CSBKE, focusing on mobile innovation community building more than on incubation and startup support. 11 3. ASSESSMENT METHODOLOGY Approach and Rationale The assessment followed a qualitative in-depth Data Collection assessment methodology based on interviews, focus groups, and observation techniques. A Primary data came from face-to-face interviews and qualitative approach was considered to be the most group discussions, as well as direct observation of effective to find answers to the posed research events. In total, 116 participants were interviewed, questions: First, the aim of the assessment most of them in semistructured in-depth interviews is to understand the nature of the outcomes, and focus groups. Interviews were conducted and for this, qualitative methods are suitable during four-day field visits in Armenia, South because they generate information with a level Africa, and Kenya. The interviews also generated of depth that purely quantitative methods cannot primary qualitative data for in-depth entrepreneur generate. Second, qualitative methods have and mobile application case studies, published been considered suitable to identify unexpected separately from this report, while 14 section 4.2 assessment outcomes, as well as complex effect includes excerpts. Annex B includes the list of all interactions. the people interviewed, organized according to these five groups. Field researchers asked questions such The assessment followed a contribution analysis as “Do you think the mLab makes a difference for [a approach 13 . To assess mLabs’ effects beyond the given startup’s] success? What factors do you think immediate realm of their activities, the Outcome made a difference?” or “Have community-building Assessment went through three process steps, initiatives lead to the creation of new start-ups? If drawing on the entrepreneurial ecosystems so, how?” as boundaries for the analysis (see box 1): The identification of main actors and subsystems Secondary quantitative and qualitative data came within the ecosystem (for example, innovation from monitoring records and statistical sources, as or startup creation) was followed by an analysis well as a desk review. Also included were secondary of roles and positions of mLabs within the qualitative data from interviews conducted for the ecosystems, before the actual assessment Business Model Evaluation and an evaluation of of mLabs’ contribution to the quality and mLab East Africa conducted by the University of effectiveness of the ecosystems was completed. Nairobi. This analytical process resulted in a mapping of each ecosystem, the identification of key actors Sampling and linkages between them, and ultimately an analysis of mLabs’ contribution from diverse The sampling method used was purposeful sampling, perspectives and parts of the ecosystem. The that is, nonprobability sampling, a method usually ecosystem maps for each mLab and the relational applied when the objective is to study information- flows among actors are presented in chapter 3. rich cases in-depth. Table 2 gives an overview of the interviewed participant groups. 13 · In order to address the questions: What difference did mLabs make? and What would have happened if the mLab had not been in place?, it was attempted to compare reports from client entrepreneurs with those from entrepreneurs that did not use mLab services. In other words, the assessment made a “best effort” to incorporate a counterfactual, even if only qualitatively, to be able to grasp the difference between overall effects and the unique effects of mLab interventions. However, the researchers were not able to construct a large enough, comparable counterfactual group to make claims about mLab’s effects on startup creation and development. Instead, the interviews from the counterfactual group were used to corroborate findings on mLabs’ effects on entrepreneurial ecosystems. Therefore, the assessment overall mainly draws on qualitative statements made by client entrepreneurs and other stakeholders (summarizing their judgments whether mLabs made a difference), and the counter-factual will not be referred to explicitly in the assessment. 12 14 · Mobile startup and entrepreneur case studies are available at http://www.infodev.org/mobilecases. TABLE 2: SAMPLE GROUPS CLIENT COMPARISON COUNTERFACTUAL GROUPS GROUPS INDIVIDUALS People directly Organizations providing Entrepreneurs, benefiting from the similar services to those startups, and businesses services provided by provided by the mLab and that have developed mLabs, for instance, their clients, for example mobile applications and entrepreneurs, intermediary organizations businesses without having startups, trainees, and beneficiaries (startups drawn on business support and event and trainees) of their services, such as incubation participants. programs. and acceleration. ECOSYSTEM END ACTORS USERS All actors populating the Individuals and businesses benefiting entrepreneurial ecosystem not from the use of mobile applications included in any of the other four or from an improved ecosystem, for categories, such as universities, mobile instance, mobile application users operators, original mobile equipment and their communities, enterprise manufacturers (OEMs), investors, customers using the applications training companies, developed, and others. government agencies, and others. A comparison between the range of actors in the ecosystem and the range of people and organizations interviewed gives an estimation of the coverage of the Outcome Assessment in each country. The relevance of the findings in section 5.3, which analyzes the influence of mLabs in the ecosystem, is corroborated by interview participants covering nearly all relevant stakeholder groups in the three countries analyzed. Limitations As a post-hoc analysis and given the complexity of the independent of each other). The assessment also effects that were of interest, the assessment does not faced inconsistencies in quantitative data collection provide a strict attribution analysis that would satisfy as well as absence of data in some areas of analysis. rigorous standards of impact evaluation. Construction Also, the overall time frame for the assignment of an immediately comparable counterfactual and (see annex C) was considered short to perform derivation of statistical information was impossible a comprehensive, in-depth analysis. Finally, the given the timing and scope of the project as well as the field researchers encountered a certain degree of intricate interactions of stakeholders in entrepreneurial evaluation fatigue and reluctance to disclose data on ecosystems (implying that the observed groups are not the part of startups. 13 4. mLABS IN THE CONTEXT OF ENTREPRENEURIAL ECOSYSTEMS The three mLabs covered in this assessment interact with idiosyncratic local entrepreneurial ecosystems, with consequences for a given mLab’s strategy, business model, and results. This chapter outlines the local specificities of entrepreneurial ecosystems and outlines each mLab’s positioning within them. In this way, the chapter constitutes the analytical backdrop for the ensuing analysis in chapter 515. In this section, a brief paragraph will outline the ecosystem of each country, followed by a snapshot of the respective mLab’s role in it. 15 · This chapter is intended to provide a snapshot of mLabs’ activities and positioning within local entrepreneurial ecosystems, as well as the broader linkages between actors. For a more elaborate description of mLab services and an analysis of their business models, refer to the 14 Business Model Evaluation, available at http://www.infodev.org/mobilebusinessmodels. SOUTH AFRICA mLab Southern Africa and the South African Entrepreneurial Ecosystem The mobile technology market in South Africa is mature Within this ecosystem, mLab Southern Africa has positioned compared to other developing countries, characterized itself as a pre-incubator that provides support to mobile by a large number of established companies and developers and entrepreneurs through a number of core multinationals dominating the ecosystem. Yet, a number services. The mLab provides infrastructure that encourages of small “pockets of innovation” in subnational regions innovation of young entrepreneurs (for example, application are not well-connected and have different strengths testing devices) and connects them with the existing and weaknesses. For example, in Johannesburg, the industry. Notably, it acts as a convener between provincial ecosystem is built around the business community; in and national government agencies based in Pretoria and Pretoria, it is built around government; while most startups grassroots entrepreneurs and ICT developers. It has also are located in Cape Town. The ecosystem is growing at a begun to interconnect innovation pockets in different areas fast pace, with stakeholders covering not only the mobile of South Africa, namely the Eastern Cape Province, startup and ICT industry, but also government, academia, and the communities in Cape Town, and mobile companies in software developer community at large. Johannesburg (see figure 1 and table 3). 15 FIGURE 1: MLAB SOUTHERN AFRICA ECOSYSTEM Innovation Hub Startup Fund Government Gauteng Province government Department of Science and 88 Government Technology (DST) MPH Incubators CSIR Meraka Institute The Innovation Hub Industry-related Academia-related Silicon Community-related Cape Initiative Public/Private nonprofit mLab Cape IT SOUTHERN AFRICA Initiative (CITI) INCUBATION INCUBATION BANDWITH MAXUM BARN PRETORIA CAPE TOWN SAINe Innovation Network JoziHub Johannesburg BB AppsLab Industry Academia Nokia Tshwane University of Microsoft Technology Qualcomm University of Pretoria Blackberry North West University Vodacom Ericsson Community Students & Startups 16 GOALS To support entrepreneurs up to a stage where they create startup businesses and launch their products to the market, and reach out to other incubators and regions in South Africa. To connect young developers with the industry. SOUTH mLab strategy and positioning in the ecosystem AFRICA NEEDS OF THE MLAB MLAB MLAB APPROACH PERCEIVED IMAGE ECOSYSTEM SERVICES TARGET GROUPS TO STARTUP OF MLAB IN INNOVATION ECOSYSTEM Difficulty to access Outreach, community Young entrepreneurs Offering one-on-one Good brand private sector building and skills and students. mentorship through reputation and finance; lack of setting through a Programmers who entrepreneur in recognition. skilled professionals; structured program develop applications. residence. missing connection (Ideas Lab, Acceleration between grassroots Lab, and Launch Entrepreneurs who Development of a vertical tech communities Lab) that prepares have skills other model that allows and industry. entrepreneurs for the than programming the mLab to plug into market or follow-on (designers, and existing incubators and support from other others.) innovation spaces with incubation centers. mobile tech-specific incubation services. mLab interventions and services MENTORING, ACCESS TO COACHING, AND NETWORKING BRANDING INFRASTRUCTURE MARKET/ TRAININGS FINANCE Business mentoring Brokerage of Brand affiliation: Provision of core Effort to leverage the and coaching through connections and infoDev/ infrastructure startups’ resources Ideas, opportunities, World Bank, (office space, and strengths. Acceleration, and connecting young Innovation Hub, connectivity, Launch Lab. developers and Gauteng province meeting room, and “Plug” into existing incubatee startups government, CSIR app testing facility). incubation centers. with potential Meraka Institute partners, customers, and private sector Connection of industry, and others. (for example, Nokia, startups with Microsoft, Qualcomm, available government Blackberry, funding. Vodacom). 17 mLab ECA and the Armenian Entrepreneurial Ecosystem Armenia has a long-standing engineering culture The ecosystem is currently in need of specialized inherited from Soviet times. The independence from and highly skilled mobile developers to be able to USSR in 1991 and subsequent years of transition compete with the global market, given the country’s were difficult, but the government’s assertion of ICTs small domestic market. The focus of mLab ECA is, as national priority in 2001 gave new impulses to the therefore, on investing in the early-stage innovation sector, which is currently undergoing a profound and high-potential entrepreneurs aiming at long- renewal. Yet, the mobile application sector is still term results and impact. nascent and started to grow substantially only in 2010 with increasing smartphone penetration. 18 ARMENIA In this endeavor, mLab ECA has benefited from the support and contact brokerage of its mother organization, Enterprise Incubator Foundation (EIF). It has also connected and complemented existing ICT developer training facilities, reached out to the few established startups, and started to support the local innovator community through events and competitions (see figure 2 and table 4). 19 FIGURE 2: MLAB ECA ECOSYSTEM Government Incubators Industry-related Academia-related Community-related Public/Private nonprofit mHub Moldova Government Deputy Minister of Transport mHub and Communication Georgia Ministry of Economy Enterprise Nonprofit Incubator CRDF Foundation Kolba Labs Eurasia Foundation Public Journalism Club mLab ECA Academia Armenian Engineering Uni- - versity, European Academy for ICT, GITC, MIC, UITE Industry Nokia Google Community Orange Students & entrepreneurs Nikita Intel IBM Microsoft 20 GOALS Building skills in mobile technology and stimulating creation of innovative new mobile applications and services to bring them to market. To have talented entrepreneurs committed to promote startup creation. mLab strategy and positioning in the ecosystem ARMENIA NEEDS OF THE MLAB MLAB MLAB APPROACH PERCEIVED IMAGE ECOSYSTEM SERVICES TARGET GROUPS TO INNOVATION OF MLAB IN ECOSYSTEM Talented local mobile Skills development, Students and Stimulating and creative Workplace for app professionals, startup creation and graduates with no environments for young young people strong mobile startup support, incubation, and work experience and and motivated people to to develop their communities. networking. people interested in develop ideas. skills. improving skills and working on mobile “Bottom up”: attending app development. young students with little entrepreneurship experience to create a startup community. mLab interventions and services MENTORING, ACCESS TO COACHING, AND NETWORKING BRANDING INFRASTRUCTURE MARKET/ TRAININGS FINANCE Provides advice Provides applied skills Brand affiliation to Office space, Helps teams to secure and connections in development for all other training centers connectivity, grants and small Armenia and ECA interns and direct and World Bank. computers (Mac seed investments (for region. mentorship for team and PC), mobiles instance, from Enterprise leaders. (Android, iPhones Development and Market Online newsletter and Microsoft), and Competitiveness (EDMC) twice a month. meeting rooms. or Civilian Research & Development Foundation Number of (CRDF) Global). community building activities. 21 mLab East Africa and the Kenyan Entrepreneurial Ecosystem The ecosystem in Kenya is relatively young but has been Since its launch, mLab East Africa has complemented thriving in recent years. Several innovation hubs and the iHub’s services by establishing a training program incubators have been launched since the establishment of in collaboration with eMobilis, running the annual the iHub in March 2010, which has taken on a dominant role regional startup competition Pivot East, and a core model and convener function. The enabling environment incubation and mentoring program for resident for mobile entrepreneurs—albeit hampered by typical startups. Through these services, the mLab reached constraints of low-income countries—is sound: favorable both more advanced, growth-oriented entrepreneurs governmental policies, foreign investment, stable economic and early-stage innovators and developers. conditions, and a growing demand for mobile applications. 22 KENYA Also through community events, the mLab aims to broker partnerships and access to additional support resources for its clients, interconnecting them with a large array of actors in the ecosystem. These actors range from mobile industry players with offices in Nairobi to other accelerators and innovation hubs such as 88mph, Savannah Fund, or Nailab (see figure 3 and table 5). 23 FIGURE 3: ECOSYSTEM MLAB EAST AFRICA Government Incubators Industry-related Academia-related Community-related Public/Private nonprofit Investors IDA World Bank Kenya ICT Nailab Pre-incubation Growth Hub Board 88mph Accelerator Industry Investors Safaricom eMobilis (Seed Stage) Nokia Samsung Savannah Fund Google Helios Investment Airtel Innovation for Africa Intel Partners Microsoft Rio Africa GSMA Victoria Solutions mLab EAST AFRICA Incubation, pre-incubation iHub iLab, iBiz, C4D Academia Community Strathmore University Students, developers, University of Nairobi Jomo Kenyata University enterpreneurs 24 GOAL To facilitate demand-driven innovation by regional entrepreneurs, ensuring that breakthrough low-cost, high-value mobile solutions can be developed and scaled-up into sustainable businesses that address social needs. mLab strategy and positioning in the ecosystem KENYA NEEDS OF THE MLAB MLAB MLAB PERCEIVED IMAGE ECOSYSTEM SERVICES TARGET GROUPS APPROACH TO OF MLAB IN INNOVATION ECOSYSTEM Access to funding; Incubation and Mobile app Identify innovative Active incubator turn “techpreneurs” acceleration, networking, developers, young solutions and turn with successful and good ideas into and partnership building. entrepreneurs, them into profitable startups. sustainable business. students for businesses. training programs, startup incubatees, Focus on applications and experienced with social development entrepreneurs. impact and mobile payment facilities. mLab interventions and services MENTORING, ACCESS TO COACHING, AND NETWORKING BRANDING INFRASTRUCTURE MARKET/ TRAININGS FINANCE One-on-one business Brokerage of contacts Brand affiliation (to Provision of core Offers exposure and mentoring and and opportunities. the iHub, large tech infrastructure linkages to investors coaching. companies, and the (office space, through networking Creation of World Bank). connectivity, and brokerage. Skills development subcommunities – meeting room, and and startup creation (Wireless Wednesday Private sector especially the app through an intensive event series for sponsorship. testing facility). training program. various sectors). 25 5. ASSESSMENT OF MLABS’ EFFECTS This chapter presents the main findings of the Outcome Assessment and is divided in three sub- of mobile applications commercialized by mLab- sections: The first presents mLabs’ effects on supported entrepreneurs; and the third outlines startup creation and development; the second the mLabs’ effects on local entrepreneurial ecosystems. development impact generated through the use While this chapter outlines detailed technical findings, chapter 6 summarizes the assessment. mLabs’ Effects on Startup Creation and Development 5.1 This section will address the research question: “What has been the contribution of mLabs to the creation and success of mobile application enterprises?” It will identify mLabs’ economic effects, providing quantitative data for the creation and development of startups that have been affected by mLabs. mLabs cannot claim to have caused the full extent Whenever this chapter refers to support by mLabs of the results elaborated on in the following. to entrepreneurs, innovators, and startups, this Rather, the data should be seen as evidence of the refers to one or more services that mLabs have achievements of mLab-supported entrepreneurs provided as outlined in chapter 316. and startups, which, together with the qualitative evidence collected, leads to reasonable conclusions It should be noted that mLabs have implemented about mLabs’ contribution to this success (or lack a wide range of services, with varying degrees thereof). While a simple approach, it follows standard of intensity of support (ranging from a client practice for accelerators and incubators that participating at a networking event to multiyear typically advertise their own success by showcasing incubation, including office space and regular the success of previous clients. Quantifying mLabs’ mentoring). net contribution to the startups’ success—and thus mLabs’ economic impact in a narrow sense—would Yet, to maintain simplicity and given the require long-term, rigorous impact evaluation interdependence of effects from different support or, at the minimum, sophisticated comparative types across client groups, this chapter will techniques and benchmarking. Yet, such methods usually broadly refer to mLab support and not are difficult to implement for interventions such as specify the exact service that a client received, mLabs, which are interdependent with dynamic and unless the evidence showed that a service made a complex entrepreneurial ecosystems. Therefore, an particular contribution. economic impact evaluation is beyond the scope of this assessment. 16 · See the Business Model Evaluation for in-depth reviews of each mLab’s value proposition, as well as reflections on the differences in the effects between different models. The Business Model Evaluation is available at http://www.infodev.org/mobilebusinessmodels. 26 The following analysis will provide few remarks on This section mainly focuses on aggregate results high-level findings, followed by considerations of the numbers and few quotes from qualitative data. following four major indicators that indirectly reflect mLabs’ effects and performance: Section 5.2 and in-depth startup case studies published separately from this report17 provide • Startups created more detailed information on mLab-supported • Revenue generated by supported startups entrepreneurs and illustrate the results and mLabs’ • Jobs created by supported startups effects. • External investment raised by supported startups Within less than three years, mLabs have managed to generate a positive contribution to the economy in terms of revenue generation and job creation of supported startups. High-Level Findings on mLabs’ Contribution 5.1.1 to Startup Creation and Development Within less than three years, mLabs have contributed examined mLabs. The contribution of mentoring, to an increasing number of new startups through coaching, and training has been less evident their interventions. Overall, the assessment finds as a contributing factor for incubatee startups, that startups supported by mLabs have managed especially for mLab East Africa. to generate a tangible contribution in the economy as reflected in the positive evolution in several The more indirect effect in Kenya could be explained indicators. The collected qualitative evidence by the higher maturity of incubated startups. The suggests that mLabs can claim to have contributed mLab’s contribution consisted of assistance with to this effect. However, the evidence is not entirely networking, branding, and infrastructure provision. conclusive, the net effect size cannot be quantified Branding played a key role as it intensified the with the means used for this assessment, and the possibility of generating networking among actors. size of the contribution varies across contexts and For instance, Sleepout, a startup with an already cases. successful product and established brand name, sought to be part of the mLab explicitly looking The Outcome Assessment addressed first whether for brand association to expand networking and startup creation and development is in any way investor alternatives. Colocation and institutional related to mLabs’s support. Second, in case of ties with key partners (in particular, the iHub) gave a relationship, the assessment identified which the mLab traction. mLab activities contributed to startup results. Qualitative evidence suggested, first, that there Startup development had a more direct link to is a link between mLabs and startup creation and the mLab’s interventions in South Africa, because development, and thus that mLabs are effectively revenue increases were often associated with contributing to startup results. Second, the business deals made among start-ups within the assessment concludes that the mLab contribution mLab, or with deals made between startups and varies by the mLab’s environment, business model, external actors thanks to mLab mediation and and service type. credibility. For mLab ECA, the mobile application enterprises’ revenue models and employment are Networking, provision of infrastructure, branding, also very directly linked to the mLab, given that it and access to markets and financing contributed plays a key role in the marketing of the applications, to startup creation and development for all three including their branding. 17 · Mobile startup and entrepreneur case studies are available at http://www.infodev.org/mobilecases. 27 5.1.2 Startup Creation Key Findings • mLabs contributed to the creation of an increasing number of new startups through their interventions. • As of October 2013, all startups that graduated from incubation programs were still active18. 2011 2012 2013 Numbers for 2013 FIGURE 4: NUMBERS OF STARTUPS CREATED BY MLABS are until June/July 2013 (% Yearly increase) 50 (316%) 12 12 8 (50%) 6 5 (100%) One of the main goals of mLabs had been the creation At the time of this analysis, all startups incubated and development of startups. The assessment by mLab ECA and mLab Southern Africa were still finds that mLabs have indeed contributed to the in business, and the ten startups that had already creation of an increasing number of new startups graduated from mLab East Africa were also active through their interventions. Figure 4 includes the as of October 201320. available results for new startups created in 2011, 2012, and 2013 at each mLab19. 18 · This statement refers only to the incubated startups and not to all the startups created with support of mLab services. “Active” is defined as at least one founder still refining the startup’s business model and working toward the commercialization or expansion of the startup’s mobile application product. In Kenya, the startups graduated from the incubation program (all of them active) were 10 out of the 50 shown in the graph. 19 · Startup creation figures are derived primarily from reporting by mLab managers. Managers were instructed to only count businesses that have formally registered. It should be noted that the burdens of registering a business vary substantially across the three countries analyzed, which makes benchmarking difficult. In internal reporting, infoDev also used the measure of “solid startup teams” to account for stable startup projects that avoid registering as a business. See the Business Model Evaluation for high-level results for each mLab, available at http://www.infodev.org/ mobilebusinessmodels. 20 · The fact they are all active is not an ultimate indicator of success as it is still quite early to judge whether they are sustainable/growing businesses. 28 Five-year time frames are usually used to assess this aspect. The figure for startup creation for mLab East Africa was much higher during the third year of operation as it includes not only startups participating in the incubation program18 but also startups that were created during the Pivot East competition. Conversely, mLab ECA and mLab Southern Africa figures include startups participating in the incubation program only. mLab Southern Africa supported the creation of 12 “ I met with the mLab manager who asked me to develop a prototype of my startups in a rather short time period. Given that mLab idea. Based on this Southern Africa focuses its support on pre-incubation, the majority of these startups are still in the product prototype, I was idea and business skills development phase. signed on at the “ “ mLab. (An entrepreneur at mLab Southern Africa) When I had the idea of MySales, I shared it with the mLab manager and other team members, who Although it started its activities in September 2011 as a program from Enterprise Incubator gave support immediately, Foundation (EIF), mLab ECA effectively started its incubation work only in May 2012, for example, by sharing before finally settling down in its current, expanded premises in June 2013. From this lots of ideas and designing point on, the mLab increased its support to entrepreneurs, which resulted in the creation the business model. The of five startups within a year. mLab then helped develop A further increase in the number of startups the website for our created can be expected in the coming years, given the trend of increasing startup creation platform and [together] numbers for each of the three observed mLabs and assuming that the first year of with the [other] mLab implementation (when mLabs are themselves startups) tends to be the most difficult. team members, I could Moreover, the ecosystem in Kenya is begin developing my app. expanding rapidly with a continuously growing pipeline of technology entrepreneurs, and I started my career in in Armenia and South Africa, the mLabs’ [the] mLab. Now I have a startup and I am more confident. “ current and planned capacity building efforts will most probably pay off in terms of more startups being created as well. (An entrepreneur at mLab ECA) 29 5.1.3 Revenue Generation by Supported Startups Key Findings • mLab-supported startups have increased revenue generation through 2012 and 2013. • Total revenue generated by m-Labs supported startups in Kenya increased four-fold in less than two years. • mLab Southern Africa has created an internal economic market place, in which entrepreneurs provide services for each other. FIGURE 5: REVENUE GENERATION BY MLAB-SUPPORTED STARTUPS AND ENTREPRENEURS (USD) 2011 2012 2013 Numbers for 2013 are until June/July 2013 (% Yearly increase) 627,000 (342%) 345,000 283,000 141,727 (67%) 55,100 (14%) A second indicator of mLabs’ positive revenues at the time of joining the mLab with performance is that their incubatee startups’ their current revenue: the aggregate revenue revenue generation has increased over time for the ten graduated incubatees went from across the board. In Kenya and Armenia in an initial $141,727 when joining the mLab to particular, the startups’ revenue streams have $627,000 in June 2013. In the case of mLab ECA, increased significantly during the period. entrepreneurs’ revenue at this point in time was at about $55,100. For mLab Southern Africa, For mLab East Africa, data broken down by startup revenues increased from $283,000 in company was available for the analysis so 2012 to around $345,000 in 2013. that it is possible to compare the incubatees’ 30 These revenues have been generated mainly Qualitative data indicate that these increases in by the launch of successful applications. Two startups’ revenue generation are due to mLabs’ entrepreneurs have launched their applications contribution, while the evidence was not entirely to the market with considerable success, with the conclusive. The assessment found it difficult to additional support of incubators that the mLab is discern a clear causal effect, given the complexity partnering with (Maxum incubator of the Innovation of entrepreneurial ecosystems that both the mLab Hub in Pretoria and Bandwidth Barn in Cape Town). and its clients operate in and the many other factors that determine a startup’s development. In The added value of mLab Southern Africa in the pre- addition, interviewees’ opinions about how strong incubation phase is the provision of an innovative the mLab’s contribution was covered a wide range. collaborative working space for incubatees with It remains that infoDev should pay close attention access to application testing devices, one-on-one to the evolution of revenue figures (including coaching, and mentoring services for business through impact evaluation), identify whether the skills and business model development, as well positive signs can be substantiated further, and as networking activities. In its facilities in Pretoria, reflect on expanding the current two-year financing mLab Southern Africa has also created an internal frameworks for mLab grants to ensure the economic space, where incubatee entrepreneurs continued positive evolution of economic results. market services, such as graphic design or programming, can work together21. 5.1.4 Job Creation by Supported Startups Key Findings • mLab-supported startups create an increasing number of high-quality jobs. FIGURE 6: JOBS CREATED BY MLAB-SUPPORTED STARTUPS AND ENTREPRENEURS 100 (79%) 2011 2012 2013 Numbers for 2013 Women are until June/July 2013 (% Yearly increase) 56 51 50 39 37 (31%) 25 (48%) 10 15 (67%) 4 18 21 · See the Business Model Evaluation for more detail: http://www.infodev.org/mobilebusinessmodels 31 High quality job22 creation has been increasing among startups supported by the three mLabs as well. Albeit absolute numbers of direct jobs created by mLab-supported startups should not be expected to go into the hundreds, this significant growth hints at a considerable potential that mLabs can stimulate the creation of numerous high quality jobs. Future studies could also examine indirect job creation, for instance, in sectors that startups draw on, such as business consulting, accounting, design, various parts of the creative industry, or industries “ Several of the people that have been through the mLab have been contracted producing technological inputs (such as local software and device manufacturers). Furthermore, through mLabs’ training and human resource building activities (see section 5.3.1.), mLabs have also improved clients’ job opportunities in cases where they do not opt to start a company themselves. This represents yet another valuable of inquiry for future assessments. to work with us. (Cofounder and CEO of PicsArt, an internationally recognized Armenian mobile application enterprise) “ 5.1.5 External Investment Attracted by Supported Startups Key Findings • Incubation graduate startups have been successful at attracting investments and raising funds. • In Kenya, the mLab’s incubatees have managed to raise more than $1.5 million in investments up to June 2013. FIGURE 7: EXTERNAL INVESTMENTS ATTRACTED BY MLAB STARTUPS (USD) 2011 2012 2013 Numbers for 2013 are until June/July 2013 (% Yearly increase) 1,500,000 (50%) 1,000,000 652,205 625,293 128,700 32,700 15,000 (293%) (118%) 22 · We define high quality jobs as white collar jobs as well as for-pay work that bring exposure to the intense and stimulating environment of a startup and entrepreneurial communities: founders and startup employees are likely to learn valuable skills and gain experience, more than in other white-collar 32 employment. Also refer to the 2013 IFC Jobs Study for a discussion of the difference between high and low-quality jobs, see http://www.fic.ba/pdf/IFC_Job_ Study_Report_12-26.pdf. “ For all mLabs, startups have been rather successful at attracting investments and raising funds. In Kenya, mLab-support startups have raised $1.5 million in investments up to June 2013, although the figure should be interpreted with caution, as the distribution Thanks to the is heavily skewed toward a few startups. Breakdown data by a company for mLab East Africa also reveal mLab publicity and that investment attracted increased for each graduated incubatee during the period that they stayed in the mLab. networking, getting With a cumulated amount inferior to the other mLabs, funds was much mLab ECA’s entrepreneurs are succeeding in attracting an increasing number of external investors, summing easier. Some of $128,700 as of August 2013. A series of startups have benefited from grant seed funding in Armenia (for the funds we got example, Microforester or Clean H20) or in the region came through this (Be Healthy – Uzbekistan). The mLab is also contributing channel, others to the development of an equity investment culture in Armenia, for instance, by taking shares in some of the startups it is supporting. Seven companies supported by mLab Southern Africa have managed to attract a total of $652,205 of external came from outside. (An incubation graduate of mLab East Africa) “ funding. The biggest portions were raised by the two most advanced companies: Afroes and GoMetro. “In the Idea2Market Contest, organized by CRDF Global (an independent nonprofit organization) and the Ministry of Economy of Armenia and facilitated by mLab in Yerevan, we won a grant of $25,500 which made it possible for us to accelerate the process of implementing our application in the country. (An entrepreneur of mLab ECA) “ 33 Development Impact through the Usage of Mobile 5.2 Applications This section addresses the second research question: “What are the development impacts from the usage of the mobile applications and software developed by entrepreneurs at mLabs?”23 Key Findings • Nearly 10 percent (29) of a total 292 mobile applications were identified to have actual or potential development impact. (This sample includes applications not yet launched to market.) • The percentage is relatively low for three reasons: (1) the criteria for inclusion of applications in the sample were based on a narrow definition of development impact, focusing on a contribution to MDGs, BoP users, or transformative potential; (2) the generation of social and environmental impact is not a precondition to access mLab support; (3) at least in the short-term, startups prioritize revenue generation over development impact. • Two factors limited the depth of observed development impact: (1) some applications were operational but had not yet been launched to the market; (2) even when applications had already been launched, the time span elapsed since launch was mostly very short. • The identified applications addressed relevant national challenges and offered practical solutions in the areas of education, health, financial inclusion, agriculture, employment, environment, mobility, and information technology. • Developers and entrepreneurs supported by mLab East Africa generated the largest number of applications in comparison to the other mLabs. This mainly resulted from to the expansive strategy adopted by the mLab in promoting startup creation. Services such as training programs or competitions (in addition to incubation) tend to generate much larger numbers of mobile applications and prototypes, even if this does not speak to the quality and sustainability of these products. • Within the identified applications with actual or potential development impact, there were a few significant success stories, several companies with satisfactory effects, and a majority of applications still in the pilot stages (that is, they are undergoing testing and readjustments previous to a fully fledged launch). The latter group of applications has not yet succeeded in the market but have potential for development impact in months and years to come. • Overall, prospects for wider impact are promising given the high relevance of the problems that the identified mobile applications intend to address as well as the high appropriateness of the proposed solutions. 23 · The answer to the question has taken into account not only mobile applications developed by mLab incubatees but also applications created 34 by developers and entrepreneurs supported through events and competitions implemented by mLabs. Identification Process for Mobile Applications with 5.2.1 Development Impact The total population of mobile applications from which mLabs’ incubation programs (applications developed development impact applications were selected was with direct support from the mLabs); (2) applications a list of 292 mobile applications linked to mLab ECA, developed through innovation competitions and mLab Southern Africa, and mLab East Africa24. challenges, such as the Regional Mobile Applications Contest in ECA and Pivot East in East Africa; and (3) This list was compiled from three data bases of applications developed by participants of specialized applications known by name to mLab managers25, training programs delivered by mLabs. Figure 8 which correspond to the three types of support provided shows the inventory of applications by type of source by mLabs: (1) applications developed by participants in and country. FIGURE 8: MOBILE APPLICATIONS BY MLAB SERVICE TYPE Participants in mLab Participants of specialized Apps developed through innova- incubation programs mLab training programs tion competitions and challenges 10 8 112 20 11 142 Total: 20 Total: 19 Total: 253 There are two reasons for the large number of the availability of the information at the time of applications in Kenya in comparison to Armenia this assessment.27 and South Africa. The first is associated with mLabs’ service portfolio. mLabs aim to promote Three selection criteria were applied to the 292 startup creation and growth using different applications in order to select the sample to be services and activities. Some of these activities analyzed in more depth: result in the generation of a large number (1) Applications that contribute to the Millennium of applications and prototypes (in particular, Development Goals (MDGs). training programs and competitions) whereas (2) Applications that address the needs of users at the others do not (such as incubation focusing on few Base of the Pyramid (BoP).28 enterprises and products). As it can be inferred (3) Applications that otherwise have a transformative from figure 8, mLab Southern Africa particularly potential in terms of improving the general conditions emphasized incubation and did not implement of people and businesses and/or the functioning of training programs or competitions to the extent society as a whole (in particular, applications that may mLab East Africa did26; this is the main reason why have a potential to alleviate poverty or deprivation mLab East Africa features a much higher number suffered by disadvantaged groups of the society). of applications. The second reason is related to 24 · Details on the list of applications per mLab and by source of applications were too extensive to be included in this document, but can be obtained from infoDev upon request. Case studies of notable mobile startups and entrepreneurs are also available at http://www.infodev.org/mobilecases. 25 · More mobile applications and application prototypes have been created at hackathons, workshops, and so forth, but not all are documented and known to mLab managers. 26 · mLab East Africa ran three editions of the regional startup competition Pivot East (yielding nearly 50 applications each) and five waves of training with approximately 20 participants each. 27 · In Armenia, the mLab’s training programs are producing mobile applications; however, these applications were not included as the data were not available at the cut-off date for this assessment. 35 28 · http://www.ifc.org/wps/wcm/connect/as_ext_content/what+we+do/inclusive+business/news+and+highlights/defining+the+base+of+the+pyramid. From here on in this assessment, development subsequently analyzed in detail. impact will refer to these three aspects. Spillover Data on impact were collected (for each application) or indirect secondary effects generated by mobile on the following of variables: application enterprises such as spurring innovation processes and creating employment as a side effect (1) Actual or potential impact, of producing and selling the application have not (2) The size of the problem (quantitative and been included as development impact. 29 qualitative), (3) Types of users (such as students, farmers, and Forty-one mobile applications with either potential SMEs), or actual development impact complied with the (4) Focus on Business to Business (B2B) or Business three criteria.30 This assessment also includes to Consumer (B2C), applications with potential development impact (5) Whether the application targets BoP users, to maintain a greater depth of analysis and avoid (6) Geographical and sector scope, limiting scope to only few of the total number (7) Application usage and reach (such as downloads, of applications, as this would imply the risk of subscriptions and user numbers), drastically underestimating the development impact (8) Period of operation, and that may ultimately be generated. (9) An assessment of the difference made in a user’s life (quantitative and quantitative). Twelve applications had to be disregarded because of substantial lack of information to conduct the The sources of information used for the collection analysis (such as absence of information on the and analysis of these variables were desk research, functioning and impact of the app, unavailability data from in-depth interviews conducted for case of entrepreneur contact details, discontinued app, studies during country visits, and remote follow-up etc.). This left a sample of 29 applications that were interviews to fill data gaps. 5.2.2 Characterization of the Sample of Identified Applications Of the 29 selected mobile applications with actual or did not apply a broad definition of development but potential development impact, two are in Armenia, one that would only include applications focused six in Southern Africa, and twenty-one in East Africa. on MDGs or users at the BoP as well as those with As shown in figure 9, the applications cover eight transformative potential. A broader definition would sectors: health (5); education (6), including nonformal have meant a greater number of selected applications. education and awareness-raising; environment (3); Second, the generation of social and environmental agriculture (6); financial inclusion (6); employment impact is not a precondition to access mLab support. (2); mobility (2); and grassroots innovation (1). Developers are held to pursue business ideas that Applications are double-counted when they cover could be successful from a commercial point of view more than one sector. 31 but not necessarily generate impact according to the applied selection criteria. Third, while many startups When comparing the total population of applications would like to develop mobile applications that have with the number of selected applications, the selected development impact (some have specific plans to do sample covers 11 percent of the applications in so), the need to generate revenue in the short-term to Armenia, 8 percent in Kenya, and 30 percent in South sustain their business is prioritized over development Africa. The percentages are relatively low for three considerations. reasons. First, the criteria for selection of applications 29 · However, employment creation has been incorporated when it was the explicit goal of the application (micro-work), and when employment creation particularly targeted the BoP. 30 · The remainder of the total 292 applications were either not related to usage that would typically lead to development impact as defined here, or they were clearly left at the prototype stage after ad hoc development at an event. 31 · As shown in the figure this is the case for Microforester, which covers environment and employment, and Dairy Sacco, which covers financial 36 inclusion and agriculture. mLab ECA FIGURE 9: OVERVIEW OF THE SELECTED APPLICATIONS mLab East Africa BY COUNTRY AND BY SECTOR mLab Southern Africa Financial Education Inclusion Health Employment Eneza (a) MyLife (p) eLimu (a) mPayer (a) MedAfrica (a) MobiDev (p) Smart Balckboard (p) Uhasibu (a) Idaktari Kopo Kopo (a) M-Chanjo Somesha Haki (a) Changamka Moraba (a) Dairy Sacco Microforester (p) Champ Chase (a) Clean H20 (p) IT Grassroots Mobility Agriculture Environment Innovaton Dairy Sacco Fishmate (a) mFarm (a) Gorecycler AftaRobot (a) GreenHouse Pro (a) mKulimabuma (a) Geekulcha (a) GoMetro (a) FarmPal (a) FarmGuide (p) Microforester (p) Note: Actually launched applications are marked with (a); potentially launched applications are marked with (p). Applications where the available information was insufficient to assess the innovation stage have no indication. FIGURE 10: OPERATIONAL LIFE OF THE 29 SELECTED APPLICATIONS Out of the selected applications, at least 66 percent (19) are operational, meaning that they are either undergoing testing or have already been launched to market. Ten 5 operational applications have been active for one to 10 three years and nine for a year or less; none of the 19 5 applications have been operational for more than three years. 9 Five applications out of the 29 selected have not been launched, of which two have been abandoned, and for the remaining five applications, insufficient information > 1 Year Not Launched is available to assess their operational status. ≤ 1 Year Unknown 37 Assessment and Illustration of Development 5.2.3 Impact through Application Usage Given the diversity in terms of maturity and Quantitative data, such as usage numbers, sector coverage of the identified applications, are included where available and meaningful. the assessment will not include estimates Because of space and scope limitations, not of the aggregated quantitative effects that all analyzed applications are covered in depth applications have had. Instead, the assessment in this report but documentation of each of the will highlight general trends in the selected applications can be obtained upon request from sample before illustrating development impact the infoDev team. Detailed mobile startup and through brief descriptions of notable cases entrepreneur case studies and success stories of actual or potential development impact.32 were also published separately.33 General Trends among Applications with Development Impact Overall, the selected applications followed a interviewed entrepreneurs. Typically, a prototype distribution where there were few significant is developed prior to market launch, which is success stories, several companies generating uploaded to an application store or Internet satisfactory effects, and a majority of applications platform in the case of smart phone applications in pilot stages, that is, they were still undergoing or made available as an SMS or USSD service testing and readjustments previous to a full- for feature and basic phone applications. Often, fledged launch. The latter have not yet succeeded during this phase previous to commercialization in the market but feature good potential for to end-users, application development might development impact if and when they launch in face technical difficulties or there might be the the coming months and years. realization that the application does not yet tally with actual market needs. At this point, several The depth and breadth of the actual, observed startups decide to conduct further market impact is limited because of two factors: (1) some research to better understand the results of the applications have not yet been launched, so only testing phase and improve previous versions. potential impact can be analyzed, and (2) even This phase is critical because it requires startups for the already launched applications, the time to further invest before they can generate span since the launch has been relatively short. revenue. Fishmate, Gorecycler, Aftarobot, Green mLabs have been operational for two years, a House Pro, and MobiAgent are examples of this rather short period of time for the development approach. impact of the usage of mobile applications to become manifest. At the time of the assessment, In fact, a recurrent aspect found when some of the applications with good prospects to interviewing entrepreneurs was the need to put make a difference were at a transitional phase. in place systems to track the impact of their Most interviews revealed that even when impact applications, as impact-monitoring systems is still at initial stages of the expected impact would enable them to learn about the uses of sequence, many applications are on the path to the application and improve its features and the make a difference, even if they currently report associated business model. Most startups did only modest levels of market penetration. not have these systems in place and required support in setting them up. Consequently, the To understand such delays in impact generation, design and technical support for impact-tracking it is useful to consider the typical launch systems for startups is an aspect that mLabs process of B2C applications as described by could consider integrating in their programming. 32 · The applications considered in the analysis are those that got beyond the prototype stage, that is, they had been fully developed, even if they were not launched to the market in some of the cases. 38 33 · They are available at http://www.infodev.org/mobilecases. Another element influencing the timing for This means that applications have to go through a impact generation are barriers of awareness process of technical adaptation, as well as gaining that stand between entrepreneurs and low- confidence and trust, which can take time and income user segments with potentially delay the generation of impact. Fishmate, Green poor knowledge of mobile technologies34. House Pro, GoRecycler, Uhasibu, Afta Robot, For instance, users may find it difficult to get used and mKulimabima are examples of applications to buying and selling through applications in the experiencing this situation. case of mobile payment applications. ARMENIA Applications with Development Impact in Armenia The main national development goals supported Project, Armenia will be left with no forest by the in the World Bank’s Country Partnership Strategy year 2030 if the current rates of deforestation include reducing inequality, ensuring human continue. Microforester has already gained nine development and accelerating economic growth. planting permissions in the country, which points In 2008, the government adopted a ten-year to good traction. The application shows good industry development strategy including the prospects to contribute to reforestation, building development of knowledge economy as a critical environmental awareness, and decreasing poverty point to form a strong and advanced information and unemployment through microwork. technology sector. 35 Clean H2O aims to allow community leaders, The two selected applications in Armenia are Clean individuals, and companies to measure the quality H20 (health) and Microforester (environment, of drinking water and generate open data access employment); both are applications with a focus information on water quality. The project integrates on social and environmental impact and potential a strong hardware component: the team designed to produce significant impact not only in Armenia an innovative water device that transmits data but also in other countries. Both applications won to mobile phones. The application is designed to awards but have not yet been launched, while store information in the cloud, contributing to the the size of the problems they intend to address creation of openly available statistics on water and the solutions they offer (business models quality. The application wants to fill the current and technological advancement) hint at good supply gap in water device applications between prospects for impact. cheap disposable water devices and expensive professional ones. Clean H2O constitutes a Microforester is a mobile app combining promising tool for nongovernmental organizations environmental protection (tree planting) with the (NGOs) and international organizations working on creation of job opportunities for communities water and sanitation in countries with particularly in rural areas through ICTs in a context where low access rates to improved water resources. residents often struggle with income generation36 and where only 8 percent of land is covered with forest. According to the Armenia Tree Clean H20 and Microforester have not yet been launched, but they point at promising impact prospects in Armenia and abroad. 34 · Also see http://www.infodev.org/articles/mobile-usage-base-pyramid-kenya and http://www.infodev.org/articles/mobile-usage-base- pyramid-south-africa for studies on user behaviour at the BoP commissioned by infoDev. 35 · http://www.eif.am/eng/resources/report-on-the-state-of-the-industry/. 36 · Armenia had a 35.8 percent poverty ratio in 2010 and a 17 percent unemployment rate in 2013. 39 SOUTH AFRICA Applications with Development Impact in South Africa Poverty reduction, improved access and GoMetro is one example of a startup that entered quality of public services (including transport the mLab as a winner of the Gauteng Innovation improvements), environmental protection, Competition. The company offers a mobile and employment are among the main national service that provides Metrorail commuters with priorities supported by the World Bank’s strategy updated information on travel routes, reducing in South Africa. Notable opportunities for impact delays in users’ daily commutes. on governance and social problems arise from mLab Southern Africa’s partnerships with The app has been operating for more than a year government agencies. and it is gaining substantial market traction and reach. For example, the Department of Science and Technology, one of the major partners of the Three aspects indicate that the impact of the mLab, has a road map for the development application is remarkable: the initial magnitude of mobile applications in key areas, such as of the problem—600,000 people using the education, health and green economy, and Western Cape Metrorail38 and suffering from recently pledged continuous support for the delays in getting to work, school, or back home; mLab37. the considerable size of the user base (450,000 application downloads); and the qualitative The mLab can link entrepreneurs with relevant improvements reported in the life of the users government programs and match developers (such as improved time management for (that do not necessarily have knowledge about Metrorail commuters) and in the quality of the social development challenges) with existing Metrorail service (GoMetro creates a positive problems. An example for the tie-in of provincial feedback loop that improves the quality of the government priorities and innovation enabled by service). See box 2 for details. the mLab over the last few years is the Gauteng Innovation Competition, jointly organized by mLab Southern Africa and the Province of Gauteng, during which developers create solutions to specific topics and existing challenges. GoMetro is a major contribution to closing the information gap between public service providers and the consumer in emerging markets. 37 · http://www.infodev.org/highlights/south-africa-and-world-bank-pledge-ongoing-support-mlab-southern-africa. 38 · Information on Cape Town Metrorail, July 2013: http://futurecapetown.com/2013/08/cape-town-deserves-a-reliable-and-predictable-train- 40 service/#.UsyAnrQf7Mo. BOX 2: GOMETRO FACTS 600,000 people use the app in the Western Cape Metrorail. 450,000 app downloads with a monthly growth rate of 10-15 percent. 90 percent of users reported using the service at least once a week, pointing to the popularity of the app. The number of GoMetro users after only the first 30 days was 19,159. EVIDENCE OF IMPACT To date the application has collected more than 10,000 customer complaints and suggestions (about 800 complaints per month), which has helped Metrorail to improve its operations. According to GoMetro, Metrorail responded to each of these complaints and sought to resolve them. GoMetro has also received 1,150 crime tipoffs through its CrimeLine service. ESTIMATES OF IMPACT Before the rollout of GoMetro, Metrorail services were characterized by cancellations and schedule fluctuations combined with the lack of real-time, consumer-facing information, leaving consumers with uncertainty and strongly affecting their daily activities. Metrorail was perceived as an organization with an indifferent attitude to customers, a problem exacerbated by insufficient communication and misinformation, not allowing users to make alternative arrangements to get to work, school, and home on time. Since the launch of GoMetro, the situation has improved significantly. According to GoMetro, information gaps have been limited substantially: Metrorail keeps users informed through the app, making the commuter experience less of a hassle. GoMetro aims to give commuters a voice through a built-in feedback service, creating a feedback loop that improves the quality of service. The application has helped close the information gap between public service providers and the consumer in emerging markets—a major problem, especially in the transportation sector. GoMetro provides commuters with accurate information to plan their daily activities, which makes Metrorail a more convenient means to commute and results in improved time management for users. No clear estimates are available on the economic benefit generated through time saved for users, but given the depth of the problem before the app launched and GoMetro’s large user base, the impact is likely to be substantial. 41 Aftarobot, another app focusing on mobility Aftarobot aims to cover 32 to 40 percent of the and transport, entered the mLab as a result market of taxi associations, which cover 65 of the Gauteng Innovation Competition. percent of all public transportation and serve 14 million passengers daily with an estimated Whereas GoMetro improves Metrorail’s 120,000 vehicles, thereby constituting the mass railway transit, Aftarobot’s attention primary means of transport in South Africa. is on minibuses—also referred to by South If Aftarobot’s plans succeed, the aggregated Africans as taxis. The app is currently in economic impact of reductions in waiting the pilot project stage39,only operating in periods and improvements in customers’ Gauteng, one of the country’s nine provinces. ability to plan their days could be significant. The app started less than a year ago and its focus is now on data gathering to understand patterns in the usage of minibuses, later informing the delivery of accurate waiting time information to users. The aggregated economic impact could be outstanding should Aftarobot expansion plans succeed. Geekulcha shows promising prospects to contribute to addressing two major problems of the IT sector in the country. community. In response, Geekulcha has built a community of IT professionals connected Geekulcha is one of the illustrative examples with the industry and other IT stakeholders. of solutions that do not directly address Approximately 20 companies have contacted the MDGs or application users at the BoP Geekulcha to headhunt students, offering but has transformative potential. The web- them participation in specific projects. Thirty based app works through Internet platforms students thus far have found jobs as a result where the IT communities get to interact, of their involvement in the community. Data have discussions, debate on IT-related on outreach and immediate results hint at subjects, and share knowledge with each good future prospects: Facebook statistics other. The application, launched in March show 1,200 and 1,700 users in the two 2013, is showing promising prospects platforms, with engagement rates revealing to contribute to addressing two major 500 active users. Geekulcha also organized ten problems of the IT sector in the country: workshops with an average attendance of 30 to its relatively small pool of skilled workers 40 people, and in less than a year 30 students and the fact that the industry lacks efficient obtained jobs through Geekulcha platforms. means to reach out to and connect talented students, as well as the grassroots developer 39 · Among the 30 to 40 routes within one taxi station, Aftarobot is currently studying the frequency of only one route. 42 Afroes is a social enterprise that develops Moraba (Afroes) addresses gender-based educational mobile games that address violence, a major problem in South Africa, important social issues in Africa, such as where according to police records, 154 women gender violence, political violence, and were raped per day between March 2010 and child abuse. Its applications are examples March 2011 and 78 percent of men admit to for transformative potential in MDG-related perpetrating some form of violence against areas. Extensive quantitative data on the women. Moraba, an initiative prompted by impact of the company’s applications are not UN Women, has 58,500 downloads and is available but there is indication that the apps likely to produce behavioral changes. make a difference for users. There are indications that Afroes applications are promoting behavioral changes, which could be sizeable given the number of downloads. “ The most powerful thing I can take from this game is what the young people write in the comments section of the game saying they have learned ChampChase (Afroes) is a game that raises awareness about the safety and security of that they have a children. It started as an initiative of the Nelson voice, that they know Mandela Children’s Fund. they don’t have to The game features 180,000 downloads in a country be taken advantage where, according to UNICEF, 54,225 crimes against children were reported in 2010/2011, with half of of as young people the cases being sexual offenses, and around 30 percent of sex crimes against children involving and that they realize victims under 10 years of age. some of what they Haki 1 and 2 (Afroes) address social rights and justice and the protection of environmental rights, respectively, with Haki 1 featuring 285,000 downloads and Haki 2 having generated 38,000. Both mobile applications aim to change the were previously doing was wrong. (Nathan Muema, software engineer and designer of Moraba) “ mindset of society on cultural practices and seek to empower youth so that they are capable to defend their rights. 43 KENYA Applications with Development Impact in Kenya According to the Kenya’s development There has been progress in the ICT agenda Vision 2030, the country aims to sector: Internet access has increased transform itself into a middle-income from 1.3 million subscribers in 2007 to country, based on enhancing agricultural 14.3 million in 2011, with 80 percent of productivity, access to health care, Kenyan adults owning a mobile phone; education, and basic infrastructure communications costs have dropped services 40. In 2013, the average Kenyan and the use of Internet information had increased access to health services, has increased (CPS Kenya report) 42. education, and social rights, but contrasts and widespread inequalities remain 41. The The mLab has supported the development slow pace of development in small cities, of a number of applications in financial towns, and rural areas is one of the biggest inclusion, education, health care, and challenges in a country where most of the agriculture that can improve users’ quality wealth and access to quality services are of life not only in Nairobi but also in other concentrated in the capital city of Nairobi. cities and rural areas. Given the great number of application cases that will be Many applications are being developed illustrated, the following is divided into the for use in regions of Kenya other than sectors: mobile education, agriculture, Nairobi, mostly applications for regular financial inclusion, health, employment cell phones instead of smart phones. creation, and environmental protection. One of the clearest examples of impact in the taking into account the current coverage of mobile education sector is Eneza Education the application, it could be estimated that, in (previously branded “MPrep”), which offers the best of cases, 33,000 Kenyan students are mobile technology tools to schools in remote likely to improve their grades in all subjects regions. According to findings of the MPrep and 700 schools are likely to increase their 2012 Impact Study Report (which analyzed Kenya Certificate of Primary School scores by three slum schools in Nairobi and measured 15 percent on average (see box 3 for details). impact in a controlled environment) and Through the use of Eneza Education, 33,000 students and 700 schools are likely to increase performance. 40· All of them are supported by the World Bank and included in the Country Partnership Strategy. 41 · http://www.worldbank.org/en/country/kenya/overview. 42 · http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2012/04/13/000386194_20120413011910/Rendered/PDF/672240C 44 ASP0P120Official0Use0Only090.pdf. BOX 3: ENEZA EDUCATION FACTS 100,000 users of which 60 percent are active; 700 primary schools integrated; 33,000 students using the SMS system. EVIDENCE OF IMPACT43 Students using ENEZA have increased their exam scores more than the average student in nearly every subject, by an average of 5 percent overall. 80 percent of students in schools were found to study more when using ENEZA. 67 percent of ENEZA students were found to study more than four hours after school per day compared to 55 percent of students who are not on ENEZA. Schools using ENEZA feature a 15 percent increase of the Kenya Certificate of Primary School scores of schools with high use of ENEZA. ESTIMATES OF IMPACT 33,000 students are likely to increase their grades in all subjects and 700 schools are likely to increase their Kenya Certificate of Primary School scores by 15 percent on average. 700 schools and 33,000 students equal coverage of 2.68 percent of schools and 0.35 percent of students in the country, which is significant for a single mobile application. ENEZA users are not evenly spread around the country; they are concentrated in slum schools, Daadab Refugee camps, and in primary schools of rural areas in Northeastern region of Kenya— areas where the poverty incidence is particularly high—for instance, in the Northeastern region, 55 percent of poor girls and 43 percent of poor boys have never been to school. Therefore, the likely impact on poverty alleviation is even higher. Enrolment, retention, completion, and progression rates are a major challenge and a concern of the MDG on education for Kenya. The percentage of repeaters in primary school stands at a 6 percent and 65 percent of students abandon after primary school. In this context, ENEZA offers a tool that prompts students to study more and could help reduce the incidence of drop-outs and repeaters in cases where these are due to insufficient studying. 43 · MPrep 2012 Impact Study Report, which analyzed three slum schools in Nairobi in order to measure impact in a controlled environment. See http://enezaeducation.com/release-of-mprep-impact-report-2/. 45 eLimu is an award-winning interactive application for children in the Kenyan Primary School education system, helping them to study and review for exams and making studying more enjoyable. Although quantitative data was not available at the time of the assessment, several factors “ Our curriculum in Kenya is like a punishment to children, they indicate that this application could have a tangible impact. The application features an feel they have to individual follow-up process that motivates kids to continue schooling, in a country do it because it’s context with a 40 percent dropout rate, a 45 to 1 student-teacher ratio, and 10 percent of the compulsory. […] standard eighth-grade students are not able With these tablets to solve second-grade math problems. [featuring eLimu], Despite nearly 100 percent of kids enrolling in primary school, less than one third of primary school pupils possess basic literacy and math skills required for their age44. In this scenario, an application that makes learning enjoyable has good prospects to our students really enjoy learning. (Headmaster of Amaf Primary School) “ make a difference. Smart Blackboard is an application that is not extremely high student to teacher ratios and yet operational but presents good prospects prohibitively expensive private tutors for the for impact. It allows students to access majority of families45. Smart Blackboard affordable tutoring while providing educators could make a difference in lowering students’ with additional income by means of micro- failure rates by facilitating a one-to-one tutorials. The application is a response to a support through the application. context where schools are overcrowded with 44 · http://www.bbc.com/future/story/20130205-t-is-for-tablet-computer. 46 45 · http://www.agra.org/where-we-work/kenya/. Aside from education, enhancing agricultural productivity is another government priority in Kenya. Agriculture accounts for 23 percent of GDP and 75 percent of the workforce46, with most food production in the hands of smallholder farmers. However, the lack of proper information is a key factor preventing farmers from making a livelihood out of agriculture47. A number of applications have focused on this gap. Some of them are already having an impact and others are on their way. MFarm provides access to up-to-date indicating a 166 percent increase in the agricultural commodity market information, selling price. Extrapolating such information links farmers to buyers through its is difficult; this could be an extreme case marketplace, and provides analyses of associated to one particular agricultural current agriculture trends. The application product. Nonetheless, it is reasonable to offers access to pricing information for assume that a significant number of farmers Kenyan smallholder farmers that usually can substantially increase their income. have to sell farm produce at low farm-gate Given MFarm’s network size, aggregate prices to middlemen, even when market farmer income increases are likely to be prices are higher. According to MFarm data, nonnegligible. the application has reached 6,300 users in its network. User testimonials show anecdotal evidence that a farmer can increase income from 30 shillings per kilogram of tomatoes to earning to 80 shillings with MFarm services, MFarm may double smallholder farmers’ incomes. The testing phase shows that the GreenHouse Pro can cut input costs by 11 percent. GreenHouse Pro has the objective of to cover the entire country. The application raising productivity in farming, especially works as a virtual agriculture extension in greenhouse farming. It is an app on officer, and the testing phase has revealed greenhouse technology providing information that the cost of inputs can be reduced by 11 on how to produce greenhouse crops. Fifteen percent when the information is provided to farmers in Kyambu are currently testing the the farmer with the app. application, and once launched it is intended 46 · In one revealing case, when the price of rice went up 20 percent in May 2011, it could have meant profits to farmers but the majority of the farmers experienced losses since they could not access reliable and timely information, according to the Economic Report on Africa 2012. 47 · In one revealing case, when the price of rice went up 20 percent in May 2011, it could have meant profits to farmers but the majority of the farmers experienced losses since they could not access reliable and timely information, according to the Economic Report on Africa 2012. 47 MkulimaBima helps farmers and insurance to take an average of five days (including one firms to track compensation claims in real- day of travel) to a process that now takes one time in an environment where smallholder day and does not require travel. This leads to farmers are particularly vulnerable to a decrease in harvest loss, since the farmer unpredictable rainfall and where crop losses can recover the funds to buy new seeds more are common, leading to food shortages and quickly, and also important travel costs even famine.48 Farmers that participated savings. Using conservative estimates and in a pilot launch are already signed up only taking into account farmers registered to the platform. Two thousand farmers during the pilot phase, the app could result are subscribed to the platform, and two in savings of more than Kenyan schillings insurance companies have been engaged. (KSh)15,000 of income a year, or about 7.5 The pilot started in July 2013 and concluded income days per farmer per year.49 in February 2014. The testing phase revealed that the process between crop loss and the claim will be faster: The app enabled farmers to move away from a claim process that used Farmers can save up to 7.5 income days a year (conservative estimate). directly to experts. The application was launched less than a year ago and data on specific impact are not available. Currently, the app is undergoing further testing, with 39 enrolled users located mostly in the small town of Syokiamu. The application’s FarmPal is another mobile agriculture prospects are promising, with features such application intended to increase income and as language options that enable users to productivity in farming. The app offers three choose between English, Kishwahili, Kikuyu, main features: a marketplace for farmers to and Kikamba, among other local languages. buy or sell second-hand farm equipment, a Reportedly, farmers who have interacted plant guide with understandable and relevant with the app were impressed and called it information for farmers, and an information “mobile extension officer.” desk where farmers can address questions Mobile money is seen as transformational for financial inclusion, which gains in importance given that exclusion from the formal financial system has been identified as a major barrier to a world without poverty.50. Large segments of the population in developing countries have traditionally been excluded from basic financial services, such as depositing, withdrawing, and transferring money easily, because they simply did not have a bank account. In Kenya, several applications supported by the mLab have focused on financial inclusion for both people and businesses. 48 · In Kenya, two-thirds of the population depends on the crops they grow and the animals they keep for their livelihoods and survival. 49 · For the estimate, it was assumed that the travel costs of a farmer coming from a village three to five hours from Nairobi would be between KSh600 to KSh1000 Kenyan Shillings (return ticket), in a context where smallholder farmers earn a daily income of KSh200 to KSh400. With mKulimaBima, farmers pay KSh10 to send the SMS and save the time and cost of travel. Assuming that one farmer processes one claim per year and that the process takes her/him three days (in which her/his opportunity cost is KSh1,200), and the travel cost is KSh800, she/he could save an average of KSh2,000, which is the equivalent to between 5 and 10 income days (7.5 on average). 48 50 · World Bank Group. 2012. Maximizing Mobile. Available at www.worldbank.org/ict/ic4d2012. Kopo Kopo is a platform that makes it easy, data is not available on these two variables. inexpensive, and convenient for a business However, the number of SMEs using the to accept mobile money payments. In app went up to 10,000 in 2013 from 50,000 partnership with payment providers, Kopo in 2012, indicating the app is likely to have Kopo aspires to create the “merchant substantial economic impact. operating system of choice” for businesses in emerging markets. It also helps clients move their business from cash to electronic payments, contributing to their integration into the Kenyan and East African markets. The app helps SMEs to expand their revenues and customer base. Quantitative The number of SMEs using the app increased to 10,000 in 2013 from only 50,000 in 2012. “ mPayer, by Zege Technologies, is a mobile and web payment gateway that enables businesses to manage mobile money, cash income, and expense transactions. The app [mPayer] has has been operational for over a year and approximately 100 companies and 3,000 helped us manage low-income clients (of which many are part of the BoP) are using it. The application our customers reduces time for payments to be reflected in a client’s enterprise system and motivates better. Now we can clients to “go green” via paperless and distinguish our cashless transactions. It has also proved to help small businesses track their income payments better better and for rural users, the app meant companies sell more and farmers get and make them farming inputs on time. Zege Technologies helps its clients adopt better accounting secure. practices by linking payments to customers and providing valuable statistical analysis We would not have and reliable reports. The app also helps customers obtain external financing: the digitalized information makes it more likely to have correct figures when presenting applications for external financing. There is no quantitative aggregated data on impact, grown as much as we have without this solution. “ (mPayer user) but anecdotal evidence and testimonials from companies using mPayer show it could be tangible. 49 Uhasibu is an Internet and cloud-based Uhasibu can reportedly enable businesses to accounting package developed specifically for cut 50 percent on accounting fees costs, which Kenyan SMEs looking for an affordable and could imply important aggregated cost savings. easy-to-use accounting solution. The company estimates that it has served between 300 and 400 enterprise customers since its creation in September 2011, and it has plans to reach up to 7,000 businesses over the next five years. Uhasibu can halve accounting costs. Given the company’s plans for expansion that implies important economic potential gains. Dairy Sacco is an app offering financial many challenges, in particular, difficulties in services to dairy farmers. It offers a mobile- accessing credit. Dairy Sacco improves access based credit tool that is modeled on to financial services tailored for agriculture agricultural cooperatives, providing input and purposes, and it is intended to increase merchandise to members on credit against income and trust for farmers by doing produce deliveries. The Kenyan dairy industry business in a more accurate way, for example, is a source of income and employment for by controlling inventory and balance of money. over 1.5 million smallholder farmers in rural Data on impact were not available to make areas. However, small-scale farmers face concrete estimates. According to the Vision 2030 government program, Kenyan health policy aims at the right to public and quality health care for all citizens. However, the country still faces a high shortage of physicians, with an average of 0.2 physicians52 per 1,000 people (as compared to 2.4 in the United States).53 Moreover, Kenya is still struggling with high incidences of fatal diseases: HIV is responsible for 29.3 percent of all deaths, tuberculosis for 14.4 percent, and malaria for 5.8 percent.53 An illustrative example of powerful mobile health of which an average of 42,000 were active users applications is MedAfrica, developed to improve who gained access to qualified doctors and the health of communities by increasing access health care self-management. The company to health care-related information. The app continued to expand in the following years, but also helps users to identify the nearest doctor, detailed usage and impact data were unavailable which is especially relevant for people in rural for this assessment. areas were the normal travel time to the nearest doctor can be one hour.54 The application also offers drug information and the app provides MedAfrica has enabled 42,000 users precise information on purchased medicines, to get access to qualified doctors and responding to the high share of pharmacies self-management of health care. that are not legally registered to sell drugs.55 In March 2012, MedAfrica recorded 70,000 users, 51 · Physicians include generalists and specialists medical practitioners (World Bank). 52 · http://data.worldbank.org/indicator/SH.MED.PHYS.ZS. 53 · http://www.who.int/countryfocus/cooperation_strategy/ccsbrief_ken_en.pdf. 54 · http://www.medafrica.org/. 50 55 · Reportedly, only 33 percent of pharmacies are legally registered to sell drugs, http://www.healthsystems2020.org/content/resource/detail/2760/. M-Chanjo seeks to reduce child mortality by As a response, M-Chanjo sends vaccination creating awareness on child immunization SMS reminders as well as basic information among parents by sending text message on health guidance. The app has reached 3,000 reminders. children since it was launched, increasing at an average of 1,000 children per year. In Kenya, children under the age of five have a mortality rate of 73 per 1000.56 One reason is that Two more mobile health applications developed children are not sufficiently protected against in Kenya complying with the assessment’s life-threatening diseases during the first years selection criteria were Idaktari and MyLife. of life. According to UNICEF, nearly 90 percent Idaktari is an application that helps doctors of un-immunized children in East and Southern and private clinics manage patient registration, Africa live in eight countries, including Kenya. including payments and reminders for Many of the cases are due to lack of information appointments. It is expected that the app will on immunization calendars or a correct follow- contribute to managing clinics more effectively up of the next vaccination due. and cut down on paper usage. M-Chanjo has reached 3,000 children and contributed to address a main cause of child mortality. Contributing to employment creation through service provided to 15 companies. The most mobile technologies appears to be an recent project is a solar lamp company seeking important agenda in Kenya, which has a youth to launch its product to all regions of Kenya unemployment rate of 35 percent. MobiAgent that have no electricity, especially in rural is a mobile application that delivers small Kenya.57 No resellers were subscribed for the tasks to the underemployed. It allows any project at the time of the assessment, as the individual to become a reseller of products product was still being advertised. Yet, if the and services offered by other companies on product is accepted, there could be substantial the platform. The application offers low-cost gains for resellers as well as cost savings for services for startups or companies who want the end consumers: resellers could earn half to push their products to the market. So far, 15 the minimum wage and households would resellers working for hosting domain services save 10 times the costs of kerosene in the first have earned 10 percent commission on each year alone.58 Impact estimates for MobiAgent are substantial: re-sellers could earn half the minimum wage and consumers reduce energy costs by tenfold in one year. 56 · http://applications.who.int/immunization_monitoring/globalsummary/countries?countrycriteria[country][]=KEN&commit=OK. 57 · 18.1 percent of population in Kenya has access to electricity (World Bank Data 2010). 58 · One lamp is sold at KSh600 at a 10 percent commission for the microworker, for a gain of KSh60. Each microworker can have 100 lamps to sell in one region, which means a gain of KSh6,000 (half of the minimum monthly income wage). The average cost for households spending on kerosene for lightning purposes is KSh500 per month. Alternatively, buying a solar lamp costs KSh600. Solar lamps are long-lasting items and therefore this entails substantial cost-savings. 51 Kenya faces a number of environmental challenges. Waste management59 and the loss of biodiversity are among them. Gorecycler and Fishmate are two applications addressing environmental protection in these two areas. Gorecycler is a web-based and mobile is not operational, as its features are still being application with the objective to enable users improved, for instance, through the inclusion of a to earn money while stimulating the mitigation routing system. Estimates point at the possibility of climate change effects. It aims to do so by of earning $0.5 per kilo of plastic and $2 per kilo providing a platform for people to find recycling of metal. The app does not feature a tracking resources and institutions near them on a map system and the profile of the users is mostly where they can sell recyclable waste. The app unknown, which makes it difficult to produce reported 200 active users during the first year aggregate estimates. of operation in 2012. It is currently online but it Fishmate connects fish farmers, fishers, up about 500 users over the next year. If the fish consumers, and industry players in one app receives traction, farmers could potentially marketplace. The application has a two-fold increase their income by 50 percent.60 But the goal: fair trade when buying and selling fish potential benefits of promoting fish-farming go and enhancing environmental awareness in beyond economic impact: the Kenya National a context where there is little aquaculture Environmental Management Authority’s State information, no online marketplace for fish of the Environment Report 2010 highlights trading, and a growing number of endangered the suitability and potential of fish farming to fish species. The app has operated for less than accelerate the delivery of the economic, social, a year and at present market research is being and political goals set forth in Vision 2030.61 conducted to collect data on users. To date, no purchases have been made through the app, but the developer of Fishmate expects to sign Fishmate could imply doubling income for fish farmers and three- fold cost reductions for consumers while protecting the environment. 59 · http://www.irinnews.org/report/97638/kenya-s-waste-management-challenge. 60 · Fishmate estimates that a fish farmer could sell a kilo of fish at an average of $3 using Fishmate via direct sales to end user, whereas they currently receive an average of $2 when selling to companies/factories. 61 · Page 72, Chapter 4 (http://www.nema.go.ke/index.php?option=com_content&view=article&id=249&Itemid=613) 52 5.3 mLabs’ Effects on Entrepreneurial Ecosystems This section answers the third research question: “What have been the effects of mLabs on the entrepreneurial ecosystem?” While this type of effect was not anticipated in the design of the mLab program under CSBKE, the assessment revealed that mLabs contribute to a range of effects on entrepreneurial ecosystems. Of these contributions, four stand out: Main outcome contributions of mLabs 1 mLabs strengthen the local talent pool and support early and very early stage innovation. 2 mLabs generate linkages between organizations in the ecosystem. 3 mLabs inspire and stimulate innovation in the ecosystem. 4 mLabs connect local, regional, and international ecosystems. The first three results have in common that by a brief description of the outcome and its they occur systematically and are substantial significance as well as an analysis explaining in each of the three analyzed ecosystems. The how mLabs contributed to these effects. fourth contribution stands out not because of The mentioned examples do not represent its magnitude but because it features effects for comprehensive depictions of all available which mLabs have a notable potential for future pieces of evidence; they aim at providing the contribution compared to other entrepreneurship reader with illustrative snapshots of each enablers. outcome contribution. The total evidence base for the statements is the analysis derived Each of the four subsections is organized as from 116 interviews with representatives of 82 follows: A summary of key findings is followed organizations in the three countries. 62 · The sample covered all stakeholder groups in each country ecosystem, that is, mLab incubatees, academia, mobile operators, mobile device manufacturers, investors, incubators and accelerators, government agencies and media. In most of the cases several organizations within each of these groups were interviewed (see annex B). 53 FIGURE 11: OVERVIEW OF MLABS’ CONTRIBUTIONS TO ENTREPRENEURIAL ECOSYSTEMS Strengthening Generating link- Inspiring and Connecting the talent pool & ages between stimulating local, regional, focusing on early organizations in innovation in the and international stage startups the ecosystem ecosystem ecosystems Broadening the Direct Among Connecting pool of startup stimulation of mLab-supported local teams and early mobile product entrepreneurs ecosystems stage startups innovations Among Creating Strengthening the Connecting mLab-supported partnerships, human resource regional entrepreneurs spaces, and base ecosystems and other actors support channels Generating Raising the Connecting multiplier effects Among profile of mobile international in the transfer of external actors innovation ecosystems knowledge 54 55 Strengthening the Talent Pool and Focusing on Early 5.3.1 Stage Innovators Key Findings • mLabs have supported local talent pools and early- and very early- stage innovators. This has increased possibilities for young developers and entrepreneurs to initiate startups and to grow them into successful companies. • Three main factors are at the heart of this contribution: the broadening of the pool of startup teams and early stage startups; the strengthening of the human resource base of the ecosystem; and multiplier effects in the transfer of knowledge. One of the most tangible outcome contributions of mLabs in all three countries is that they have helped talent development and early stage innovators that otherwise would not be served by the ecosystem. They have done so by providing entry channels and capacity to young developers and entrepreneurs, thereby increasing the likelihood of startup creation in the mid- to long-term: more people entering the ecosystem, with more capacity and with incentives to create and develop startups. All in all, this means that the possibilities for young developers and entrepreneurs to initiate startups and to grow them into successful companies have been enhanced. There are three factors that are at the heart of the contribution of mLabs to their enhancement of local talent pools and support of early stage innovators: 1) mLabs have broadened the pool of startup teams and early stage startups. 2) mLabs have strengthened the human resource base of entrepreneurial ecosystems through training, workshops, and event-based skill- building. 3) mLabs have generated multiplier effects in the transfer of knowledge from incubatees and other supported entrepreneurs to other early- stage innovators. 56 1. Broadening the Pool of Startup Teams and Early Stage Startups mLabs have addressed early stage innovators focus on set startups, either with initial market and entrepreneurs that often start off with not traction or with clear evidence of high payoff much more than an idea and a rough business potential. model for a mobile application. For example, in Armenia and South Africa, mLabs focused In other words, mLabs have widened the their activities mainly on young developers and pipeline for startup market entries by including entrepreneurs. In Kenya, while the mLab is also players that had been excluded previously, supporting more advanced startups, training thereby broadening the pool of early stage programs for individuals have mainly targeted startups and startup teams. For example, mLab young developers, with a recent change of Southern Africa created new channels to bring emphasis toward business graduates and non- the previously isolated programmer student technical entrepreneurs. Aside from training community into the entrepreneurial ecosystem. programs, mLabs have also experimented with While it is not possible to say with certainty how events and competitions to create more concrete many new startups are created in this way that channels toward founding a startup (see box 4). otherwise would not have been created, it is This is different from the typical approach of likely that mLabs’ emphasis leads to an increase profit-oriented accelerators and incubators that in the overall number of potential and actual are usually much more selective and tend to startups over the long-run. BOX 4: CREATING INCENTIVES AND CHANNELS FOR STARTUP CREATION AND DEVELOPMENT THROUGH EVENTS AND COMPETITIONS Events and competitions by mLabs have been a channel for developers to get interested in entrepreneurship and move into startups. The immediate objective for the majority of events and competitions is not startup creation, but rather to spur mobile product innovation and engage participants in competitive exposure. Yet, even if startup creation was more of a long-term rather than an immediate goal, it occurred in several cases. The five teams generating revenues within mLab ECA at the time of this assessment were all formed at hackathons and competitions. Such events have not only contributed to the creation of startups but also to their evolution. For example, the idea of the Gardener Application (Nako Games) was generated in a hackathon before a revenue model was developed in the mLab. The startup Nako Games was then incorporated and incubated at the mLab and is today on the path toward becoming a promising game application development company. In Kenya, a large number of young innovators are already seeing technology, including mobile and web based applications, as a way to make a living. 88mph, a well-known accelerator in Nairobi, received around 400 applications during their last program round in 2012, a number that would have been unthinkable a few years ago. As a reason for this change, mLab East Africa’s Pivot East competition is widely considered to be a major driving force, complementing universities’ efforts and the buzz prompted by the iHub. Data from a recent evaluation conducted by the University of Nairobi revealed that a number of Pivot East winners continued and grew their startup with the help of the mLab’s incubation program, and also 80 percent of interviewed past participants who did not join the incubation program were still in business (as of June 2013), with the great majority of them (90 percent) feeling that their business was growing. For mLab Southern Africa, there is anecdotal evidence that competitions have led to the creation of startups. An example is the case of MFactory. The founder joined the mLab based on his participation in a hackathon and has now created a successful service company. The same is true for GoMetro and Aftarobot, which joined the mLab having been selected as winning teams of the Gauteng Innovation Challenge 2012 (see section 5.2 for more information). 57 57 2. Strengthening the Human Resource Base in the Ecosystem mLabs have particularly focused on providing skills, such as workshops, mentorship, working spaces, and meaning they have strengthened the human resource testing facilities, assisting developers to evolve into base in the ecosystem, making it larger, more inclusive, startups and helping early stage startups to further and more apt for startup creation. The human resource develop and consolidate. base has mainly been strengthened by means of As shown in table 6, mLabs have had considerable training programs and by engaging developers reach through their training and coaching programs in hackathons and competitions featuring skill- for developers and entrepreneurs. In most cases, development elements, such as hands-on coaching. individuals reached and trained were young developers Moreover, core incubation services have offered assets and entrepreneurs. TABLE 6: DATA ON TRAINING AND COACHING DELIVERY Teams applying for in-depth support through mLab 78 216 460 Total: 754 Teams that received in- depth one-on-one support 21 57 119 Total: 197 Number of people trained 223 260 462 Total: 945 Number of developers and entrepreneurs reached 1,578 1,000 5,000 Total: 7,578 Source: mLab scorecards, values as of June 2013. The expectation is that at some point, some of the skills and knowledge levels to overwhelmingly good trainees will set up businesses. This important link and very good levels through the training. Moreover, between broadening and strengthening the human subsequent to leaving the mLab, most trainees have resource base, which later translates into startup maintained close contacts with their classmates and creation, can already be seen in Kenya. An evaluation 68 percent of the interviewed graduates continue to undertaken by the University of Nairobi revealed that attend mLab events, with the Wireless Wednesday 28 percent of the past participants to mLab East series as the most commonly cited. Africa’s training program have created a startup and are involved in running it—more than half of them on Moreover, mLab ECA offered two different types of a full-time basis. Together with qualitative evidence capacity building: a four-month internship program that mLab trainees value the training highly in its where practical and theoretical training were contribution to their skills, confidence, and market combined as well as specific demand-oriented exposure, there is indication that the mLab contributes training sessions. Trainees considered the internship to creating incentives for startups to develop even program unique in its offering of practical experience through services focused on skill building. and job perspectives. There is a widespread perception that the mLab will not only contribute Moreover, according to the same evaluation, for to startup creation but also to the development of all areas covered in the mLab East Africa’s training highly qualified professionals ready to fit into other program (technical, business skills, and others), past newly started companies searching for young and participants reported to have moved from average innovative talent. 58 3. Generating Multiplier Effects in the Transfer of Knowledge The mLab pilot experiences have revealed that mLabs are also able to generate multiplicative knowledge flows where expertise acquired by startups and entrepreneurs during mLab support programs feeds back into the startup ecosystem at large. FIGURE 12: AN EXAMPLE OF MULTIPLIER EFFECTS IN THE CREATION AND TRANSFER OF KNOWLEDGE 88mph iBiz Growth Africa Growth Africa (accelerator) (incubator) (incubator) (incubator) Peer mentoring Peer mentoring Graduated Pratical knowledge Successful startups transfer through startups peer-mentoring Networking Pivot Coaching Pivot winners Mentoring East One-month accelerator An example such effects can be found in Kenya (see East Africa are part of this multiplier effect: during figure 12). The Pivot East competition identifies the competition, participants gain knowledge from candidate teams that have potential to become the Pivot East one-month acceleration program, successful startups. Some of the winners join mLab which some of them transferred as peer mentors in East Africa’s incubation program and benefit from other incubation centers, such as the C4D Lab at the networking, coaching, and mentoring. Eventually, University of Nairobi. Therefore, entrepreneurs that these incubatees graduate and often become received support from mLab East Africa—in one successful startups (such as Kopo Kopo and way or another—are transferring that knowledge mFarm). Importantly, some of the founders and back to the ecosystem through peer mentoring. employees of these startups then become mentors in other incubators and accelerators (iBiz, Growth The cycle of practical knowledge sharing described Africa, 88mph), thereby allowing the knowledge and above has an important characteristic feature: practical experience generated within the mLab to the networks and linkage that the mLab creates feed back into the startup ecosystem. Even Pivot East for client startups become the channels for the participants who did not win the competition and knowledge generated to be transferred inside the thus did not enter the incubation program at mLab ecosystem and produce multiplier effects. In other 59 words, the high degree of exposure to ecosystem in place in South Africa and in Armenia. Thus, it actors during mLab incubation programs facilitates appears that it is a matter of time for these effects the linkages that subsequently connect mLab to reveal themselves for these two mLabs as well. graduate startups with other incubation and acceleration organizations in the ecosystem, That said, knowledge transfer through peer enabling knowledge transfer to multiple parties. mentoring and learning inside mLabs is already happening for mLab Southern Africa and mLab The effects of significant knowledge transfer by ECA. For instance, in Armenia, interns have mLab support alumni outside of the realm of often become trainers and team leaders that mLabs have not yet happened in Armenia and then mentored or trained new incoming interns. South Africa, simply because no incubatee of these In this regard, mLab ECA has institutionalized two mLabs has graduated yet. The emphasis on peer mentoring and learning within the mLab— knowledge creation and the generation of linkages although this has not happened in the context between startups and other ecosystem actors (that of startup creation, as was the case for mLab is, the precondition for the effects in Kenya) is also East Africa. 60 60 Generating Linkages between Organizations 5.3.2 in the Ecosystem Key Findings • mLabs have made a difference in creating linkages that increased the density of the ecosystem in contexts where the lack of connections to professional networks is a major hindering factor for startup development and resilience. • Linkages created have been created at three levels: among mLab-supported entrepreneurs, among mLab-supported entrepreneurs and other players in the ecosystem, and among different actors of the ecosystem without involvement of client entrepreneurs. A limiting factor for startup entrepreneurs in general63 mLabs in the three countries have all made a and in Africa in particular is isolation and lack of difference in this regard by generating linkages connections to professional networks, or “relational where isolation was a barrier for entrepreneurs assets,” that make them prosper and more resilient. to flourish and develop. mLabs have generated Most interviewees in the counterfactual group were linkages at four levels: active startups that had not gone through incubation. (1) mLabs brought startups together, that is, they They often pointed out that the connection of the linked incubatee startups with one another within startup with other startups, partners, and funding the mLabs. sources (usually family and friends) is a key factor for (2) mLabs linked startups to actors in the ecosystem continued entrepreneurship. A product or business beyond the mLab. model may be tweaked or redesigned relatively (3) mLabs generated linkages between ecosystem quickly, but a valuable network is not easily created. actors outside of the mLab. 1. Linkages among mLab-Supported Entrepreneurs Bringing startups together in mLabs has proven to generate a series of intangible effects valued by creates a collaborative environment and strong ties startups across mLabs: among co-incubatees. This effect is more tangible • Office and common working spaces as well as in mLab ECA and mLab Southern Africa. access to basic infrastructure (such as Internet or meeting rooms) under the umbrella of a • This collaborative environment at times goes recognized institution (the mLab) have made a beyond peer support and exchanges of ideas difference. Beyond the direct monetary value, it and translates into specific initiatives and even also had a motivational effect for most interviewed business deals. An illustrative example comes from entrepreneurs, as it meant that they would take mLab Southern Africa: GoMetro and other startups their startup seriously and work on it as a full-time inside the mLab subcontract fellow mLab startups business. for mobile application development. Similarly, Jatamobile, another startup inside the mLab, • Common workspaces and the fact that most provides graphic design services to other incubatee incubatees come in through the same pipeline peers, subsidized by the mLab. This creates an (hackathons, competitions or challenges) also internal market place within the mLab. 63 · http://www.businessinsider.com/loneliness-and-startups-2013-7. 61 61 2. Linkages among mLab-Supported Entrepreneurs and Other Actors A second level of linkages brought about by mLabs is the opinion of the mLab manager, “in the past Nokia would linkage of entrepreneurs and startups to the ecosystem have most probably brought companies from Finland to beyond the mLab. Most interviewed startups and conduct training; now it is relying on local resources.” entrepreneurs expressed that the most difficult part is to “enter the loop,” that is, to be effectively connected with mLabs have also had an important role in linking startups the ecosystem. This is a role that mLabs have played very with contract opportunities. An example is GoMetro in well. Through connections to the ecosystem, startups South Africa that entered into a contractual agreement benefit from exposure to new actors and linkages with with the Gauteng province. This contract was directly other organizations, for instance, through directed brokered by the mLab and the Innovation Hub, which contract brokering or participation in fairs and events. provided the necessary credibility and backing to the startup. For instance, at mLab East Africa, meet-ups prompted by the mLab between education startups (both within The link between startups accessing external funds and and outside the mLab) have led a group of them the mLabs is, at times, a clear direct link and at other (Eneza Education, Shoolbox) to work on establishing times the result of a more indirect effect prompted by an association with the objective of doing advocacy, increased exposure to networks. Examples of indirect leveraging the interest of the sector, and stirring dialogue links usually entail that networking facilitated by the mLab with the Ministry of Education. increases the number of possibilities for new exchanges and interactions that may help startups to attract funds. mLab ECA also had an important role in integrating actors from the broader mobile app ecosystem. Eurasia For example, at mLab East Africa, Eneza Education’s Foundation and Kolba Labs are two important social exposure to the mLab is one of the enabling factors that innovators that want to integrate mobile technologies to allowed the company to obtain resources from Savannah develop solutions. They are not technology specialists and Fund. Similarly, Kopo Kopo sees a direct relationship did not have any linkages with mobile developers before between the increased exposure generated thanks to the engaging with the mLab. Both organizations contacted the iHub/mLab tandem and attraction of investment funds. mLab to ask for technical support in order to transform existing ideas into app products. Microforester was developed during the m2work competition organized by mLab ECA. Microforester then In South Africa the mLab is connecting startups with decided to continue development of its prototype within established industry actors. For instance, the startup the mLab’s facilities and with the mLab’s help. The app MFactory is working as a direct training service provider was awarded a grant by CRDF Global to support the for Nokia. The contract has been established by the mLab, continuation of the project. The application was also which provides the necessary credibility and legal liability presented in the local government´s STEP competition to MFactory to be able to work for such a large firm. In the where it won $15,000. 3. Linkages among External Actors mLabs have also created events specifically The third level of linkage building occurs when designed to link actors so as to broaden the the mLab (or any of its direct clients) generate ecosystem and enhance its quality. Wireless linkages among ecosystem actors external to the Wednesday in East Africa is a clear example. Wireless mLab. Wednesdays is an initiative that brings together the tech community, industry practitioners, and mLabs bridge and link external actors by several ecosystem actors (such as government agencies or means: The most common are hackathons, investors) to discuss problems in the sector. By means innovation challenges, and competitions, which of such integration, the events intend to contribute bring together developers, entrepreneurs, mobile to generating channels, improve communication operators, mobile device manufacturers, and between industry and tech players, and ultimately potential investors. bring about pertinent mobile solutions. 62 5.3.3 Inspiring and Stimulating Innovation in the Ecosystem Key Findings • mLabs have actively stimulated innovation across the three countries and they have done so in three ways: by directly stimulating product innovations; by creating partnerships, spaces, and support channels for innovation together with universities and innovation hubs; and by raising the profile of mobile innovation in the ecosystem. • The increase in the profile of innovation in the ecosystem and the inspiration effects generate snowball effects of subsequent innovation processes. In many cases, these ecosystem stimulation effects arise from community-building activities. A main outcome contribution of mLabs across the three countries is that they have actively stimulated innovation. They have mainly done so in three ways: by directly stimulating product innovations, by linking innovation stakeholders such as universities and innovation hubs, and by raising the profile of mobile innovation in the ecosystem. 1. Direct Stimulation of Mobile Product Innovations With regard to product innovations that mLab-supported entrepreneurs have developed, data on application prototypes and applications brought to market indicate that mLabs have indeed had an effect in promoting innovation in the mobile space: innovation has happened within and around the mLabs. Intermediate values Final values (either 2011 or 2012)65 (Numbers for 2013 are TABLE 7: NUMBER OF NEW MOBILE APPLICATIONS until June/July 2013) BROUGHT TO MARKET64 Source: mLab reporting scorecards. Number of new mobile applications brought to market 14 64 110 117 115 195 Number of mobile application prototypes 22 74 117 131 60 301 64 · Apps brought to market (438) and prototypes created (442) reported in mLab scorecards differ from the population of applications from which applications with development impact were selected (292, see section 5.2.). The former sample of applications includes a wider range of applications created at events and competitions (such as hackathons) as well as trainings and workshops; the latter includes only applications that mLab managers kept record of by name. 65 · Depending on the availability of the data. 63 Similar to startup creation in terms of absolute mLab. Yet, mLab East Africa has indeed played a numbers, the extent of mobile product innovation role in product innovation, only at later phases of varied depending on how mLabs positioned the innovation cycle: it supported more advanced themselves in the process of translating ideas teams and focused either on bringing innovations to into customers’ products. With regard to product the market or in supporting the design of business innovation coming from mLab incubatees, the models ensuring product innovations are better direct impact on generating ideas has been more monetized. Moreover, direct product innovations discernible for mLab ECA and mLab Southern have also happened outside the incubation program; Africa, given that the generation of ideas has mainly for instance, mLab East Africa gave participants happened inside the mLabs. In contrast, for mLab an incentive to develop applications in the training East Africa, incubatees’ innovative ideas had in most program and in the Pivot East competition. cases already happened before clients entered the 2. Creating Partnerships, Spaces, and Support Channels for Innovation mLabs have actively created partnerships and young entrepreneurs without access to professional support channels for innovation in the ecosystem. The labor markets but willing to develop ideas into processes that they have promoted include, among businesses. Universities are foundational knowledge others, research on new applications, the creation of providers and laboratories transform this knowledge spaces that stimulate the generation of new ideas, into marketable products. In this regard, the mLab is and the organization of challenges and competitions. located right at the start of innovation processes. Often, what is enabling mLabs to be stimulators and “ prompters of innovation are their institutional or physical proximity to innovation hubs or universities, which can be the initial nodes for innovation processes within the ecosystem. For instance, mLab East Africa is located in the same building as the iHub, one of the four members of the mLab consortium and a reference organization for the tech community. This There are several close physical and institutional proximity allows the mLab to develop innovation-related joint initiatives successful with iHub Research and the User Experience Lab (UX businesses in there Lab), both linked to the iHub. (in the mLab): While mLab Southern Africa’s location in Pretoria is far from ideal (complicated access for young developers mFarm, Kopo without access to their own transport), the closeness to the Innovation Hub, a government-funded science Kopo. I think they and technology park that brings together high-tech industry, academia, and entrepreneurs, is a plus all go through factor when it comes to generating innovation in collaboration with the provincial government (see example of GoMetro). mLab ECA is located within the State Engineering University of Armenia campus, sharing the building a mentorship program. (Developer and hackathon winner in Kenya) “ with the Microsoft Innovation Center, other engineering laboratories, and several tech companies. This set- up allows the mLab to be close to its target group: “ Community building initiatives have been a core activity for mLabs in promoting partnerships and support channels for innovation within the ecosystem. mLabs organized hackathons, innovation challenges, and competitions and also linked them with the After the Microsoft incubation processes, which made it possible for winners to develop their innovation beyond the initial Innovation Center and mLab idea. Overall, community building activities turned out to be a versatile instrument with several positive were created in Armenia, effects (see box 5). if one follows the events An example comes from mLab East Africa. In 2011, Pivot 25 was the first regional startup competition in East that they organized, such Africa. It was launched at a very incipient stage of the ecosystem and has had an important effect in prompting as hackathons, mLab software developers to innovate. For instance, Pivot East competitions, you can see (which became the brand for the competition from 2012 “ “ that the number of people that create social innovation is increasing hugely. For me the mLab (Program Development Manager of an NGO in Armenia) is a place, where I can learn from my peers. I want to learn from as many onwards) triggered the founder of Ma3route, an mLab people as possible “ startup in Kenya, to delve into the software solution that became the well-known mobile/web/SMS platform and the mLab helping citizens to share and access information about transport and traffic conditions. Ma3route is not provides this space. the only example; there are several others such as Eneza Education, MedAfrica, or Zegetech, who started (An entrepreneur at innovating around Pivot East and continued the process mLab Southern Africa) at the incubation program in the mLab. BOX 5: COMMUNITY BUILDING: A KEY EXPLANATORY FACTOR BEHIND SEVERAL EFFECTS Community building, including competitions and events, has been referred to several times throughout the report. The reason is that hackathons, competitions, and innovation challenges are a main factor contributing to more than one outcome. Community building has been mentioned as: 1) A mechanism that creates incentives and channels for startup creation and development 2) A means to strengthen the ecosystem’s human resource base 3) As a core element enabling mLabs to promote innovation within the ecosystem, and in particular, as a factor that increases the profile of innovation in the ecosystem and generates inspirational effects that lead to further innovation. 65 65 3. Raising the Profile of Mobile Innovation mLabs have also promoted innovation indirectly by and Aftarobot in South Africa. These success means of two wider positive side effects of innovation stories have then inspired many other students and challenges, competitions, hackathons, and other young developers and motivated them to embark outreach. First, mLabs have been able to broadly on product innovation. The importance of locally raise the profile of innovation in the ecosystem, relevant role model startups was widely cited. making the case for innovation in the media and putting mobile and software related innovation on It is clear that mLabs have increased the number of the host countries’ agendas. Second, events and people working on innovations. What is less clear competitions had important inspirational effects,66 is to what extent these innovations have been of as they produce winners that eventually become high quality. Whether the effects of competitions successful entrepreneurs. Examples range from are viable market-based innovations or just pitches Mysales and Nako Games in Armenia to Kopo to win a prize without follow-up is a controversial Kopo and Eneza Education in Kenya or GoMetro aspect.67 66· Although inspiration is difficult to measure, in-depth interviews showed that the extent of this effect is present. 67· Distractions from real market/product/ business model development innovation processes are at times pointed out as adverse side effects of 66 competitions. See http://blogs.worldbank.org/psd/evolution-startup-competitions-case-pivot-east and http://www.infodev.org/m2Work. 5.3.4 Connecting Local, Regional, and International Ecosystems Key Findings • Given their transnational scope, mLabs have a high potential to connect local, regional, and international ecosystems. However, this potential has not yet been fully realized. • infoDev’s global activities have shown promising ways to increase the recognition of mLabs through exposure to international ecosystems as well as to increase international recognition of local startups. mLabs have a high potential to connect local, The two main factors explaining why regional regional, and international ecosystems. Their activities have been relatively limited are core mandate is regional, but they have a local insufficient funds available and mLabs’ need and an international dimension. This outcome can to focus on the development of a viable value be expected to increase given mLab’s regional proposition for clients in the cities and countries mandate, initial palpable evidence from mLab that mLabs are located in. This is in line with the East Africa as the longest running mLab, and pilot and innovative nature of the mLab concept mLabs’ role as part of an international network that first needed to be established locally before (the four mLabs worldwide). Yet, at the time of this it would be able to start working regionally and assessment, there was no widespread evidence of internationally. this effect occurring across the board yet. Yet, a number of promising regional activities The format of Pivot East, mLab East Africa’s illustrate the potential for future linkages between annual regional startup competition, has so far different ecosystems. Aside from mLab East Africa’s shown the best results for regional connection approach based on Pivot East, mLab Southern of ecosystems. The third edition of Pivot East in Africa has also reached out into the ecosystems 2013 was held in Kampala, which increased the of neighboring countries. It provided coaching in regional nature of the event and implied practical the first hackathon ever held in Mozambique and knowledge sharing between the mLab and Outbox generated a series of linkages. Young developers Hub in Kampala. from Mozambique have visited the mLab in South Africa out of their own initiative and interviews in The Pivot East Concept was also adopted and Mozambique showed an enormous interest in the emulated in Nepal and in Vietnam as an example mLab’s activities. of international coordination, but this was not grounded in a structured knowledge exchange mLab ECA’s mandate covers over 10 countries between mLabs. There are potential opportunities in the region: Ukraine, Moldova, Belarus, for experience sharing between Vietnam and Kazakhstan, Uzbekistan, Tajikistan, Kyrgyzstan, Kenya, particularly with regards to the challenges Turkmenistan, Azerbaijan, Georgia, and Armenia. in agriculture addressed by mFarm. However these mLab ECA organized a Regional App Mobile potentialities and the ensuing mechanisms for Contest in Armenia in 2013 where, for example, sharing the experience have not yet been explored. an innovator from Uzbekistan presented the 67 prototype of Be Healthy application that Africa, the Global Forum in East London enabled ultimately received funding of $70,000 from the mLab Sothern Africa to establish relationships Ukrainian incubator, Happy Farm, which had with a number of subnational regions within South been contacted by the mLab for the contest. Africa, especially the provincial governments Likewise, mLab ECA is working with enablers in of the Eastern and Western Cape, which would two countries: it opened an office in the Ukraine probably not have been aware of the mLab and is also expanding to Moldova. without this event. Moreoever, the possibility for local startups to participate in the VentureOut International linkages created by infoDev through Challenge provided them with the possibility its global activities as well as the Global Forum for of promoting their products in international Innovation and Technology Entrepreneurship show markets. GoMetro was one of the finalists of more potential of linking different ecosystems. the VentureOut Challenge and had a chance to Instead of facilitating a continuous exchange and participate in a global infoDev event in Moldova. collaboration between local enablers, infoDev Such activities have been perceived as an added supported mLabs through global activities, which value of infoDev as they laid the groundwork include the Global Forum, as well as innovation to enable the mLab to devise an international competitions such as the m2Work project, and strategy and to create awareness for the mLab the mAgri and VentureOut challenges. In South within the local ecosystem. 68 6. SUMMARY AND CONCLUSIONS This chapter summarizes the main conclusions and messages of this Outcome Assessment, providing applications, as outcomes that related to the high-level answers to the research questions posed by original objectives of the mLab program. The infoDev. For reference, the assessment examined (1) assessment also provided (3) an analysis of mLabs’ mLabs’ effects on startup creation and development effects on entrepreneurial ecosystems, since (as the best available proxy measures of increased evidence pointed towards potentially substantial competitiveness in mobile application enterprises) and effects that were not originally envisioned under the (2) the development impact generated through the use of CSBKE program. Broader goals of CSBKE, such as applications developed by mLab-supported developers improvements in the knowledge economy or reaching and entrepreneurs, including the number of relevant disadvantaged populations with mobile innovations, were covered implicitly under these three effect types. 6.1 mLabs’ Effects on Startup Creation and Development The assessment shows that mLabs’ interventions revenues over time across the board, while are positively associated with variables, such creating high quality jobs and raising investments as startup revenue generation and job creation and seed funding. through the creation and growth of startups. At the same time, it is worth noting that only Indeed, after less than three years of operation, a rigorous impact evaluation could quantify economic results generated by mLab-supported the economic value that mLabs add. However, startups are tangible, even if the available data mLabs are interrelated with entrepreneurial and the assessment methodology did not allow ecosystems in complex ways, so the design of for precise specification of the net quantitative a reliable impact evaluation study would be a effect that mLabs have had on them. Over difficult task. Still, infoDev and mLabs need to time, mLabs contributed to the creation of an continuously monitor quantitative results data, increasing number of new startups through their at the minimum allowing for intertemporal interventions, and the startups supported in their observations and the setting of benchmarks for incubation programs have managed to increase comparative studies. 69 69 Development Impact through the Usage 6.2 of Mobile Applications About 10 percent of the 292 mobile applications In conclusion, there is evidence that the effects of identified as developed by innovators supported by mLabs also include contributions to development mLab ECA, South Africa, and East Africa had actual impact through the use of mobile applications or potential development impact, defined as a supported by means of incubation, training contribution to the Millennium Development Goals programs, and competitions. As a caveat, the (MDG), addressing users at the Base of the Pyramid analysis was limited by the unavailability of data (BoP), or otherwise having a transformative and the diversity of applications and development potential in improving conditions for people and stages. Measuring the contribution of applications businesses. to development impact merits a separate impact assessment in the future, once a larger number Most applications with development impact of applications are launched and they have been analyzed are from Kenya. The main reason is likely used in the market for long enough to demonstrate that the total number of applications developed by usage results beyond early traction data. entrepreneurs that received support from mLab East Africa is much larger than for mLab ECA and mLab Southern Africa. This appears to be the result of the expansive strategy adopted by mLab East Africa, promoting startup creation and growth largely via training programs and competitions (in addition to the incubation program), which have many participants and lead to application development as immediate outcomes. Among the 29 applications analyzed, there were a few significant success stories, several companies generated satisfactory effects, and a majority of applications were still in pilot stages. The latter were found to not yet have succeeded in the market but to feature some potential for development impact. The fact that applications have not yet been launched in some cases, and the short time span elapsed since the applications were developed and launched in other cases, has led to limited depth of development impact observed to date. In Kenya, direct development impact could be observed for some mLab-supported mobile applications that, for instance, improved users’ access to fundamental services, such as health, education, and financial services. In South Africa, three applications are generating positive impact on mobility and employment opportunities, and another two are raising awareness on crucial problems, such as gender violence and child abuse. In Armenia, applications are at a more incipient stage and only potential impact could be considered. 70 6.3 Effects of mLabs on Entrepreneurial Ecosystems The influence and effects of mLabs on stakeholders, opened channels and pipelines for entrepreneurial ecosystems varied across startups to emerge that did not exist previously, countries given that each mLab has operated and stimulated new innovation processes in the in a very specific and different ecosystem (see ecosystem. All these contributions have meant ecosystem snapshots in chapter 3). Yet, in all adding actors, linkages, and features to the cases, mLabs made palpable contributions to entrepreneurial ecosystem. the talent pool and early stage innovators, to the generation of linkages in the ecosystem, and to This conclusion is of relevance given that, in the stimulation of innovation. the first stages of implementation, “lack of insight into the local innovation ecosystem or It is difficult to discern any kind of net contribution a limited understanding of what an ecosystem to an entrepreneurial ecosystem, as one could support project should usually entail”69 made it argue that any actor in an ecosystem has a hard for some mLab managers and consortium positive effect on it merely because they are part organizations to identify what activities should of it. However, the three mLabs stand out as be prioritized. In this regard, it appeared that deliberate contributors with a strong positive net mLabs have been able to adapt timely. effect. In other words, they have been ecosystem Moreover, the pilot experiences have shown builders as a result of mLabs’ business models68 that the regional and cross-national nature and value propositions that focus on an expansion of mLabs leads to a high potential to connect of entrepreneurial ecosystems: mLabs have local, regional, and international ecosystems. created new spaces for exchanges, brought in However, this potential has yet to be realized in new actors, generated new linkages between its full extent. 6.4 Overall Assessment In sum, it was assessed that mLabs did make a have a high potential to establish themselves difference. Importantly, the analysis was able to as ecosystem builders that fill the void among identify unexpected and indirect contributions, universities, mobile technology companies, and including for client and beneficiary groups that are startup accelerators. In this vein, the evidence only indirectly connected with the mLabs (such as collected for this Assessment suggests that the mobile application users or innovators learning return on investment for mLabs should not be from incubation graduates). While this assessment measured only in terms of their direct economic does not present precise quantifications, overall, effects represented in startup creation and the evidence suggests that the direct net economic access, but also account for the positive effects effect of mLabs through the creation and on entrepreneurial ecosystems. The mLab grants development of mobile application startups is only turned out to be seed investments with substantial one part, and potentially not even the largest part, multiplier effects for mobile application innovation of mLabs’ contribution. in the mid- and long-run. The model of mLabs as ecosystem builders presents good prospects for The highlighted conclusions point to two important replication in countries and regions with incipient aspects to be considered for a decision whether to and latent yet promising mobile application and invest further in the mLab concept. First, mLabs entrepreneurial ecosystems. 68· Refer to the Business Model Evaluation for more details on the notion of an “ecosystem builder,” available at http://www.infodev.org/ mobilebusinessmodels. 69· Quoted from the Business Model Evaluation, available at http://www.infodev.org/mobilebusinessmodels. 71 Second, mLabs have shown ability to foster the generation of mobile applications that in themselves have development impact; however, this ability has not yet materialized to its full extent as the lion’s share of relevant applications are still in the pilot and testing stages. Moreover, through such activities as innovation competitions, hackathons, and training programs, mLabs support the generation of applications only at initial development stages (except for applications developed in the context of longer-running incubation programs). Only once these applications surpass testing and are launched to a wider marketplace can they generate more substantial impact. Yet, neither mLabs nor infoDev are providing substantial support for most applications in that critical transitional phase, which is ultimately a precondition to further increase the commercialization rate of applications that have the potential for significant development impact— one of the original objectives not yet achieved in full. Targeted commercialization support to socially and environmentally beneficial applications would offer infoDev the possibility of reaching further beyond its original goals. 72 73 ANNEXES ANNEX A: CHARTS ON MLAB-SUPPORTED STARTUPS’ CREATION AND DEVELOPMENT 1.1 Economic effects: Results for mLab-supported entrepreneurs in East Africa, Kenya 627,000 50 141,727 6 12 Startups created Revenues generated by startups ($) 1,500,000 1,000,000 100 56 50 External Direct jobs investments created by raised by startups startups ($) KENYA (up to June, startups created Women up to September) 2011 2012 2013 74 1.2 Economic effects: ARMENIA Results for mLab-supported entrepreneurs in ECA, Armenia 5 55,100 Startups created Revenues generated by startups ($) 128,700 37 32,700 25 15,000 15 18 External investments raised by Direct jobs created by startups startups ($) 2011 2012 2013 (Up to July) Women 1.3 Economic effects: Results for mLab-supported entrepreneurs in Southern Africa, South Africa SOUTH AFRICA 345,110 652,205 625,293 12 51 283,000 39 8 10 4 Startups created Revenues generated External Direct jobs created by startups ($) investments raised by startups by startups ($) 2011 2012 2013 (Up to September) Women 75 1.4 Mobile applications: KENYA Results for mLab-supported entrepreneurs in East Africa, Kenya 263,000 78 38 Consumer app users Pro-poor / BOP / rural population reached / social development apps 301 196 115 70 60 New mobile apps brought to market Number of app prototypes 2011 2012 2013 (Up to July) 1.5 Mobile applications: Results for mLab-supported entrepreneurs in ECA, Armenia 1,700,000 6 6 ARMENIA Consumer app users Pro-poor / BOP / rural population reached / social development apps 131 117 117 110 12 New mobile apps brought to market Number of app prototypes 2011 2012 2013 (Up to July) 76 1.6 Mobile applications: Results for mLab-supported entrepreneurs in Southern Africa, South Africa 598,000 8 6 2 Consumer app users reached Pro-poor / BOP / rural population / social development apps 64 2011 74 2012 22 10 14 2013 (Up to September) New mobile apps brought to Number of app prototypes market 1.7 Capacity and skill building: Results for mLab East Africa, Kenya KENYA 460 119 338 7 Number of teams applying for Teams that received in-depth, 2011 in-depth support through mLab one-on-one support 2012 462 5,000 2013 (Up to June) Number of people trained Number of developers and entrepreneurs reached 77 1.8 Capacity and skill building: ARMENIA Results for mLab ECA, Armenia 57 216 15 8 25 Number of teams applying for Teams that received in-depth, in-depth support through mLab one-on-one support 2011 260 1,000 2012 180 2013 24 (Up to July) Number of people trained Number of developers and entrepreneurs reached 1.9 Capacity and skill building: Results for mLab Southern Africa, South Africa SOUTH AFRICA 78 21 12 32 2 Number of teams applying for 2011 Teams that received in-depth, in-depth support through mLab one-on-one support 2012 223 1,578 1,122 2013 108 (Up to September) Number of people trained Number of developers and 78 entrepreneurs reached ANNEX B: LIST OF INTERVIEW PARTICIPANTS ARMENIA List of organizations interviewed: Armenia TYPE ORGANIZATION / INITIATIVE TARGET GROUPS (10) Mobile Enabler mLab ECA Entrepreneur Nako Games Entrepreneur MySales Entrepreneur Microforester Entrepreneur My News Entrepreneur Letsein Team Entrepreneur Users mLab Entrepreneur Clean H2O, Paint the Music Entrepreneur Stories Hub COUNTERFACTUAL (6) Company Unicomp (Buy.am) Company Idram Mobile Wallet app Company Appolosoft Company Ineed Company PICS’ART Company Altacode COMPARISON (4) Company X-Tech Team Company Instigate Mobile CJSC Start Up 4 car Incubator GITC ECOSYSTEM (5) Business association UITE Private sector Orange Foundation Eurasia Other Google Development Group International Organization UNDP 79 KENYA List of organizations interviewed: Kenya TYPE ORGANIZATION / INITIATIVE TARGET GROUPS (18) Mobile Enabler mLab East Africa Mobile Enabler Akirachix Mobile Enabler eMobilis Entrepreneur 1 Kopo Kopo Entrepreneur 2 Kopo Kopo Entrepreneur 3 Kopo Kopo Entrepreneur 4 Kopo Kopo Entrepreneur Zegetech Entrepreneur 1 Eneza Education Entrepreneur 2 Eneza Education Business Development MedAfrica Technologies Ltd. Entrepreneur Sleep Out Entrepreneur 1 Ma3route Entrepreneur 2 Ma3route Entrepreneur Hackathon winner Entrepreneur 1 Akirachix Entrepreneur 2 Akirachix Entrepreneur 3 Akirachix COUNTERFACTUAL (4) Entrepreneur JuaKatibaYako Entrepreneur Hackathon winner Entrepreneur 1 Changa Entrepreneur 2 Changa COMPARISON (4) Incubator/Startup support 1 Growth Africa incubatee Incubator/Startup support 2 Growth Africa incubatee Incubator/Startup support 3 Growth Africa incubatee Incubator/Startup support 88mph/Startup Garage Incubator/Startup support C4D Lab Incubator/Startup support 1 Growth Africa Incubator/Startup support 2 Growth Africa Incubator/Startup support Nailab Africa Incubator/Startup support iLab Africa Incubator/Startup support Savannah Fund 80 TYPE ORGANIZATION / INITIATIVE ECOSYSTEM (14) Ecosystem Ushahidi, iHub Private sector Nokia Private sector Intel Business Development Viktoria Solutions University University of Nairobi Private sector Samsung Research Consultant/Entrepeneur iHub, Sprint Interactive (Research consultant) Girl Afrika (Cofounder) Business Development Afrinnovator Journalist Journalist in social innovation,technology, entrepreneurship, business Public sector Kenya ICT Board Private sector Microsoft Private sector Wayra Incubator/Startup support iHub Research Private sector GSMA Mobile for Development FINAL USERS (9) User Founder, Pete’s Coffee User Founder, Kasap Butcherry Shop User 1 Milimany Primary School User 2 Milimany Primary School User Nurse User Volunteer at Afrikarsi Nursery School, Student at Akirachix User Volunteer at Health Center, Student at Akirachix User ICT Student at Akirachix training program User Eneza User 81 SOUTH List of organizations interviewed: South Africa AFRICA TYPE ORGANIZATION / INITIATIVE TARGET GROUPS (13) Mobile enabler The Innovation Hub Mobile enabler 1 mLab Southern Africa Mobile enabler 2 mLab Southern Africa Entrepreneur Geekulcha Entrepreneur Mfactory Entrepreneur 1 Sowertech Entrepreneur 2 Sowertech Entrepreneur Appchemy Entrepreneur Med-TechEngineers Entrepreneur Jatamobile Entrepreneur BookBay Ecosystem Angels Fair Entrepreneur GoMetro COUNTERFACTUAL (3) Entrepreneur (counterfactual) Expenza Entrepreneur (counterfactual) Tech4Africa Entrepreneur (counterfactual) Praekelt Foundation COMPARISON (5) Incubator/Startup support 1 Blackberry Applications Lab Incubator/Startup support Jozihub Incubator/Startup support 2 Blackberry Applications Lab Incubator/Startup support RIIS Incubator/Startup support 88mph ECOSYSTEM (9) Private Sector Nokia Private Sector Microsoft Government Department of Science and Technology Private Sector Microsoft Nongovernmental Institution SAINE Private Sector Ungana Africa Private Sector Vodacom Government/Founding Partner CSIR Private Sector Abaphumeleli Business Consultants, EmpowerWeb, South African Black Entrepreneurs Forum (SABEF) 82 TYPE ORGANIZATION / INITIATIVE FINAL USERS (5) User 1 Geekulcha User User 2 Geekulcha User User 3 Geekulcha User User 4 Geekulcha User User GoMetro User 83 ANNEX C: RESEARCH PROJECT TIMELINE Field mission to Armenia, Start of Backgorund Questionnaires Kenya, Post-field the project research & Interview South Africa Research Gudielines Briefing Desk Review No. of interviews No. of interviews with Task (Internet, Design conducted conducted Managers available templates (total): 111 (total): 5 and ToR documents) clarification Identify & In Armenia: 24 For Armenia: 1 Inception contact Report interviewees In Kenya: 53 For Kenya: 2 Agendas In South Africa: 34 For South Africa: 2 September 27th September September - October 11th September 20th - 27th October 16th - 27th (2 weeks, 1 13th - 19th 2013 2013 (2 weeks) week each (1 week) (1 week) country) 84 Internal Peer Analysis Reporting Review Review Final Report February 11th - March 6th 2014 November October 28th - 2013 October November 1st February 20th - 27th to 1st - 10th 2013 January (1 week) (1week) 2014 2014 85 ©2014 infoDev / The World Bank | 1818 H Street, NW | Washington DC, 20433 Email: info@infoDev.org | Tel + 1 202 458 8831 | Twitter: @infoDev 88 www.infodev.org