Value, growth, impact: how the lean startup approach can transform social enterprise
The lean startup framework has been used to transform the field of entrepreneurship, and is now being adopted by large enterprises as well. A number of social innovators and funds are also advocating and implementing the lean startup model for the social enterprise sector.
Ann Mei Chang provides a comprehensive framework and extensive case studies for social innovation in her book, Lean Impact: How to Innovate for Radically Greater Social Good. The 14 chapters are spread across 290 pages, and make for a useful and inspiring read.
Ann Mei Chang has worked in the startup sector, for giant tech firms, the government, and at non-profits. She was earlier Chief Innovation Officer at USAID, Executive Director of the US Global Development Lab, and Chief Innovation Officer at Mercy Corps. She spent more than 20 years at Google, Apple, and Intuit, and serves on the boards of BRAC USA, and IREX. She has a BS degree in Computer Science from Stanford University, and is currently the Executive Director of Lean Impact at Lean Startup Company.
Here are my key takeaways from the book, summarised in Table 1 (below) as well. See also my reviews of the related books The Lean Startup Way, The Lean Startup, Lean Startups for Social Change, A World of Three Zeroes, Do Good, Scaling Up, and The Prosperity Paradox.
Innovation applies not just to technology but business, politics, education, culture and social change, as explained in the foreword by Eric Ries, author of the bestseller Lean Startup. The lean approach is not just possible for social entrepreneurship, but preferable as well.
Part I: The Lean Startup framework
Successful Silicon Valley companies not just think big, but have perfected the art of testing, learning and iterating faster. The lean startup approach can reinvent our approach to social good as well, Ann begins. “Think big. Start small. Relentlessly seek impact,” she urges.
The world needs not just altruism and philanthropy but systematic ways to maximise impact by accelerating learning, managing risk, and forming synergistic partnerships. “Social innovation involves iterative testing and improvement, refining business models, influencing partners and policy, fine-tuning logistics, and many other practicalities,” Ann explains.
“The aim of Lean Impact is to find the most efficient path to deliver the greatest social benefit at the largest possible scale,” she defines. “By combining scientific rigour with entrepreneurial agility, we can dramatically increase both the depth and breadth of our impact,” she adds. This calls for humility, flexibility, and grit.
For example, d.Light began as a low-cost LED light whose batteries could be powered by diesel generators. But Founder Sam Goldman pivoted to solar lanterns when he saw diesel generators were not as readily available in India; the company now has sold 20 million solar light and power products in 62 countries.
Understanding community needs for water supplies and designing along with farmers has helped Proximity Designs become the largest social enterprise in Myanmar. Proximity to the customers also helps build trust, gather creative insights, promote participatory research, and even enable serendipity.
Health in Harmony dug deeper into deforestation dynamics in Indonesian Borneo to discover that trees were being cut to raise funds for healthcare costs. The solution was to develop alternate livelihoods and local health clinics. The inquiry approach was based on the Five Whys technique of Toyota.
“Great ideas often blend a direct need with a creative vision and technical expertise,” Ann explains. Insights must be brought in from end-user communities. For example, protective suits initially used during the Ebola crisis were designed for air-conditioned buildings, and not humid climates; new designs were submitted via a Grand Challenge.
It is important for social innovators to replicate or build on existing solutions, rather than hoard insights and be bound by considerations of ego and politics, Ann advises. Innovators need to creatively find ways to get users to communicate exactly how they feel about the product, factoring in cultures where people may not want to be openly critical, eg. giving them multiple choices of features. Funders also need to go beyond endless pilots, and work on partnerships to scale.
The author illustrates the lean startup principles of experimentation and fast learning in action at South Africa’s Harambee Youth Employment Accelerator. The steps are: identify assumptions, build an MVP (test riskiest assumptions quickly and cheaply), use validated learning (MVP performance data), build-measure-learn across iterations, and pivot or persevere. Harambee focused not just on getting youth jobs, but finding out how to increase retention, improve training based on job conditions, change job locations, and identify appropriate price points.
Ann cautions against relying only on vanity metrics (eg. number of units sold, number of people reached, etc), and advocates focusing instead on innovation metrics (eg. willingness to pay, unit economics, profit per sale, referral rates, etc). Such data should not be confused either with the reams of compliance data required by funders.
For example, Revolution English switched from classroom-based teaching of English to an approach based on real-world content, online interaction, and consistent feedback. It was able to show that the solution both works and scales well (value hypothesis, growth hypothesis).
TOMS Shoes initially donated a pair of shoes to the needy, for each pair bought by a consumer. It later pivoted to local manufacture of shoes, and giving other products like clean water or eyeglasses.
Part II: Validate
The three pillars of social innovation are value (is there demand?), growth (will it scale?), and impact (does it work?). It is important for social innovators to move quickly from vision to validation, Ann urges.
Faster validation at early stages speeds up innovation. Field data are more important than boardroom debates about what solutions work; evidence is better than conviction. The order of assumption tests should be based on risk, time, and cost. “Failure is a natural and essential part of the process,” Ann cautions.
For example, One Acre Fund assumed that smallholder farmers in Africa could benefit from growing passionfruit, but there were challenges with training, quality, and transportation, and the experiment failed. It now carries out smaller tests in nursery settings, followed by a few dozen farmers’ fields, before scaling up.
Tenofovir gel, intended to prevent transmission of HIV, received a lot of funding but failed to take off in the market due to issues of impracticality in use. Understanding issues of user preference earlier could have helped redesign or even abandon the gel.
Digital tools now enable better data collection, and can also be used within organisational networks. “Good, timely data gives us an unbiased view into experiments,” Ann explains. With digital networks, data can move at “light speed versus bicycle speed,” according to Chuck Slaughter, Founder of Living Goods. This can increase the pace and accuracy of learning.
Nexleaf Analytics uses IoT to monitor cold-chain storage of vaccines. Harambee uses text messaging to stay in touch with job recruits for two years, and glean insights about its employment services.
Designing with users and in context helps come up with relevant and usable solutions. Reel Gardening’s Garden in a Box improves garden yields by encasing seeds along with fertilizer in strips of paper, with specific instructions for spacing and watering. Eco-Fuel Africa helps farmers start their own micro-businesses in creating fuel briquettes from agricultural waste by using press machines.
The author describes a range of MVPs in use for validation of value: marketing (TaroWorks mobile CRM), concierge (MicroMentor’s introductory and semi-automated calls), Wizard of Oz (Code for America’s GetCalFresh website; Copia Global’s paper catalog for m-commerce), hardware (Simprints fingerprint scanner for farmers), and improv (mobile video education modules for Indian patients’ families, by Design for Extreme Affordability researchers; home-delivered intervention by Roca to reduce incarceration cycles in Boston).
The Fund for Shared Insight has created an initiative called Listen for Good to help nonprofits develop a five-question survey based on NPS. This helps benchmark performance in the social service sector.
There are limits to relying on charity for growth; more sustainable engines are market-driven (Off Grid Electric’s financing options via private debt and equity), voluntary contributions (crowdsourcing, eg. DonorsChoose for classroom supplies), cross-subsidy (Aravind Eye Hospital, BioLite stoves), replication (Alcoholics Anonymous), franchising (BRAC), open source (PATH water filters), government funding (Zipline drones for medicine delivery in Rwanda), and government support (India’s National Deworming Day with school participation).
Technology-based solutions can also scale well, eg. Khan Academy (12 million learners a month, on a $37 million annual budget). “Small is beautiful, but big is necessary,” in the words of Fazle Hasan Abed, founder of BRAC. The Abdul Latif Jameel Poverty Action Lab (J-PAL) curates a library of practical resources for researchers, and has a network of over 160 affiliated professors from universities around the world.
Validating impact of social innovation is a challenge due to complex multi-dimensional linkages and long gestation periods, but the author describes a number of useful approaches. It is important to classify results into outputs, outcomes, and long-term impacts. Efficiency and efficacy are important considerations to assess.
Recommended approaches are rapid-cycle evaluations (IDinsight and mosquito net distribution in Zambia), early indicators (regular student performance tracking by Summit Public Schools), working with constraints (One Acre Fund and tree growing in African farms), and multi-dimension tracking (student food habits and school menus, by FoodCorps).
Instead of selling pre-mixed packets for oral hydration therapy to tackle diarrheal mortality in Bangladesh, BRAC taught mothers how to make the mixture themselves. In addition to giving rural women loans to buy cows, they started a dairy processor, Aarong Dairy, with chilling centres to collect milk.
Part III: Transformation
The last part of the book takes the discussion beyond individual social enterprises to re-architecting the entire ecosystem of funding, policy, alliances, and hybrid structures. Diverse interests and agendas lead to funding that is often siloed, splintered and even misdirected, Ann laments.
Duke University’s Centre for the Advancement of Social Entrepreneurship (CASE) has developed a suite of online tools called Smart Impact Capital to help entrepreneurs navigate the funding landscape.
Many social innovators receive funding for a fixed timeline, with no sustainable path forward after the project is completed. Fortunately, some non-profits like Medic Mobile have found committed donors who offer unrestricted funding for building complete platforms.
mPower Social Enterprises negotiated the right to use funds for prototype development which included the freedom to fail. Many tech philanthropists give support for innovative ventures with room for experimentation and adaptation. Open Impacts’ report, The Giving Code provides some such insights.
The EMCF foundation provides flexible capital for innovations helping economically disadvantaged youth. Advanced market commitments (AMCs) involve contractual guarantees for product purchases, eg. vaccines. Social Impact Bonds (SIBs) and development impact bonds (DIBs) are other innovative mechanisms for social innovation.
Blended finance options are used by the Omidyar Network. Kenya’s M-pesa was initially funded by the UK’s DFID as a grant to Vodafone to develop a system to repay microfinance loans by SMS.
A major cultural shift is needed in the social sector to deal with failure, according to the author. It leads to risk aversion and lack of transparency. “Knowing when to let go or pivot, particularly in the face of strong personal attachment, is one of the hardest decisions for leaders to accept,” Ann explains.
MoveOn has a “joyful funeral” to honour risk-taking, effort and creativity even in the face of failure. There are public forums like Fail Festival and also internal forums on failure at organisations like USAID.
The final chapters call for going beyond CSR to regular social contributions across the board. Impact investment funds (‘do good’) are the need of the hour, not just responsible investment funds (‘do no harm’). New forms of blended and hybrid finance are needed, but those that social entrepreneurs can tap into without a degree of financial engineering. Examples include low-profit limited liability companies (L3Cs) and community interest companies (CICs).
In sum, it is important to bridge profit, purpose, and social good, the author urges. “What matters is setting our sights high, learning as fast as we can, and finding every possible way to maximise value, growth, and impact,” Ann signs off.