But there is another number worth looking at: Out of nearly 95.2 million enterprises registered in the country, over 99% are micro.

As of September 9, 2026, the Ministry of MSME Dashboard reported 95,236,393 enterprises across the Udyam Registration Portal and Udyam Assist Platform, which aim to give MSMEs a formal identity and registration number, along with access to lending and other benefits. Of these, 94,651,631 were classified as microenterprises — about 99.4% of the total. 

However, the word “micro” needs some care here. Under the revised MSME classification effective from April 1, 2025, a microenterprise can have investment of up to ₹25 million (~US $265,000) and annual turnover of up to ₹100 million (~$1.06 million). So being classified as micro does not, by itself, tell us that a business is tiny. What the numbers do tell us is that the overwhelming majority of registered enterprises (about 73 million) are still at (or technically “under”) the first rung of the MSME (micro, small and medium) classification. They are classified as “nano,” a subset of “micro” that includes businesses with less than ₹10 million in annual turnover. These enterprises are mainly cash-based and family-run, and the majority of them are rural. They have the capacity and aspiration to grow, but no dedicated ecosystem to help them transition and scale.

India has clearly become very good at bringing enterprises into the formal system. What we have not become equally good at is helping viable enterprises move beyond micro and into their next stage of growth: becoming large enough to attract formal investment.

The Gap Between Nano and Investable

At the end-point of that journey sits the Self-Reliant India Fund. The fund was set up to channel growth equity to MSMEs through a Fund-of-Funds structure. According to government data from July 30, 2026, 766 MSMEs had been assisted between the fund’s launch in 2021 and June 30, 2026, with more than ₹30 billion invested ($316 million) by the government. An additional $900 million of government funding is yet to be deployed.

Put these two numbers next to each other and the gap becomes hard to miss: Tens of millions of enterprises have entered the formal system. But just hundreds have grown large enough to obtain equity funding from a government platform designed to provide them with easy access to this support.

What happens to everyone in between?

That question has stayed with us through much of our work with rural and semi-urban businesses. At Quiver Catalyst, a rural enterprise scaling platform, we work with growth- oriented entrepreneurs who are ready to move beyond day-to-day survival. Our support combines practical mentorship, business support and advisory, investment readiness, and growth capital in the form of micro-equity — because we have found that capital alone rarely addresses the barriers that emerge as a business grows. 

We conducted a year of fieldwork with more than 850 enterprises across four states of India, combining the findings with national data to produce our latest white paper, “Beyond Survival: Unlocking Growth Pathways for India’s MSMEs,” which examines the transition from microenterprise survival to sustained growth. In the process, we’ve seen that many Indian enterprises are not struggling to survive. They have customers. They sell something people are willing to pay for. Some have been around for years. The entrepreneur has already done the difficult work of getting the business off the ground.

And yet, when the time comes to grow beyond the nano stage, things begin to get complicated. The books that were perfectly adequate for running a nano business may not be enough for an investor trying to understand its margins, working capital or cash flow. The entrepreneur may know exactly how to sell within her district but have no idea whom to call when she wants to expand into the next one. A large order can look like an opportunity until the business realises it does not have enough working capital to fulfil it.

Then there is the question of the money itself: The business may have outgrown the microloan that helped it get started, but that doesn’t mean it is unbankable. It may just be too large for the smallest loans and still too small, informal or unstructured for many institutional investors. For instance, it may need $5,000 or $25,000 to make the next jump, not $50,000 and not $1 million.

This is where a lot of businesses get stuck.

Other Obstacles Faced by Growing MSMEs

These MSMEs also face an obstacle less visible than finance. In a small enterprise, the founder is often doing almost everything: speaking to customers in the morning, buying raw material in the afternoon, checking production, following up on payments and somehow maintaining the accounts in between.

As the business grows, the decisions this entrepreneur must consider start to change:

• Should I hire someone now or wait?
• Can I afford more equipment?
• If I sell through a distributor, what margin do I give up?
• Can I take this loan if my sales drop for three months?
• How much stock should I hold before the main selling season begins?

In a larger company, there are people whose entire jobs are built around answering questions like these. For a first-generation entrepreneur, they can become decisions that determine whether the business grows at all. And where the entrepreneur is located changes how difficult it is to find those answers.

Ambition is not a Mumbai privilege — Access is

For instance, we have repeatedly seen that MSMEs located outside of major population centres face unique challenges that aren’t shared by their urban counterparts. This is not because entrepreneurs in smaller towns lack ambition or capability. It is because the networks that support business growth are far less dense and often much harder to access outside India’s major commercial centres. These enterprises are hampered by the three Ls: location, missing market linkages and lack of leverage.

Consider what an entrepreneur’s journey can look like in a major city like Mumbai (India’s financial capital) compared to a small, remote Indian town. At different points in this journey, a growing business needs a banker who understands it, an accountant who can help put the numbers together, a distributor willing to take a chance, a mentor who has solved the same problem before, and a network that can provide connections to investors and buyers.

In places like Mumbai, many of these people are relatively close. In a small town like Gadchiroli, the entrepreneur may have to travel much further — both socially and physically — to cover each of these needs.

Supporting MSME Growth in India Requires More than Just Capital

The question then is: What would make that distance shorter?

Expanding access to incubators and accelerators is clearly one part of the answer. Jagriti Enterprise Centre – Purvanchal, PRIME Meghalaya, Villgro and a growing number of incubators use different models, but all point to a simple need: Entrepreneurs outside the large business centres need access to mentors, markets, knowledge and networks, not only classroom training.

But incubation cannot end with a workshop and a certificate. An entrepreneur trying to grow a business needs someone to turn to when the distributor does not pay, when the margins change, when new equipment suddenly looks necessary, or when the loan that looked affordable on paper begins to feel very different in an off-season.

Finance needs some rethinking too. A rural agricultural processor can have a perfectly healthy business over the course of a year, and still have very uneven cash flows within those 12 months. But a fixed installment payment does not care whether it is peak season or lean season: The monthly repayment amount remains the same.

Government programmes have done a great deal to expand formal credit and credit guarantees for small businesses. But access to credit and access to the right kind of capital are not always the same thing. For some businesses, revenue-linked or cash-flow-linked structures may make more sense. For others, debt may still be the right answer. The larger point is that the financing has to understand the business rather than forcing every business into the same repayment pattern.

And even the right capital will do very little if the enterprise has nowhere new to sell.

Market access sounds like a broad development-sector phrase until you sit with an entrepreneur and realise it can mean something as simple as finding the first reliable distributor outside her district.

Government channels have opened useful doors for businesses navigating the challenges of growth. The Public Procurement Policy for Micro and Small Enterprises sets a 25% annual target for the amount of goods procured from MSEs across Central Ministries, Departments and CPSEs, while the MSME Ministry’s Procurement and Marketing Support Scheme is intended to help MSMEs improve marketability and reach new markets. But getting through those doors still requires documentation, quality control, working capital, competitive pricing, fulfilment capability and relationships.

That is why we have slowly come to believe that the transition from nano to investable cannot be solved by capital alone.

Sometimes the business needs money. Sometimes it needs six months of better record-keeping before taking on that money. Sometimes it needs a distributor, someone to sit with the entrepreneur and work out whether buying that new equipment will actually improve the business. And sometimes the most useful thing for an entrepreneur is simply helping her understand how an investor or lender is likely to perceive her business before she sits across the table from one. We have therefore started one step earlier than the question, “How much capital does this business need?” We ask: “What is actually preventing this business from growing? What needs to change before more capital can genuinely help?”

The answers to these questions are different for every enterprise. That is precisely the point. Every MSME faces its own challenges in moving from nano to small and medium-sized. India needs to do a better job of addressing those obstacles if it hopes to see its successes in formalising MSMEs translate into actual, sustainable business growth and job creation.

Quiver has embarked upon a journey to develop a new capital stack led by risk capital, along with an enabling system to help transition these enterprises. But India needs tens (or hundreds) of like-minded organisations to join us, if we hope to scale the country’s millions of nano enterprises. 

Read our full white paper at https://www.quiver.in/insights

Pranshu Chhabra is a development professional working with Quiver, Rudra Midhun Kumar is part of Quiver’s Diligence Team, and Chiropriya Dasgupta is the Director of Investments at Drishtee Foundation and Co-Founder of Quiver.

Photo credit: Quiver

However, scaling these innovations could also provide NGOs with a solution to that funding challenge. The development sector has reached a firm consensus that it needs to depend less on grants, so hybrid business and operational models are becoming increasingly important for NGOs seeking to remain sustainable, relevant and adaptable as the aid landscape evolves. There’s a growing need for these organisations to explore approaches like rethinking their funding architecture, claiming a clearer niche, and evolving the support systems they can draw on. Strategies to generate earned revenue could play an important role in that ongoing evolution toward diversified funding.

But this transition often costs money before it makes money. And this “messy middle” — the period where early support falls away before a business model can stand on its own — is where social innovations stall. This is the stage the Enabel Innovation Hub focuses on. Enabel is the Belgian Agency for International Cooperation, and its Innovation Hub supports mission-driven organisations in Africa and the Middle East in scaling innovative solutions to complex societal issues. The Hub’s research looked at what has to change from within for NGOs to execute this transition — an organisational and mindset shift the sector has been far quicker to call for than to fund. And what we learned is that the transition between grant dependency and financial sustainability goes beyond finding a different type of payer. Instead, it may require an organisation to change what it provides and to whom, how it is governed, who it employs, and what it chooses to say “no” to.

There is no clean path from grants to revenue 

Since 2018, the Enabel Innovation Hub has provided “transition-to-scale” support to over 50 non-profit social innovations across 18 African countries through grant funding and capacity-building technical assistance. 

For a long time we worked on the assumption that scale — funded by a larger grant with more ambitious targets — was the next natural step after a successful pilot. But our grantees’ lived experience showed us that doing more of the same thing in more places is rarely what the job requires. The harder task is changing what an organisation is and where its money comes from.

Those changes were easier to contemplate when grant budgets were growing. Now they are shrinking while expectations rise, and NGOs are being told to diversify as though there were a clear route from grant money to a sustainable model, in markets where programme beneficiaries cannot afford the full cost of the service or product being provided.

To better understand how NGOs are successfully diversifying, we undertook a study involving 18 mission-driven organisations operating across Africa, with varying degrees of financial diversification and at different stages of their journey. Our goal was to better understand how NGOs can scale their innovations while pursuing financial sustainability.

We found that as organisations scaled, financial sustainability was often a condition they continuously managed, with diversification often being the first visible shift. Of all the organisations we spoke to, almost none had moved neatly from grants to earned revenue. They layered and rebalanced income streams as circumstances changed, whether due to funding volatility, policy shifts, affordability pressures or operational demands. Shocks such as losing a grant often prompted diversification and model evolution more than strategy alone.

Financial diversification is not the same as scaling, but it brings many of the tensions involved in scaling to the surface. Deciding to sell something forces an organisation to figure out what it is uniquely good at, who will pay for it, and whether it is prepared to stop some of its regular programming to make space for this new focus (or to hand over its innovations to external partners that will take them to market).

Diversification did not relieve short-term financial pressure for the NGOs we spoke with, because it introduced complexity and demanded capabilities the organisation did not yet have. However, we did find that those NGOs that were more financially diverse were often the most resilient.

The Anchoring-Balancing-Compounding Archetypes

To make sense of the patterns we were seeing, we developed the Anchoring–Balancing–Compounding (ABC) archetype model. It describes three archetypes that non-profits occupy and move between as they diversify their funding sources. And crucially, it identifies the different demands each one places on the organisation and its backers — e.g., showing funders the different types of support required.

Anchoring organisations are predominantly grant-led, bringing in a small share of earned income that helps subsidise costs while grants remain their main source of revenue. Their most pressing need is room to experiment: unrestricted or flexible capital to test revenue models, support to work out pricing for users who cannot pay the full cost, and permission to fail without a restricted grant’s deliverables pulling them back toward premature scaling.

Balancing organisations run grants and earned revenue in rough parity and are actively holding mission and commercial logic together, often through dual legal structures. Here the constraints shift from experimentation to management, creating a need for other types of support — e.g.: governance and legal support for a hybrid structure, working capital to run several income streams at once, and assistance with the harder, less fundable task of managing the tension between mission and revenue as both grow.

Compounding organisations are led by earned income, retaining grants for what markets cannot or should not fund, and they often present as social enterprises with a strong core mission. Their needs are different from those of the other two archetypes, and they include: strategic, catalytic grants ring-fenced for any non-commercial work, the partnership and government-relations capacity to embed into public programmes or procurement systems at scale, and the evidence infrastructure that performance-linked or outcome-based finance demands.

These archetypes raise two key questions for non-profit organisations: Which financial model best serves the organisation’s mission, and what would it need to change to operate with that model?

The ABC Archetype Model, developed from Enabel’s Pathway to Sustainable Scale research and refined through Enabel’s implementation with the Fit4Scale portfolio.

The capacity limbo that no one pays for

An organisation moving into Anchoring, Balancing or Compounding brings assets a new commercial entrant would need years to build. Most have spent a long time working in contexts where markets are thin and public budgets are constrained. They’ve also built relationships a newcomer simply doesn’t have. They know how to design with users rather than for them, and how to translate between what is happening on the ground and what a donor’s reporting template will accept.

However, none of that transfers automatically to a commercial model. Knowing a community well will not tell you how to run three income streams at once, price a service that has never had a price before, or judge when another organisation is better placed to carry an innovation forward. Those are different jobs, and they usually need different people.

This challenge can be seen in the insights we learned from Aflatoun, a global education organisation that empowers children and young people through social and financial education. It set up AflaVentures as a profit-making arm to generate earned income, and then could not staff it with the parent organisation’s existing team. According to a representative at the organisation: “AflaVentures has mixed success, partly because as an NGO … you really need to have commercial people to be successful in that, and finding a person who is interested in doing commercial activities for a nonprofit isn’t easy … at the moment we are not doing fantastic in that area but it’s purely a staffing issue more than anything else.”

Hiring is one part of the transition. A commercial function also has to sit inside an organisation whose systems, incentives and culture were built for non-profit delivery, and that can require a second legal entity with its own board.

In situations where the current team cannot stretch far enough, the grant work and the commercial work compete for the same people. For instance, Green Farmlands has a core team small enough that staff double up across roles. Yet it cannot resource everything at once, so it chooses between its non-profit activities and its revenue-generating ones month by month.

The organisations we interviewed described this as capacity limbo: more demand and opportunity than the team and its systems can carry. This situation is most challenging when an organisation is moving from founder-led delivery to formal operations, or running an innovation alongside core programmes with no dedicated capacity for either. At myAgro, for instance, roles shifted every few months and process knowledge sat informally in people’s heads, which made it difficult to bring in new staff.

This is why the form of support offered to these organisations matters as much as the amount. NGOs at this point need funding that’s flexible enough to let them build capability and change course when a model fails. They also need backers who behave as learning partners rather than compliance monitors checking whether fixed project deliverables have been achieved.

The hidden challenges behind financial diversification

To help address these interrelated challenges, Enabel launched its Fit4Scale initiative earlier this year, which aims to build the capabilities NGOs actually need as they pursue financial diversification and scale, working directly with the leaders who have the decision-making power to make these calls.

We’re supporting organisations’ efforts to: test new commercial business lines which bring in earned revenue, establish commercial enterprises, set up governance mechanisms that can subsidise their non-profit activities, and explore partnerships and alternative funding models to scale their innovations.

But these structural model shifts are only part of the decision. Underneath lie harder challenges.

One involves a change of mindset. Much of the NGO world is built around a particular idea of social mission: doing good for its own sake, delivering projects, meeting donor commitments, etc. But earning revenue requires a different mindset that enables the marketing and selling of products and services — one that holds a commercial logic alongside the social one. This not only changes what an organisation does, it also changes the kinds of people it needs to do it. As Mark Thomson, Director of Business Strategy and Delivery at IWMI, a global research organisation addressing water challenges in developing countries put it: “The hardest shift for IWMI hasn’t been strategic … it is cultural. We had to stop thinking of our digital tools as project outputs and start thinking of them as products that need to be maintained, improved and sustained over time to deliver ongoing impact. That reframe sounds simple, but it cuts against decades of how research institutes are funded and how our people are incentivised. Grant cycles reward novelty. Sustaining what works well requires a completely different mindset.”

The other challenge is often more emotional: balancing mission alignment. Financial diversification forces NGOs to confront questions around mission drift — e.g., as you bring in earned revenue and new capabilities to scale your innovation, does this protect your mission or shift you further away from your original purpose? As Ian Pringle, Executive Director of Farm Radio International described it: “Our expertise in participatory communication, audience engagement, radio, digital platforms and data, for example, can be viewed not only as tools for delivering development programmes, but as capabilities that can solve problems for other organisations. The difficult part is knowing what to commercialise without commercialising our purpose. For Farm Radio, the test is not simply whether something can generate revenue. It is whether we can build a viable offering around what we do exceptionally well, while continuing to reach the people and communities who are unlikely to be served by a purely commercial model.”

Scaling can also mean partnering with others to bring in the skillsets the NGO doesn’t have, sometimes handing them a piece of what the NGO built. And that’s often where the discomfort sits. Handing an innovation to someone else to scale can feel like losing it, even when another organisation is better placed to carry it forward. As Martin Jacobs, Global Rehabilitation Specialist at Handicap International put it: “For the first time, we are looking at our eHealth application through a scaling lens. Instead of simply asking how to do more with more inputs, such as replicating a pilot elsewhere, we are examining our unit cost of delivery and our long-term financial model. This puts our mission as an NGO into question: Where does our mandate end? When and how should we let go and hand the solution to others better equipped to commercialise it? And how do we ensure those partners prioritise impact over profit?”

Commercialisation can pull financial resilience and organisational values into direct conflict. Earned revenue can be less reliable than a long-term grant, particularly in thin markets with volatile demand. And diversification only leads to scale when there is a market to diversify into, which can rule out a substantial share of the work the development sector exists to do. Additionally, though deciding to spin off a commercial activity is perfectly valid (and often represents the most sustainable route to scale), this is not an outcome that is typically rewarded by grant funders.

So the financial support for commercialisation efforts has to reach further than revenue. NGO leaders need the space and the support to decide how their impact should scale, and to shift both skillsets and mindsets across the organisation in line with that pathway.

What funders would have to change

The overarching question is: What types of support are funders prepared to provide, to help organisations implement these changes? We see three core areas where this support could focus:

Funding the transition itself: As mentioned above, moving to new financial models generates costs before it creates returns. These organisations need transition capital: temporary funding that enables them to learn what type of organisation they need to become, and then funds their efforts to become that organisation.

None of this requires the sector to spend more than it already does. Most grant budgets allow a small amount for capacity building, and it is usually spent on compliance training to meet the donor’s reporting standards. The work that moves an organisation from one funding model to the next is different: It involves everything from pricing and financial modelling, to the slow business of learning to sell something. It produces no immediate beneficiary numbers, and it’s difficult to attribute to a funder’s intervention, which is why it is often overlooked.

Designing for learning, not only delivery: How money is designed matters just as much as the amount. Field realities move faster than reporting chains, so a leader who discovers in month eight that the business model does not work has two bad options: Acknowledge that the attempt has failed, and risk being seen as a poor planner. Or keep struggling to deliver the original deliverables across the original grant period and lose two more years.

Funding arrangements shape whether organisations have the space to make these choices well. As one leader at myAgro, which helps smallholder farmers save for agricultural inputs, put it: “If you come up with a list of requirements and deliverables, we’re just going to steer towards those requirements, and not towards finding a real solution.”

Timing matters as well. An organisation moving towards the “Balancing” stage of the ABC archetype takes on cost and complexity well before it sees any return, which is exactly the profile a one to two-year grant cycle punishes. Judging its finances at the moment the grant ends catches it at its worst point.

The organisations that shifted the furthest across archetypes in our study shared one thing: leadership that had accepted that the organisation needed to become something different. That mattered more than market opportunity did, and it strengthens the case for funders to work with NGO decision-makers themselves to get shared buy-in for full organisational change, rather than treating this diversification like another small-scale innovation project within the organisation’s broader programming. But this is as much a behavioural shift for funders as it is for organisations. It requires a new psychology of grantmaking, where the goal stays fixed, but the ways to get there can change based on what you learn.

Redefine what scaling success looks like: Funders also need to broaden what counts as progress. During a transition to new revenue streams, most of an NGO’s progress is invisible on a standard grant report. This reporting isn’t designed to capture the capabilities and systems an organisation needs to deliver at scale, such as a viable pricing model, new commercial capability, stronger governance, a partnership that establishes who carries the work, or the evidence that a business model should be abandoned before more is spent on it. Sometimes the end result of this process is that the NGO decides that it should continue to be predominantly grant-funded, or that an external partner is better placed to take an innovation forward.

The ABC model makes that conversation more specific. Instead of assuming every organisation should travel from grants toward earned revenue, it encourages funders to ask three questions: Where is this organisation now, where does it need to go to achieve greater impact, and what capabilities, structures and funding will it need to get there?

If funders are serious about scaling impact, they have to pay for the transition itself, not just the pilot that precedes it and the scale they hope will follow. That means transition capital: flexible funding, committed for the years an organisation spends becoming something different, that pays for the partnerships, governance, structures and skills the ABC process demands. Until that money exists, the sector will keep asking NGOs to cross a gap it refuses to fund.

Disclosure: Enabel is the Belgian Agency for International Cooperation. The Innovation Hub is Enabel’s innovation unit, funded by Belgium and the EU, supporting mission-driven organisations active in Africa and the Middle East in scaling innovative solutions to complex societal issues. The Enabel Innovation Hub is supporting the organisations mentioned in this article, IMWI and Farm Radio International, with grant funding and capacity building technical assistance. It also commissioned and published a mixed methods study which interviewed Aflatoun, Green Farmlands and myAgro. 

Toon Driesen is the manager of the Enabel Innovation Hub, funded by Belgium and the EU; Alex Losneanu is Innovation Director and Pritika Kasliwal is Innovation Lead at Brink, which is a partner to the Innovation Hub.

Photo credit: cagkansayin

That realisation made us rethink everything. My co-founders and I saw that, if we wanted children to arrive at school ready to learn, we had to reach them earlier, which meant supporting their parents from the time of the child’s birth. To help meet that need, in 2016 we founded ToguMogu, an AI-powered digital parenting platform in Bangladesh that has reached more than 476,000 registered users over the past six years, at parenting stages ranging from pregnancy through age 10. Alongside its AI assistant, it connects parents to a network of experts, including paediatricians available through teleconsultation and a community of hundreds of thousands of other parents, while also offering growth tracking, expert-reviewed content and other services. 

Light of Hope and ToguMogu were not separate ideas. They emerged from the same insight: that a child’s development is continuous and starts at birth. We simply applied that insight to two different stages of childhood: the early years from 0-5, and the primary school years from 4-12.

After nearly a decade of building services for parents navigating both of these stages of childhood, we’ve developed a clear view of where parenting technology is failing families, not just in Bangladesh but globally. Despite being worth hundreds of millions of dollars and growing at double-digit rates, the parenting technology industry has three structural blind spots that are getting more serious the faster the sector grows. These blind spots exist in wealthy and emerging markets alike, but their consequences are most severe in low- and middle-income countries where families have the fewest options for child development specialists or other alternative professionals to turn to.

The First Blind Spot: Parenting Apps Are Built for Mothers, Not Families

Walk through the onboarding flow of almost any parenting app and you will notice something immediately: The language, the content and the community features assume one user, the mother. This reflects a real imbalance in caregiving, where mothers worldwide perform 2.5 times more unpaid care work than fathers. But by designing only for mothers, these platforms reinforce that imbalance rather than helping to close it.

Research from developed markets reveals how this gender bias is often built into popular parenting apps, as shown in a 2021 study of digital parenting guidance in Canada. This design choice not only reduces the chances of fathers getting involved, it also often gives mothers the added responsibility of “curating” digital parenting guidance for their partners — a dynamic that doesn’t serve the interests of either parent (or their children).

We saw this pattern in our own platform and set out to change it. Of ToguMogu’s nearly half a million registered users, over 164,000 are men, roughly 35% of our user base. We deliberately designed the app to include fathers, framing them as active participants in a shared parenting journey instead of secondary observers of the mothers. Once we designed for fathers, they showed up.

This matters far beyond app design. When fathers are engaged and supported in early parenting, children do measurably better. Research on parenting programmes indicates that engaging fathers, not mothers alone, improves both the effectiveness of the programme and the outcomes for the child. A parenting platform that engages the whole family is not just being inclusive; it is investing in better developmental outcomes for the child, which is the entire point.

The Second Blind Spot: The Age-Five Cliff

Parenting apps tend to be heavily concentrated on children’s early years. This focus is well founded: Around 90% of a child’s brain development happens before age 5, and early connections lay the groundwork for later cognitive ability. But treating age 5 as a finish line is a mistake.

Development does not stop when a child starts school. Research on middle childhood, which spans the ages of 6-12, often describes this period as the “forgotten years,” since it is overlooked by most education research despite being a time of rich cognitive, social and emotional growth. Middle childhood is when children consolidate learning habits, self-regulation and social skills, yet it is precisely the stage at which most parenting apps disengage from the family.

When a child enters school, the parenting technology sector largely hands the family over to the education technology sector. Some edtech tools connect to families through school-to-parent messaging and home engagement features, but that is a different job from supporting a parent in the developmental work of raising a 7-year-old at home: managing big emotions, building independence, and navigating a child’s first real social conflicts. Currently, few tools serve that need, which leaves parents of primary-school-aged children with less support than they had when their child was an infant.

At ToguMogu, our usage data showed that families did not stop seeking guidance once their child started formal school; questions simply shifted from topics like feeding and sleep, to issues like behaviour, learning and schooling. In response, we expanded ToguMogu to serve children up to age 10, adapting our content and expert network in recognition of the fact that guidance for ages 0-5 does not transfer neatly to the school years. However, the broader sector has largely not made that realisation, and families in emerging markets, where school systems are least able to fill the gap, pay the price.

The Third Blind Spot: AI Without Accountability

As the head of a digital platform that integrates artificial intelligence into its service delivery, I’m familiar with both the upsides and downsides of the widespread incorporation of AI into parenting apps. ToguMogu’s AI assistant provides parents with insights based on their child’s developmental progress, and I have seen firsthand how useful this can be, and how easily it can go wrong. UNICEF’s guidance for child-centred AI calls for regulatory oversight, safety, fairness, transparency, explainability and accountability. Yet no consistent standard, either regulator- or industry-led, holds parenting tools to those principles, even when they leverage AI to offer developmental assessments, behavioural guidance and health recommendations for children — areas where the wrong advice can have significant negative consequences.

The risk is real, and it is particularly relevant in the Global South. Studies show that while parenting programs are generally successful in improving children’s behaviour, their effectiveness may not be easily replicated in low- and middle-income countries, where ignoring unique cultural contexts can worsen existing inequalities and cause harm. The difference between developed and emerging countries here is cultural, not technological: Many parenting interventions in the Global North are based on values like individuality and independence, which run contrary to the collectivism, emotional restraint and joint-family child-rearing that characterise parenting across South Asia and other emerging regions. A tool that treats one culture’s parenting norms as universal therefore risks judging another culture’s practices as deficient. That’s why parenting technologies must be tested locally, built on local data and kept under human oversight: As research has found, this sort of deep local adaptation yields better results than surface-level changes, like simply translating an interface from one language to another.

Education technology already offers a cautionary tale. UNESCO has warned that generative AI is advancing faster than national regulation and that many educational institutions remain unprepared to validate the tools they adopt. Parenting platforms risk repeating that pattern with even higher stakes. The children affected include infants and preschoolers, and in many emerging markets, families have no clinician, health visitor or specialist to turn to for a second opinion when an app gets it wrong. The app may be the only source of guidance available.

What Funders and Development Organisations Should Do

Over a decade of building ToguMogu and Light of Hope, we have fallen prey to each of these three blind spots ourselves, and we know they are fixable. But they will not fix themselves. The funders, governments and development organisations that support and partner with these platforms should put their leverage to use to advocate for the following changes.

Require father-inclusive design as a condition of support: Any platform receiving development funding or an institutional partnership for early childhood outcomes should have to show active engagement from fathers and co-parents, not just mothers. Measure it. Fund it. Expect it.

Fund the 5-8 window deliberately. The move into school should not be a drop-off point in support. Development organisations working on education should actively seek out and back platforms that bridge parenting technology and school-age support. That infrastructure barely exists today, and it needs to be built.

Set accountability standards for AI in early childhood before a crisis forces them. Funders and institutional partners should require, at a minimum, that AI-driven parenting tools disclose the make-up of their training data, undergo independent clinical validation before deployment, and tell parents clearly when guidance is AI-generated rather than reviewed by a clinician. The sector will not do this quickly on its own. External pressure is what will move it.

None of this requires punishing innovation or demanding costly trials for every feature. It requires that the organisations funding this sector attach basic expectations to their support. The families using these platforms in low- and middle-income countries are not beta testers. For many, a parenting app is the closest thing to professional early childhood guidance they will ever receive. That responsibility deserves more care than the sector currently brings to it. At ToguMogu, partners, including the United Nations Population Fund and Bangladesh’s Directorate General of Family Planning, have pushed us toward higher standards than we would have set alone. That kind of institutional pressure works. The development community should apply it far more widely.

Zillul Karim is Co-Founder and Strategic Advisor at ToguMogu, and serves as Board Director at Light of Hope Limited.

Photo credit: Pressmaster

Staff are regularly using free, publicly available AI tools to summarise evaluation reports, translate documents, draft proposals, tidy up their field notes and prepare donor updates. Some do this carefully. Others do it with little sense of how outputs should be checked, where their own human judgement must be used, and what types of training data should never be uploaded.

Leadership responses to this off-the-books AI use vary. Some leaders I’ve worked with know it is happening and look away. Some suspect it and hope the question resolves itself, or may even embrace staff experimentation with AI tools in the absence of an organisational policy. A few others have restricted or banned staff use of AI tools.

That last response deserves attention. In a sector rightly alert to technology’s risks, restricting or even banning staff use of AI tools may feel like the responsible position for many development and impact organisations. In practice, it is often the opposite. Saying no does not keep AI out of an organisation; instead, it keeps AI use invisible, and invisibility is the one condition under which none of AI’s risks can be managed.

The AI ‘Governance Vacuum’ in Global Development

This issue isn’t merely anecdotal. In a major survey by the Humanitarian Leadership Academy and Data Friendly Space, covering 2,539 humanitarian workers across 144 countries, seven in 10 reported using AI daily or weekly, while fewer than a quarter of their organisations had any formal AI policy in place. The same survey found that almost two-thirds of organisations offered staff little or no AI training, even as most staff used the tools regardless. The researchers referred to the resulting misalignment as a “governance vacuum”: individual use running well ahead of the institution’s capacity to guide it.

The timing makes this lack of guidance harder to ignore. Official development assistance from DAC members fell by 23.1% in 2025 to US $174.3 billion, the largest annual contraction on record, with core contributions to the UN system down 27% and bilateral programming cut deeply. Organisations are being asked to protect service delivery with fewer people and tighter budgets, as well as to justify every cost while doing so. Meanwhile, AI tools keep getting cheaper, more capable and easier to access. Anyone with a browser can use one, and many staff already do.

So the question of whether these organisations will adopt AI was settled some time ago, without fanfare, one prompt at a time. What remains unsettled is the governance that should underpin this growing AI use. On one side sit staff with heavier workloads and rising expectations. On the other sit boards and senior leadership teams that still tend to treat AI as a future strategy question, a procurement decision or a reputational risk to be contained. Between them sits the real-world organisation: people using AI tools informally because they help, while policy, training and accountability trail behind.

The Three Downsides to Restricting AI

None of this is an argument for rushing in, no holds barred, to adopt AI. Organisations operating in the international development and social impact sectors handle sensitive data, work with vulnerable communities and operate in charged political contexts. Badly governed AI usage can result in exposed data, fabricated evidence, biased analysis, and weak conclusions concealed by false confidence. These risks are real. But addressing them will require clear safeguards and informed oversight rather than disengagement.

That’s because AI restrictions and bans are weak forms of protection and carry costs of their own. Three stand out.

The first falls on staff. When an organisation avoids the issue, every judgement call devolves to individuals. The programme officer using AI to summarise a 90-page impact evaluation must decide alone whether the summary can be trusted. The fundraiser drafting a concept note must decide alone what AI outputs are safe to paste in. The monitoring officer experimenting with qualitative analysis must work out, unaided, whether the AI tool has flattened what respondents actually said. These are hard questions even for specialists. Organisations that refuse to engage with them are not sparing staff the burden; they are transferring that burden to them, without providing the rules, training or backup that could help staff navigate it.

There is an equity dimension here too. When organisations provide neither practical guidance nor safe routes to experiment, staff are left to make their own calculations about AI’s risks and rewards. More confident users may save time, but they may also expose sensitive information or rely on plausible-sounding but erroneous outputs; more cautious colleagues may avoid those risks, but also miss legitimate opportunities to work more effectively. The result is neither fair nor controlled. Individual confidence and risk tolerance rather than organisational priorities and consistent safeguards determine who uses AI, how they use it and who bears the consequences.

The second cost involves organisational learning. The most useful applications of AI in social impact and global development work are rarely the controversial ones. They sit in the unglamorous middle of these organisations’ work: synthesising lessons across projects, making institutional knowledge findable, drafting first versions of routine documents, comparing proposals, and pressure-testing assumptions before a decision is made. These workflows are precisely the ones many organisations struggle to resource, and precisely where AI can help without displacing expertise or automating anything sensitive.

Leadership teams that either avoid AI or respond to staff AI usage mainly by restricting tools are unlikely to discover this middle ground. They also lose the chance to build the judgement and safeguards needed for higher-stakes uses through supervised practice on low-risk work. If organisations defer that learning, they may eventually face a consequential AI decision with no institutional experience to draw on.

The third cost lands on partners and communities. All social impact organisations — from large UN agencies to grassroots NGOs — worry, rightly, about extractive or carelessly used technology. But unmanaged AI is more dangerous than visible, bounded AI. When staff paste a partner’s internal documents into public tools with no data rules, when community feedback is summarised without anyone checking for lost meaning, or when AI-polished language smooths the uncertainty out of a report, the risk is carried by the people least able to see or challenge the process. Workplace AI bans prevent none of this. They simply guarantee that nobody is watching for it.

Some organisations have grasped this. Mercy Corps developed internal generative AI chatbots grounded in its own digital library, giving staff a safer way to use organisational knowledge while reducing the risk of sensitive data being pasted into public tools. The United Nations Development Programme (UNDP) took a more specialised approach: Its Artificial Intelligence for Development Analytics platform gives staff and the public multilingual, source-linked access to evidence from nearly 7,000 evaluation reports. Launched in 2022 and substantially upgraded since, more than three quarters of users reported that it made evaluation evidence easier and quicker to use, although UNDP acknowledges that uptake remains uneven.

Guidance for Responsible AI Adoption

Responsible adoption starts from a different premise: AI is already inside the organisation, so leaders first need to understand how it is being used. Strategy, policy and procurement still matter, but they should respond to real workflows and risks rather than preclude any investigation of them.

It begins with mapping real staff use. This should be a practical exercise to establish where staff already use AI: which tools they prefer, what information they’re feeding in, and which outputs they’ve come to rely on. This exercise will highlight both where staff are successfully using AI, and where they feel out of their depth. The aim is not to shame experimentation. Rather, it is to make invisible practice visible enough to govern.

The next step is to select a small number of existing or proposed workflows for formal, supervised use. Strategy, policy and procurement should inform that choice, but they should not substitute for examining the work itself. Good starting points are low-stakes, internal tasks grounded in source material and easy for a person to verify — for example, summarising public documents, preparing internal meeting notes, compiling donor information from published sources, or checking a human-written draft against agreed-upon criteria such as factual accuracy, required content and donor rules. These are useful entry points because staff can practise source discipline, verification and human review before using AI with sensitive data or consequential decisions.

Alongside this, staff need simple data boundaries. Public information, internal operational material, personal data, partner documents and community-level data should not be treated alike. Most teams can begin with plain-language categories, concrete examples of what staff may and may not upload in common situations, and a clear route for advice when the answer is uncertain. They also need defined human review points. The higher the stakes, the stronger the requirement that a person checks sources, tests assumptions and makes the final decision. AI should be treated as a tool that can accelerate parts of a workflow while producing errors plausible enough to escape casual review — not as a colleague whose judgement can be trusted.

Finally, boards and senior leaders need enough AI literacy to govern its use. They do not need to become technologists, but they must be able to: distinguish between controlling access to tools and controlling the risks within particular workflows; understand how data protection, model error and human review requirements vary by use case; and judge whether staff have the training and approved tools to work safely. They should be asking: Do we know where AI is already used? Which tools and data are permitted? Where must a person verify sources or make the final decision? And who owns the quality, compliance and consequences of each AI-assisted workflow?

This is a manageable agenda. But it requires leaders to move from avoidance to stewardship.

The central question for impact-driven organisations is no longer whether or not AI should be used — that ship has sailed at most organisations across the sector. The challenge now is to determine where AI can responsibly improve their work, where it should be kept out, and what staff and leadership capability is needed to tell the difference. Saying no may feel prudent, given the organisational risks this technology can bring. But if prohibiting AI has the practical effect of increasing unmanaged use, fostering uneven staff practices and weakening oversight, it is not prudence. It is risk in another form, borne by the people the sector exists to serve.

Loksan Harley is the Founder and Principal of Homelands AI.

Photo credit: wildpixel

So how big an impact is this challenge having on women entrepreneurs, and what can be done to address it?

We already know that fraud is among the well-documented barriers women face when accessing digital financial tools and services. What’s missing is data on the unique risks women experience once they have already gained access to formal financial services or adopted digital tools — particularly in terms of their exposure and response to digital financial fraud, and their ability to recover from it.

Both CARE and Innovations for Poverty Action (IPA) have recently collected more data on this issue, via the Midline Survey from Strive Women, a Mastercard Strive program implemented by CARE, and from IPA’s Financial Consumer Protection Surveys. We combined our datasets to answer a few questions: Is there a specific story for women around digital risk and fraud? Are they more vulnerable? And what can we do to improve their resilience?

Are women more likely to be victims of fraud?

IPA’s surveys from seven countries find that women, on average, report fraud attempts 6 percentage points less often than men. But this gap shrinks once you account for differences in education, country, employment and financial literacy — factors closely tied to digital and financial platform usage. In other words, once you compare women and men with similar levels of education, employment and financial literacy, they report fraud attempts at about the same rate. Women aren’t exposed to fraud less because they’re women; they’re exposed less, in part, because they use formal digital and financial tools less regularly.

Similarly, data from the Strive Women program shows this pattern directly. Surveying over 1,000 entrepreneurs from Peru, Pakistan and Vietnam, it found that their reported rates of encountering fraud (attempted or successful) in the past 12 months are low (21% in Vietnam, 19% in Peru and 12% in Pakistan). A key determinant of fraud exposure is the entrepreneur’s country context and their usage of digital platforms: Entrepreneurs who used more digital tools for their business were more likely to experience fraud, regardless of gender. In other words, the very tools driving business growth are the same ones driving fraud exposure.

However, CGAP research shows that once contact from a fraudster happens, women are more likely to be affected by fraud: less equipped to identify a scam or resist social engineering tactics that exploit trust or urgency, and less supported by the systems meant to help when something goes wrong.

How does fraud affect women and men differently? 

Our findings reveal several areas where fraud has different effects on women and men.

Being prepared: Women may be slightly more attuned to digital risk than men. Data from Strive Women shows that 75% of women rate cybersecurity as important to their business, compared to 68% of men. But that added awareness doesn’t translate into more action: About a third of both women and men still take zero protective steps, and when entrepreneurs do act, they almost always default to the same single habit — a strong password — with far fewer using two-factor authentication, backing up their data or using antivirus software.

Knowing what to do next: Twice as many women as men in the Strive Women sample said they would not know what steps to take if their digital account were compromised, despite reporting high confidence using digital tools day-to-day. IPA’s data across several countries tells a more mixed story on formal complaints: Women were somewhat more likely than men to file complaints in Ethiopia and Pakistan (note: the data from these countries is not yet publicly available), roughly on par in the Philippines and Kenya, and less likely in Tanzania, Bangladesh and Uganda, where the gap reached about 17 percentage points. This spread suggests that the broader cultural contexts and financial systems women are navigating may affect their ability to seek help. Evidence from India points to one manifestation of this challenge: Women are more likely to turn to family or friends before pursuing a formal complaint, and fear of family backlash can discourage them from going further on their own — meaning the decision to seek redress is often a social one, not just a practical one. And even when a woman does clear that hurdle, the system doesn’t always meet her halfway. As one Strive Women entrepreneur in Peru put it: “There was money taken from my card. I went to the bank. They told me I had made that withdrawal, but I had not.”

Bouncing back: The Strive Women sample shows that men also have a larger financial runway to cope with a shock or emergency: 22% of men have enough capital to run their business for six months or more if something goes wrong, compared to only 14% of women. Without that cushion, a single fraud event threatens the sustainability of their business. IPA data also finds that women across the surveyed countries are 8 percentage points less likely to have capital reserves than men.

What actually helps women avoid fraud

Consumer education can help women detect and avoid fraud, depending on how it is designed and delivered, but education alone can’t solve the problem. In rigorous evaluations of fraud education programs in Kenya and Nigeria, IPA found that standalone tip sheets and training increased women’s confidence in recognizing fraud, bringing it closer to men’s. However, these initiatives did not similarly improve women’s ability to accurately identify fraudulent messages. In other words, participants felt more able to spot fraud but were not measurably better at it.

The exception was a program in Uganda that used interactive, story-based lessons built around realistic fraud scenarios. Users navigated common scams by making decisions at key moments, allowing them to practice how to respond before encountering fraud in real life. The program reduced the share of users who lost money to fraud by 1.4 percentage points overall and by 2.6 percentage points among women. Beyond training, a well-timed nudge or notification at the right point in a user’s digital journey can build awareness just as effectively. Better still is to build security into the product itself, so that safety doesn’t hinge on a behavior change — e.g., through default protections like automatic multi-factor authentication, transaction alerts, safer onboarding and role-based account permissions.

Redress mechanisms also need to be offered through channels women already trust. Many women still prefer an in-person agent to a chatbot, and routing support through community structures such as women’s groups and peer networks can reduce the stigma that keeps fraud victims from reporting. Peer networks and endorsements can also support women’s financial safety. In Ghana, an IPA evaluation found that training group leaders in mobile banking and encouraging peer endorsement were especially effective at building women’s confidence to handle fraud. IPA research also finds that offering free, easy-to-access legal aid can increase redress resolution rates for women slightly more than men.

But in many cases, providers do not need new programs or policies: They just need to make better use of the data they already collect. IPA’s work with a Ugandan telecom regulator found that categorizing and segmenting existing complaint data by sex could reveal where to focus next in fraud prevention policy.

Let’s stop treating digital safety as an afterthought 

Women entrepreneurs do not need to be persuaded to use digital tools. They already are. But if we want them to continue reaping the benefits of digitization, what they need are systems that make those tools safe.

That means moving beyond consumer education as the primary line of defense. Instead, providers should embed protections into products from the outset through secure defaults, clear transaction alerts, intuitive reporting channels, and recovery mechanisms that are free and easy to navigate. Regulators should require providers to collect and analyze sex-disaggregated fraud and complaints data, and set institutional standards for delivering effective consumer protection — not just access — across genders.

The success of digital financial inclusion should no longer be measured only by how many women adopt digital tools, but by whether they can use those tools safely, with confidence — and whether they have the ability to recover when something goes wrong.

For more information about the research in this article, contact us by email at financialinclusion@poverty-action.org and entrepreneurship@care.org.

Tanvi Jaluka is the Research and Learning Lead for CARE USA’s Women’s Entrepreneurship team; Lauren Perlik is an Associate Program Manager for Innovations for Poverty Action’s Financial Inclusion Program.

Photo credit: AndreyPopov

But laying the groundwork alone will not deliver transformation. As AI moves from a research frontier to a practical tool for governments and service providers across the Global South, MDBs are confronted with a harder challenge: What will it take to actually deploy AI at scale in LMICs, and are they set up to finance that?

Investing in components rather than systems

The AI investment landscape within MDBs today is wide but thin. There are pilots, such as an AI tool for tax administration in Armenia, responsible AI experiments with municipalities in Chile, and innovation challenges for road safety in Asia. There is infrastructure, such as data centers in Thailand and Azerbaijan, broadband expansion across West Africa, and digital identity programs in the Philippines. There are enabling environments, such as cybersecurity support across dozens of countries, regulatory sandboxes, and AI policy advisory.

Each of these initiatives is valuable, and they span across the “AI stack,” including AI applications themselves; enablers like policies, regulations and digital public infrastructure; and foundational technologies like internet access, data and compute infrastructure, and digital devices. Taken together, however, they show that MDBs are still in the discovery phase, investing in components rather than systems, and in pilot programs intended to demonstrate feasibility rather than programs that aim to scale.

What’s largely missing is a financing approach that connects AI use cases to the underlying foundations they depend on (i.e., the enabler and foundational technology components of the full AI stack), and that moves from proof-of-concept to national-scale deployment within a coherent MDB program.

What scaling AI actually requires 

It will take more than funding individual use cases and building data infrastructure to achieve large-scale AI deployment in an LMIC context. Large-scale AI roll-out will require MDBs to get five things right simultaneously, spanning across multiple sectors and working with government ministries that rarely plan together. These include:

• Use cases that are technically sound, locally adapted and designed for the realities of end-users in often under-resourced environments — e.g., community health workers with basic smartphones or smallholder farmers with intermittent connectivity.

• Data foundations or interoperable platforms that connect to existing government systems, with governance arrangements that determine how data is collected, stored and used.

• Compute and connectivity calibrated to the deployment context. Consider a national deployment of AI in healthcare: If it can only run in urban environments, it hasn’t solved the problem of access across the country.

• Devices such as smartphones and tablets in the hands of frontline workers. This remains the most consistently underfinanced layer in the stack, despite being the final link in the chain that decides whether AI-based systems reach traditionally excluded communities or not — and despite the fact that smartphone ownership across LMICs stands at 50%, and just 24% in Africa (as of 2024).

• Enabling policies such as data protection regulations, procurement frameworks, and sector-specific guidelines that allow governments to deploy AI responsibly and at speed.

Many MDBs’ digital and AI strategies articulate these layers clearly but fall short on operationalization. In practice, ensuring that an AI strategy can be effectively executed means addressing these layers in a single, cohesive program as opposed to having them spread across separate projects, teams and financing instruments.

What an AI system with large-scale public benefits could look like 

Consider what an integrated approach might look like in primary healthcare, a sector where the development case for AI is strong, and the infrastructure gap — including both ill-equipped facilities and constraints on provider availability and capacity — is well-documented.

The goal of such an approach would be to create a system that gets ahead of illness instead of just responding to it. Every citizen, regardless of where they live, would receive continuous, personalized health support. That would include preventive care through ongoing monitoring and early detection, delivered by community health workers at people’s doorsteps.

Achieving such an ambition would mean investing across a suite of interconnected AI use cases rather than a single tool, including:

• A clinical decision-support system that helps community health workers diagnose conditions accurately in low-resource settings.
• An administrative layer that reduces the documentation burden on those same workers, enabling them to spend more time with patients and less on paperwork.
• A referral coordination tool that ensures that patients who need higher-level care actually get to the right facility, and that their records follow them.
• A patient tracking system that enables longitudinal monitoring, flagging individuals with deteriorating health or those who have missed critical follow-ups.

Individually, each of these use cases makes an interesting pilot. Taken together, they make up a transformational system.

However, building this system requires a shared infrastructure: a unified data platform that interoperates with the government’s existing health information systems to link a person’s longitudinal health data to their existing national ID, enabling continuity of care between providers and informing the government’s population-level research; data protection regulations that govern how patient data is collected and used; and devices in the hands of every frontline worker.

In addition to infrastructure, capacity building must be a core investment. Community health workers need to be genuinely equipped to use these tools on an ongoing basis, and government health teams must have the ability to manage, evaluate and own these systems without relying on external technical partners to keep them running. At both the grassroots and local governance levels, this requires structured, continuous support that reflects varying levels of digital literacy and local language needs, while accommodating the often-unpredictable realities of fieldwork.

At the same time, project timelines need to shift to ensure that solutions do not become obsolete before they are even launched. With AI performance doubling every seven months, having a six-to-12-month lead time from approval to first deployment, rather than the two-plus years typical of large MDB operations, matters enormously. And active technical support cannot end at go-live; it must continue through the scale-up phase. This can help governments avoid vendor lock-in and dependence on a single provider, manage data governance, and progressively own these systems over time.

Five shifts worth considering to scale systematic AI implementation

Moving from experimentation to systematic AI financing at scale is more an institutional challenge than a technical one. It involves rethinking how MDBs frame ambition, scope programs, structure financing, run operations and deliver advisory services, and includes the following five shifts.

From component funding to sector transformation: This shift reframes the question from “what AI investment can we structure?” to “what would it take to transform this sector through AI?” That reframing shapes everything downstream, including the scope of the program, the mix of instruments, the partners involved and the metrics of success.

Across the full stack, in a single program: This shift requires an institution to take stock of what already exists in a country (e.g., internet connectivity, data systems, workforce capability, regulatory environment) and build a curated set of investments that address the specific gaps between where things are and where they need to be for AI to work at scale.

Financing calibrated to scope: Compute infrastructure, data platforms and AI applications have fundamentally different risk profiles and cost structures — from tens of millions of US dollars for sub-national phases to billions for a national rollout — and a one-size financial instrument is unlikely to serve all three well.

Procurement and operations designed for AI realities: This shift requires procurement that: prioritizes digital public goods, reusable building blocks from comparable countries and MDB-developed tools; treats interoperability as a non-negotiable requirement; and avoids vendor lock-in. It also requires lead times under 12 months; project durations that account for active support beyond the launch date; and evaluation frameworks that capture adoption, use and deployment.

Deep technical advisory front-loaded: This shift involves helping governments: understand AI’s cost reality from the outset; make smart early decisions on model selection, dataset localization and the use of existing digital public goods; and manage the risks that are specific to AI, such as data interoperability failures, low user adoption, and systems that work in pilots but cannot sustain themselves after project financing ends. Two areas deserve particular attention:

• Supporting governments in choosing the right AI architecture: MDBs must help governments work through on-the-ground realities such as network connectivity in rural areas, the processing capability of devices used by frontline workers, power infrastructure, and available budgets. These factors inform choices between frontier, cloud-enabled large language models; offline-capable, cost-effective small language models; or AI-in-a-box solutions that may be equally or more effective depending on the context.

• Helping governments understand the sovereignty stakes of AI procurement: AI procurement carries strategic implications that physical infrastructure never did, and most governments are ill-equipped to navigate them. MDBs should help governments think through where data will be stored, where compute will reside, and who will control the models underpinning critical public services. These decisions, once locked in, are costly and difficult to reverse.

The development opportunity in AI transformation 

MDBs have been helping to build the infrastructure layer of the digital economy across the Global South for years. The next step is to ensure that the communities this infrastructure was built for actually benefit from AI transformation at scale.

AI can extend diagnostic reach in healthcare, strengthen agricultural advisory, and improve government service delivery across other sectors and societal needs, while supporting evidence-based policymaking — at a scale and cost few other interventions can match. As the 2030 deadline for the Sustainable Development Goals approaches, LMICs continue to face an estimated US $4 trillion annual SDG financing gap. AI alone won’t close that gap, but used well, it can make every dollar of development financing go further.

Kunal Walia is a Partner at Dalberg Advisors.

Photo credit: Urupong

Consider a four-year-old dried food cooperative CARE worked with in West Africa: It has active export contracts to the European Union, audited financials, and a defaulted loan rate of zero. Its immediate financing need is $40,000 in working capital to meet seasonal demand. Yet it has struggled to acquire this funding, as there are few commercial products designed for that ask. The cooperative is too large for microfinance, too small for micro, small and medium enterprise (MSME) funds, and too aggregated for direct equity. So the capital does not flow.

This challenge goes beyond one cooperative. It affects an entire asset class that the impact investing field does not name, size or build products around: aggregated MSMEs. These consist of individual enterprises that have joined together to contract and borrow as one, either formally as a cooperative, or informally through a producer association or collective enterprise that brings multiple individual MSMEs together around a shared market, value chain or commercial activity.

At CARE, we work across all of these different forms of aggregation, and see them emerging across low- and middle-income markets. We help these MSMEs organize and strengthen their businesses, often leveraging funding from the savings groups we help them join or organize. We also support them through other producer and enterprise networks, and connect them to financial institutions and markets. As a result, we’ve had a direct view of both the investment opportunity these aggregated enterprises represent, and the reasons many funders are missing this opportunity.

The Asset Class Hiding in Plain Sight

Aggregated micro, small and medium enterprises exist in sectors like renewable energy, digital commerce and the care economy, but they have primarily been concentrated in agriculture, trade, processing and related services. Whatever their focus area, these aggregated MSMEs sit in a structural gap that current investing architecture has not been designed to fill, with loan needs that typically range from $500 to $50,000.

However, their financial behavior is becoming increasingly visible. For instance, Root Capital’s track record over two decades demonstrates that aggregated agricultural enterprises can be a viable lending segment when financing models are designed around their realities. As an impact investor, it has provided debt financing to hundreds of agricultural MSMEs, including some that aggregate smallholder producers, and evidence from its portfolio shows that many of its borrowers (42%) later gain access to commercial finance.

This ongoing traction with funders shows that the aggregated MSME segment includes viable borrowers with evidence of successful repayment. The missing piece is a recognition from other impact and commercial asset managers of these enterprises’ investment potential.

Why the Gap Persists

The impact investing field has organized itself around instruments rather than enterprise types. Conferences debate first-loss structures, guarantees, viability gap funding and concessional debt mechanics. Far less attention is paid to the question of what asset class those instruments are designed to serve, including at the bottom of the value chain. In the investment structuring conversations I sit in, the debate is almost always about the design of the instrument: where first-loss capital sits, how guarantees are priced, how the tranches stack. The questions of which borrower segment the deal is meant to serve — or whether anyone has even defined the segment — rarely come up.

The result is a market with sophisticated supply-side architecture but a less-developed demand-side infrastructure for aggregated MSMEs. The market lacks widely adopted underwriting standards specific to this asset class, as well as shared methodologies for sizing the addressable market, and for pricing and compensating the origination work required to quantify these borrowers’ informal economic activity so they can become investable. Of course, there are real frictions — like high transaction costs, currency risk, weak legal enforceability and limited exit pathways — that compound the problem and reduce investors’ incentive to solve it. These are not insurmountable, but they remain unaddressed because impact investing has not yet widely recognized aggregated enterprises as a distinct investable segment.

A recent piece in NextBillion by leaders at Pangea Africa and Social Finance International made an important contribution to the conversation around strengthening the broader impact investment pipeline, by proposing a financing model for business advisory service providers serving African MSMEs. The argument here goes one level deeper: Before investment facilitation can scale, the asset class itself needs to be defined. This requires a shared understanding of the enterprise type, common approaches to underwriting and measurement, standard risk and return characteristics, and the infrastructure to connect capital with investable opportunities. Without these shared definitions and connections, capital providers may continue to view these enterprises as fragmented, high transaction-cost opportunities that do not fit their established investment models.

What Building the Aggregated MSME Asset Class Requires

To build this asset class, the impact investing sector must focus on the three key priorities that would change the status quo and make aggregated MSMEs more visible and accessible to investors: formal definitions, underwriting standards and pricing the origination function.

First, a formal definition of the asset class — with sizing data — is needed. This will require a credible global sizing of the addressable market, along with consistent reporting categories that allow capital allocators to compare risk, return and performance across investment opportunities (including across geographies and aggregation models), and that facilitate portfolio construction at scale. The SME Finance Forum provides a useful precedent: By collating official MSME definitions and related data across economies, it has made a highly diverse segment more visible and measurable, even though national thresholds for what qualifies as a “micro, small or medium” enterprise vary. The Global Impact Investing Network, Convergence and leading development finance institutions already perform similar field-building functions across the capital ecosystem by convening market actors, developing common frameworks and aggregating data, making them well-placed to establish shared definitions, sizing methodologies and reporting categories for aggregated MSMEs.

Next, we need to build underwriting standards for aggregated enterprises, not just individual ones. Aggregation changes the underwriting calculus in ways that traditional MSME credit models do not capture, because characteristics that exist at the collective level — group governance, cooperative financial discipline and verified repayment behavior — become material indicators of creditworthiness that may stand alongside or in place of some indicators used to assess individual enterprises. These signals should be formalized into standard underwriting frameworks that ultimately apply to both impact investors and commercial lenders, so capital providers can assess aggregated MSMEs’ creditworthiness and price these loans more consistently, extending the pool of available capital beyond specialist lenders.

Finally, it is imperative to price in the origination layer. The work of moving enterprises from early-stage and financially underserved into investable form is typically financed by donor-funded technical assistance and business development services rather than through a market mechanism that explicitly prices the origination function. These funding sources will remain important, but they should be only one part of the financing model, particularly as concessional resources become scarcer. Origination has a price that ought to be considered, structured into deal architecture, and rewarded as a critical component of the capital deployment process. If the functions that create investment-ready enterprises remain uncompensated, they will continue to be underprovided despite the value they create for capital providers. Origination fees, technical assistance funding structured within deal architecture, and performance-based payments tied to capital deployment are all workable models worth exploring.

A Working Example of How to Finance Aggregated MSMEs

My perspective on how best to finance aggregated MSMEs comes from leading CARE’s Rise platform. At the heart of Rise is a simple observation: Many financially underserved microenterprises become investable when they are aggregated through savings groups, producer organizations or cooperatives and connected to financial institutions as a visible, assessable pipeline.

CARE has enabled access to savings groups for over 30 million people across 67 countries, with about 25 individuals in each group — many of whom use these funds to establish individual and group enterprises. Aggregating these MSMEs into producer associations, cooperatives and other organized enterprises creates an investable unit that can be assessed, strengthened and connected to commercial finance. These aggregated MSMEs form a pipeline through which enterprises can access growth capital from financial institutions, and enter commercial value chains by engaging with input suppliers and buyers.

We work in both directions to build this pipeline: On the MSME side, CARE helps enterprises strengthen governance, achieve investment readiness and digitize their financial records. On the funder side, we work with financial institutions, buyers, governments and local partners to improve market access and connect enterprises to appropriate financial products. Where aggregated MSME markets have not yet matured, we use instruments such as time-limited first loss guarantees, origination incentives and structured technical assistance to help bridge the costs of market entry. Taken together, these activities (among others) perform the origination function by transforming fragmented enterprises into investment-ready opportunities.

Our results illustrate what building an asset class looks like in practice. For instance, CARE partnered with Kazi Yetu, a Tanzanian tea social enterprise, to help women tea farmers move beyond primary production into value addition and formal markets. As part of this effort, members of some of the local savings groups we facilitated evolved into a cooperative that collectively invested in tea processing, developed stronger demand linkages through an auction, and participated in a women-led processing factory. The cooperative increased its income by 546% by selling processed tea rather than green leaves, and the Government of Tanzania committed to replicating the model through five additional processing factories. Women also gained stronger links to formal business registration systems and financial services, making future access to capital more feasible.

In Vietnam, CARE supported women coffee farmers’ efforts to organize into producer groups, leading to the formation of the Ara Tay Cooperative, which leverages regenerative agriculture, enterprise development, technical assistance, market intelligence and commercial partnerships to move producers into higher-value specialty coffee markets. As part of this work, we recruited an international coffee expert to provide onsite coaching in processing and roasting techniques, enabling the cooperative to meet specialty coffee standards and compete nationally. Rather than financing individual farmers in isolation, this aggregated approach led to the creation of a commercially viable cooperative enterprise capable of producing, marketing and selling higher-value products.

In both cases, CARE performed the aggregation, enterprise development and market linkage functions that helped turn fragmented producers into viable, investable enterprises. Yet despite the value it created for market actors, our work was subsidized by donor funding rather than financed as part of the investment process. These experiences support our broader conclusion: Building an investable asset class requires more than financial instruments. It requires deliberate investment in aggregation, enterprise readiness, standardized underwriting and origination. Once those market functions exist, guarantees, blended finance and commercial capital become considerably more effective because they are financing enterprises that have already been organized into a form the market can recognize.

Conclusion

The aggregated MSME asset class is not waiting to be invented. The borrowers exist, the demand is documented, and the capital is sitting on the sidelines waiting for a market to receive it. What remains is the work of defining the asset class, building standards around aggregated enterprises, and pricing the origination work that makes deployment possible.

That work is unglamorous, but it is the precondition for everything else. Until the field treats these enterprises as an asset class, the impact investing market will keep recycling through the same deals, and cooperatives, producer associations and savings-group-graduated enterprises will remain systemically overlooked.

Dr. Adanna Chukwuma is Associate Vice President for Economic Growth and Private Sector Engagement at CARE USA.

Photo credit: Sebastian Soto

Given the natural alignment between the region’s smallholder-driven supply base and its growing supermarket demand, one might expect smallholder farmers to be among the primary beneficiaries of formal food retail’s growth. Yet despite this growth, the formal retail market remains inaccessible to many smallholders, as a range of barriers prevent them from supplying these vendors. The agricultural production is present; the coordination is lacking.

As part of the FCDO’s Africa Food Trade and Resilience programme, which aims to increase intra-African food trade and improve smallholder income and resilience, AGRA and TechnoServe ran a “Kenyan High-Value Market Pilot” to test what is required in practice to link regional smallholder suppliers to formal retail markets.

We took a market-based approach, partnering first with Kenyan supermarkets and the aggregators that supply them to understand their procurement needs and the barriers that prevent them from sourcing from smallholders. Our pilot was then able to use this knowledge to support smallholders in reaching real sales opportunities, with the aggregators consolidating and coordinating supply from multiple producers to allow the supermarkets to source at scale.

Over four months, the program facilitated the sale of 63 metric tons of smallholder produce, with a projected annual volume of 586 metric tons, based on advanced orders and the trajectory of supply agreements. Through aggregators, smallholder producers are now supplying previously inaccessible supermarkets.

From our experience working closely with aggregators and supermarkets in this pilot program, we identified some key constraints that limit market linkages between smallholder farmers and formal retailers. Below, I’ll explore those constraints, and share some of the approaches we developed to address them.

The five missing links between farmers and formal retail markets

The program identified five key “missing links” that separate farmers from formal retailers: market information, quality, supply and logistics coordination, cross-border trade, and financing.

Market information: In our experience, buyers and suppliers require more than an introduction to begin working together. Concerns over non-payment, non-delivery and quality failings make new relationships risky. Moreover, supply, demand and price variation create challenges for medium- and long-term planning.

To facilitate new partnerships, we screened, vetted and matched buyers and suppliers, taught smallholder-linked suppliers how to present their products and negotiate prices with formal retailers, and facilitated coordination to match buyer demand with smallholder production.

As buyers and suppliers continue to interact and build trust through networking and trial and error, there is opportunity for private innovators, industry associations, development organizations and governments to develop digital or low-tech market linkage solutions (e.g., a market actor directory showing historical transactions and reviews from prior trading partners) to lower coordination barriers and facilitate further partnerships.

Quality: To meet the requirements of high-value retailers, smallholders needed to make immediate improvements to their produce quality, with more sophisticated refinement likely needed over time. In our pilot program, the changes needed to meet current formal retail requirements were relatively minimal, and we were able to facilitate them by training smallholders to improve their post-harvest handling practices and use food-grade crates to minimize produce bruising.

As formal retailers across the region continue to face food safety scrutiny, further quality criteria are likely to become differentiators, if not requirements, for suppliers. The program assessed smallholders against the Kenyan Good Agricultural Practice standard to identify farm-specific actions to improve food safety and traceability. For example, smallholders were trained to document pesticide application timing to ensure a sufficient gap between application and harvest.

As high-value retail produce quality expectations evolve, investment for the necessary training and equipment for smallholders may come from buyers looking to strengthen their supply bases, farm associations improving their members’ commercial viability, development partners, or some combination of the above.

Supply and logistics coordination: The price and availability of logistics services, particularly for cross-border trade, remain a key constraint. For many suppliers, inconsistent produce volumes make it difficult to secure regular, cost-effective transport, while fragmented supply increases inefficiencies across the value chain.

In our pilot program, what proved effective was coordinating suppliers to aggregate volumes and align delivery schedules to buyer demand. This planning allowed suppliers to fill truckloads, reduce per-unit transportation costs, and provide the predictable supply desired by formal retail customers. Aggregating supply also facilitated access to cold-chain logistics, to maintain quality and deliver over longer distances.

Private sector actors can profit from providing these shared logistics and cold-chain services on a fee-for-service basis. Improving access to these services would not only reduce costs and losses but also enable more suppliers to participate in regional trade.

Cross-border trade: We worked with aggregators in Tanzania to supply Kenyan markets — a process complicated by non-tariff barriers to trade. By organizing a stakeholder forum and a workshop at a one-stop border post (where regulatory officials maintain a presence), we allowed traders to sit down with regulators to understand cross-border trade requirements.

From working with these parties, we observed a clear desire for stronger regional government coordination. Suppliers and buyers alike want access to regional market opportunities, and expect their governments to see not only the fiscal benefits of customs revenue, but also the benefits of neighborly cooperation to food system resilience.

In these conversations, traders and regulators highlighted practical opportunities to improve the efficiency and predictability of border crossings, such as the implementation of the East African Community’s harmonized sanitary and phytosanitary regulations, and the consolidation of border interactions under a lead agency. Both parties agreed that regional horticultural shipment procedures should be adjusted to make it easier to consolidate consignments in one container, facilitating smallholder farmers’ access to regional markets.

Financing: Both buyers and suppliers sought financing to meet working capital needs and fund necessary investments. During the pilot, we introduced market actors to financial service providers and provided loan application support, but longer-term solutions are needed at greater scale.

Many smallholder-linked aggregators lack the formalization (e.g., record keeping, governance, succession planning) required to access funding. In response, lenders should partner with technical assistance providers to overcome these access barriers and build a pipeline of bankable businesses. Aggregators’ relationships with established market actors remain an underutilized avenue for unlocking finance; financiers can leverage these relationships to develop or more widely deploy tailored value chain financing products for agriculture (e.g., tripartite arrangements, supply chain financing). Solutions like technical assistance and value chain financing unlock banking services to businesses that would otherwise lack access, while reducing financier risk. The financial institutions that pursue them stand to grow while simultaneously improving portfolio risk.

The road to formal market inclusion for East African smallholders

Formal food retail is expanding in East Africa, with Kenya’s supermarket sector growing from $9.6 billion in 2022 to $12 billion in 2024, and modern outlets gaining traction across Uganda, Rwanda and other countries. At the same time, regional coordination is gaining momentum, with intra-East African Community (EAC) trade growing from $12.1 billion in 2023 to $19.3 billion in 2025, and the EAC reaffirming its commitment to regional integration earlier this year. Even as non-tariff barriers persist and implementation lags policy ambition, corridor upgrades (e.g., road capacity expansions, digitized border procedures) and harmonized regulations are creating a plausible backdrop for predictable cross-border sourcing. The opportunity to link smallholder farmers to the predictable revenue streams of high-value retailers has never been more accessible.

Smallholder suppliers have long been excluded from high-value markets due to constraints in quality, coordination, logistics, border processes and access to finance. Our pilot demonstrates that these are not structural limitations: They are system constraints that can be addressed.

As the program has shown, there is no longer any doubt about whether smallholder suppliers can serve formal markets. The question now is whether we can build the systems that allow them to do so consistently and at scale.

Daniel Njiwa is Director for Inclusive Markets, Trade and Finance at AGRA; Kris Ansin is the Country Director at TechnoServe for Kenya, Tanzania and Uganda; Audrey Tsoi is a Fellow with TechnoServe.

Photo credit: Dragos Condrea