Mobile-network partnerships with social sector organizations

What Digital Green’s FarmerChat partnerships with mobile operators shows about scaling social sector innovation.

Many social-sector interventions demonstrate promising results and could benefit far larger populations, particularly when delivered through phone-based products that can scale cost-effectively. But reaching end-users at scale requires more than one delivery or financing pathway. Partnerships between nonprofits, Mobile Network Operators (MNOs), and funders can offer one route to scale by bringing together complementary capabilities.

Learnings from SME-MNO engagement

Start-ups and small and medium-sized enterprises (SMEs) in low- and middle-income countries (LMICs) are building mutually beneficial partnerships with MNOs. Since 2015, GSMA’s Mobile for Development Foundation has funded 175 companies through the GSMA Innovation Fund, helping reduce the risks of innovation in LMICs. Drawing on its mobile industry network, GSMA has supported over 110 partnerships between grantees and MNOs. These partnerships give start-ups access to MNOs’ customer networks, distribution channels, and services such as mobile money. In return, MNOs gain new revenue opportunities, expand their customer base and offer distinctive services.

This has implications for the social sector too, for example, Vitara, a West African agritech company, had an established farmer network and digital tools but needed trusted payment infrastructure. MTN Ghana wanted to expand its MoMo mobile money service into rural and agricultural markets. The two integrated MoMo into Vitara’s platform and jointly developed financial literacy workshops, giving female smallholder farmers access to financial tools. For MTN Ghana, the partnership offers a more cost-effective way to reach rural customers.

What transfers to nonprofits

A similar partnership model can work in the social sector. Nonprofits bring the product and sector expertise, while mobile operators provide infrastructure, customer relationships, and distribution channels that can lower the cost and effort of reaching users. What differs is the commercial sustainability model. Philanthropic funding can support early product development, testing, and distribution. Such partnerships are especially useful early on, when interventions are still being tested and the risks of scaling remain high. Over time, sustainability may depend on two reinforcing sources of funding:

• government systems that adopt, finance, and integrate services with demonstrated social value
• end-user engagement that supports continued service use and creates commercial value

Rather than replacing one another, these actors can share the risk of taking an innovation to scale.

For mobile operators, sustained engagement can create commercial value through greater data use, device uptake, customer retention or, where appropriate, direct user payments through freemium or subscription models. These benefits can complement public financing, including integration into government programmes or budgets.

Ecosystem catalysts such as GSMA, can play a central role in helping social sector organizations build relationships with mobile operators. They bring longstanding expertise in developing mutually-beneficial value propositions and connections with the right decision-makers within MNOs. They can coach organisations on their MNO engagement strategy, from developing value propositions to understanding operator priorities.

The farmer’s experience is the test of whether the arrangement works: if the service is easy to access through a mobile operator they already know, can be used without worrying about data charges, and offers timely support when the farmer would otherwise have to wait for an extension worker.

Digital Green has pursued several forms of mobile-network partnership, organised around who pays and who distributes FarmerChat. In Zimbabwe, Econet offers FarmerChat through its EcoFarmer brand and leads customer acquisition. In Kenya and Ethiopia, Safaricom sponsors farmers’ data use, reducing a barrier to continued use.

For philanthropic funders, these partnerships also offer a potential sustainability model. This includes supporting proven nonprofits to reduce reliance on philanthropy while reaching more users and markets through institutions that can finance and distribute their services at scale.

Data are through 24 August 2026. Evidence is observational except where noted.

Digital Green and Safaricom, in Kenya and Ethiopia: removing data costs is associated with higher FarmerChat usage

FarmerChat remains Digital Green’s product, while Digital Green pays Safaricom for the data farmers use. Farmers can therefore access FarmerChat without the charge coming from their own balance. This reduces a barrier for farmers, especially women, while giving Safaricom a commercial role in delivering a useful service to rural customers without having to build the advisory product itself. Farmers on other networks continue to pay standard data rates.

There could be an additional benefit for mobile operators, as Sila Bitok, a farmer and private extension provider in Nandi County, points to an interesting use case:

“I had a phone that did not allow me to download FarmerChat so I changed it immediately… however, I had survived with it and did not see a need to change it until I needed FarmerChat.”

While this is anecdotal and further evidence is needed, it raises an important question: can useful sponsored-data services influence device upgrades, paid data use or customer retention?

Other farmers emphasised the value of having agricultural advice available when they need it. As Susan Wothaya, a farmer in Nyeri County, shared that “within a minute I have the help I need.”

Existing data show a clear association between sponsored (free) data and FarmerChat usage. In an analysis by The Agency Fund and Digital Green covering 814,000 farmers over twelve months, usage was higher among Safaricom users where mobile networks could be identified. They asked 27% more questions in Kenya and 19% more in Ethiopia than farmers on other networks, with the difference remaining relatively stable over time, and women and men benefiting equally.

A second signal comes from Ethiopia. During a four-week temporary disruption to sponsored data, image and voice queries among Safaricom farmers fell by around 30%, while text queries fell by 13%. No comparable decline was observed among farmers on other networks.

All query types declined during the disruption, but voice and image queries fell much more sharply than text. For farmers accessing FarmerChat’s more data-intensive features, this suggests that cost is an important constraint: when sponsored data was disrupted, use fell most sharply for the features that consume more data. The pattern is consistent with sponsored data increasing FarmerChat use, although the observational evidence cannot establish causality. Planned randomized experiments should provide a clearer estimate.

Where mobile data is a material expense for lower-income farmers, removing that cost can make FarmerChat easier to access and keep the service available when a farmer has run out of airtime or an active data bundle. A recent survey of 161 lapsed FarmerChat users reinforces the importance of these access barriers: 57% cited phone or network constraints as a reason for stopping use, while 15% specifically cited high data costs. Field testing in Ethiopia also confirmed that Safaricom users with a zero balance could still open FarmerChat and receive a response without drawing on their balance.

Gender inclusion was also part of the case for sponsored data. The increase in use was similar for women and men, but this does not necessarily mean the value of removing data costs is the same. Women farmers often have less access to mobile data and less control over spending on airtime, so removing the cost may address a greater barrier for them. A follow-up report by Digital Green from August 2026 reinforces that data costs are a bigger barrier for women: 25% of women who stopped using FarmerChat cited data costs, compared with 8% of men. Sponsored data could therefore make FarmerChat more accessible to women even if the observed increase in usage is similar across genders.

For Peris Wachira, a farmer in Kenya’s Nyeri County, the benefit was straightforward:

“I do not have to worry about buying data bundles to access it.”

One limitation of the current arrangement is that farmers were not systematically told that their FarmerChat data was sponsored. Many users may therefore have been unaware that they could use the app without incurring data charges. Digital Green and Safaricom need to align on communications to determine which organization will inform farmers about the benefit, through which channels, and how prominently. Word of mouth, however, plays a crucial role.

Sila, the Kenyan farmer and private extension provider, added that farmers often ask him:

“Now without data, how will I be entering FarmerChat?”

He further shared that farmers’ biggest surprise was learning that “it is free and they can use it anytime.”

The service is free to the farmer, but the data is not free to the operator. Digital Green pays the sponsored-data cost.

Econet, Zimbabwe: distribution as reach

Another successful model of MNO partnership is white-labelling. Econet offers a version of FarmerChat under its own EcoFarmer brand, using its existing customer relationships and distribution channels to reach farmers. The underlying advisory engine remains the same, but the service is branded and distributed by Econet.

The arrangement creates mutual value. Through Econet’s distribution network, more farmers can access FarmerChat’s advisory service, while Econet adds an AI-enabled agricultural product for its rural customers. Digital Green and Econet have a commercial revenue-sharing arrangement: farmers are charged a small fee to access EcoFarmer and Econet shares part of these revenues with Digital Green to cover platform management costs.

From January through 24 August 2026, the pilot reached approximately 14,100 farmers who asked about 55,900 questions, providing the clearest case in which actual user behavior (rather than anticipated uptake) can be studied. Registrations rose sharply in February, with more than 2,000 farmers joining in that month alone, and rose again by about 2,700 from April through May. The available usage data show these surges but do not identify what drove them.

The model also creates a dependency on Econet's distribution efforts. While Digital Green can reach farmers faster through the operator's existing channels, further acquisition is likely to depend on how actively Econet markets and integrates the service. White-labelling may therefore work best when the product fits an operator's existing commercial offering.

What farmers do once they begin using the service is a separate consideration. Image-based diagnosis, where farmers photograph affected crops, accounted for around 28% of queries but, because image queries are more expensive to serve than text or voice, they accounted for 71% of observed serving costs. Usage limits therefore played an important role in keeping the service economically sustainable.

The broader lesson extends beyond this case, the features that generate the greatest value for users may also carry the highest delivery costs, so partnerships designed for scale need to account for both from the outset.

The distinction between reach and retention

Taken together, the two models show the clear potential of MNO partnerships to expand reach and reduce barriers to using critical social services like FarmerChat. Econet shows what can happen when an operator takes an active role in distribution, while the Safaricom experience suggests that removing data costs may support higher usage. Upcoming sponsored data randomized experiments should provide a clearer estimate of how much lower access costs affect farmer engagement.

Reach, however, is only part of the problem. In the Econet pilot, 1,863 farmers asked at least one question, but only 15% returned the following week. By week four, fewer than 4% were still active. Similar drop-off can exist outside the MNO partnerships, suggesting that sustaining that retention issues are not unique to the Econet model. Mobile-network partnerships can help get farmers through the door, while keeping them engaged remains the harder problem.

Designing MNO-Nonprofit partnerships for success

These cases show that successful partnerships require a clear value proposition for both users and partners, and that alignment on distribution, user communication, data access, and commercial responsibilities shapes how the product performs in practice. The cases also highlight differences between the two partnership models:

• Sponsored data requires careful implementation to ensure farmers can use the service end-to-end without unexpectedly reaching features that consume their own data.
• White-labelling places greater weight on branding, customer ownership, commercial terms, and the operator’s role in distribution.

For white-label partnerships in particular, aligning early on product roadmaps, shared objectives, and product ownership can help prevent the partnership from drifting into bespoke development.

They also illustrate the value of complementary roles across the partnership ecosystem for sustained delivery at scale. Digital Green brings the product, sector expertise, and understanding of user needs. GSMA contributes mobile-industry expertise and relationships with operators. The Agency Fund supports partnership development, evidence generation, and coordination across actors.

These models nevertheless involve important trade-offs. Sponsored data is limited to users on participating networks, for example, while white-labelling can reduce a nonprofit’s visibility into the end-user relationship. The right arrangement will also vary by country and operator, depending on the regulatory environment and commercial priorities. Regardless, such partnerships can create mutual value: operators can extend social services to their customers, while nonprofits gain access to established distribution channels and infrastructure. Where incentives are well aligned and all partners remain engaged, mobile-network partnerships can provide a credible pathway to sustainable delivery at scale.