(Illustration by Harry Campbell)
This article is from SSIR’s forthcoming Fall 2026 print issue. Subscribe today for full access online and to receive your own print copy.
* * *
The maternal mortality ratio in Bangladesh stands at 123 per 100,000 live births. The country has roughly 6.7 physicians per 10,000 people, among the lowest ratios in South Asia. A national survey published in the Journal of Global Health found that 38 percent of maternal deaths occur on the day of delivery, with hemorrhage and eclampsia accounting for over half of those. These are preventable conditions that kill women because the right information and care do not reach them in time.
The standard development response is supply-side: Build clinics, train birth attendants, wait a decade. But the deeper problem is demand-side. Across South Asia and sub-Saharan Africa, families do not trust the health system enough to use it consistently. A first-time mother in rural Mymensingh, Bangladesh, is more likely to follow her mother-in-law’s advice than a government health worker’s. No amount of clinic construction fixes that.
The dominant response from the social innovation sector has been to import digital health tools built for Western markets and localize them. This approach assumes that a platform designed in San Francisco can be adapted elsewhere through translation and interface tweaks. A World Economic Forum analysis of MedicineAfrica, a nonprofit that connects UK physicians with medical students and clinicians in countries like Somaliland, found that participants described much of the transferred knowledge as inapplicable because it relied on equipment and medicines they could not access locally. The curriculum assumed a clinical environment that did not exist in their hospitals, and the practical effect was a platform that talked past the people it was built to serve.
What’s more, a 2023 Conflict and Health scoping review—a systematic survey of existing research—found that in digital health interventions for displaced populations, local participation was largely limited to surveys and focus groups within predetermined project designs. In both cases, the communities these tools were meant to serve had almost no role in shaping them. The working assumption that these tools should work is faulty: The infrastructure is different, the trust dynamics are different, the literacy patterns are different, the payment systems are different. What is needed is not adaptation but ground-up design by people who understand these realities because they live them.
Imported platforms don’t fail in emerging markets because of bad translation. They fail because they were never designed for these places to begin with. The ones that work are built from the ground up by founders who confront the same constraints they’re designing for.
Building From Within
ToguMogu, the parenting and family-health platform that I cofounded, launched in Bangladesh in 2017 to help the families that international platforms consistently fail to reach. While global tools like BabyCenter serve English-speaking, smartphone-savvy urban users (roughly 10 percent of Bangladesh’s population), ToguMogu was built for the other 90 percent: women with basic phones, limited data plans, varying literacy levels, and a deep preference for guidance in Bangla that reflects local medical practices and cultural norms. As a Future Startup profile noted, ToguMogu recognized a fundamental market gap: the absence of a comprehensive, localized parenting-support ecosystem.
The design choices were shaped entirely by constraint. Content was developed with local doctors and traditional birth attendants, not imported from Western medical databases. It was delivered through the communication channels that families already used, primarily Facebook Messenger and WhatsApp, because asking rural Bangladeshi women to download a standalone health app created an adoption barrier that no marketing budget could overcome. AI-driven personalization matched guidance to each woman’s pregnancy stage and local context. The interface worked on 2G networks. Families did not want a platform that looked like a hospital. They wanted a trusted voice that felt like a knowledgeable neighbor.
We got things wrong early on. Our initial assumptions about smartphone penetration and literacy levels forced significant redesign. We had built for the user we imagined, not the user who existed. The women we were trying to reach often shared a single phone with their household and had limited time to browse. The version that finally worked was stripped down, delivered content in short Bangla-language segments, and treated the platform not as an app but as an extension of the community-health conversations these women were already having. Designing for constraint turned out to be the innovation, not the obstacle.
The platform now serves over 200,000 registered families with week-by-week pregnancy tracking, doctor consultations, expert-verified baby products, and a community of parents. It generates continuous behavioral data: session patterns, content engagement by pregnancy stage, purchasing decisions, and return visits. This is not reach. This is sustained engagement with a population that most digital health interventions struggle to find at all.
In 2022, the United Nations Population Fund (UNFPA) was looking for locally rooted digital platforms that could strengthen sexual and reproductive-health services in countries where conventional delivery models were falling short. Rather than building its own tool, the agency ran its Equalizer Accelerator Fund jointly with the World Intellectual Property Organization, the International Telecommunication Union, and the International Trade Centre, drawing 300 applications from 61 countries. ToguMogu was one of 10 projects selected. What followed was not a grant-and-forget arrangement. UNFPA Bangladesh embedded the platform into its reproductive-health programming, using it to reach women who were not showing up at clinics. “It’s not business-as-usual for UNFPA to engage and partner with private companies,” said Kristine Blokhus, UNFPA’s then-representative in Bangladesh, at the November 2023 partnership signing.
A parallel track developed around Bangladesh’s garment sector, where four million workers, most of them women, have limited access to reproductive-health services. BRAC, the H&M Foundation, and The Asia Foundation funded ToguMogu through their STITCH for RMG Global Innovation Challenge to adapt the platform for factory workers’ pregnancy health and childcare. That work laid the foundation for something larger. In July 2025, Bangladesh’s Directorate General of Family Planning launched a digital monitoring tool built with technical support from ToguMogu and international reproductive-health nonprofit Jhpiego, in collaboration with UNFPA, to deliver real-time contraceptive supply tracking across garment factories. The tool reached 22,000 women in its first year. A local platform originally built for parents had become infrastructure for national health delivery.
The Way Forward
Based on my experience with ToguMogu and my knowledge of the field, I would offer three recommendations for social innovators working on digital health in emerging markets.
1. Partner with what already works. | The instinct among many international organizations is to develop a digital health tool at headquarters and deploy it across country offices. This approach consistently underperforms. UNFPA took a different path: Instead of building a parallel system, it partnered with an existing local platform and embedded it into national reproductive- health programming. Bangladesh’s government did the same when it needed digital infrastructure for garment-factory health services. Even Grameen Danone Foods, the social business established to combat child malnutrition in Bangladesh, chose to partner with a local digital platform to reach parents rather than building its own. The pattern is clear: Organizations that treat local platforms as partners rather than grant recipients get faster adoption, deeper community trust, and systems that outlast the project cycle.
2. Measure retention, not downloads. | The social sector’s obsession with reach metrics misses what matters. Downloads and registrations tell you almost nothing about whether a platform is working. A platform with 200,000 families who return week after week is worth more than one with two million downloads and a 5 percent monthly active rate. These sustained-engagement platforms generate continuous real-world data on behavioral change: which content mothers engage with at each pregnancy stage, whether they return for postnatal guidance, how session patterns shift over time. Funders who measure success by reach alone are optimizing for the wrong outcome.
3. Support human-AI collaboration, not automation. | In low-resource settings, AI that replaces human workers fails. The model that works is augmentation. ToguMogu’s platform flags high-risk-pregnancy indicators and surfaces them to local health providers. It does not replace the community-health worker. It makes her more effective by ensuring that warning signs are caught early and that women who might otherwise fall through the cracks are connected to care.
Consider what becomes possible in 10 years if these shifts are adopted. A network of locally built digital health platforms, each designed for its own context, could cover maternal and child health across South Asia and sub-Saharan Africa—not through one global product, but through dozens of community-rooted ones that share knowledge and interoperate. These platforms would not just serve individual families but function as the digital infrastructure through which governments and international organizations deliver public health programming. The continuous behavioral data they generate would reshape how the sector measures impact, moving from periodic evaluations to real-time population-level evidence. And the artificial walls dividing health, education, and economic support would begin to dissolve, as integrated platforms serve the whole family. A pregnant garment worker receives prenatal guidance through the same system that later delivers early-childhood content for her child, rather than slicing their needs into separate funding categories. None of this requires new technology. It requires a shift in where the sector places its trust.
The next generation of impactful digital health platforms will not come from Silicon Valley scaling down. They will come from Dhaka, Nairobi, and Lima, built by entrepreneurs who understand their communities because they live in them. The social innovation sector can continue importing solutions that look impressive in a grant proposal but fail in the field. Or it can start backing the people who are already doing the work.
Read more stories by Md Zillul Karim.