Suvita

This is about impact and evidence. We don’t know if the model has impact yet. Suvita’s model builds on a compelling 2019 J-PAL RCT in Haryana showing a 22-percentage-point increase in full vaccination (from ~40% to 62%) when trusted influencers were engaged. That study found no SMS effect, though global evidence suggests SMS reminders can drive modest gains; Suvita’s internal modeling estimates an additional ~2-point lift. Their implementation differs from the original trial , and they don’t yet have evidence from their own delivery yet. Suvita is now testing if they can reliably identify true influencers—people who spread information measurably more than random individuals.

This is about scope. India has cut under-vaccination rates by more than half over the past decade, yet ~4 million of the country’s 22 million children each year still miss their full set of recommended vaccines. Coverage varies widely by state, and in every state, there are hard-to-reach groups—such as migrant workers, marginalized communities, and tribal populations—who remain underserved and are concentrated in specific districts in each state. Some of these communities migrate in and out of states more frequently, making them harder to target consistently. Overall though, Suvita’s model is well suited to reach them. With high phone penetration, digital outreach can scale efficiently. But success depends on accurate phone numbers linked to households, which is harder in states with weak public health infrastructure and limited digitization.

This is about whether government can deliver the model. Direct SMS to parents has been easy to deploy and scale in Maharashtra, where the health system is digitized, but likely harder in Bihar, where systems remain manual. The national government, with UNDP, has launched U-WIN—a nationwide website and app to digitize vaccination records, schedule appointments, and send reminders. Suvita is exploring how to support U-WIN’s rollout in Bihar. Meanwhile, they are still in R&D with the influencer aspect of the model—testing how to recruit, whom to recruit, and which households to target. They are currently testing approach in one district per state and do not plan to scale until they’ve landed on an approach they feel confident works.

‍

This is about what the model costs if delivered by government and whether government will pay. Suvita’s estimates for direct SMS delivery are $1.01 per child enrolled in Bihar and $0.25 in Maharashtra. Costs are higher in Bihar because health records are not reliably digitized, requiring significant staff time. The influencer model is currently much more expensive ($2.17), though it’s early and costs should fall as the approach is refined. Suvita does not yet know what a realistic target cost for governments to pay would be.