From Harvest to Table: Why Better Data Is the Missing Ingredient in Reducing Food Loss

By Alan de Brauw and Carolyne Maina

New research tracing tomatoes, mangoes, and dark leafy greens through Nairobi's wholesale markets reveals where losses happen — and who is best placed to fix them.

Food loss and waste is one of the most consequential—and most poorly measured—challenges in global food systems. TechnoServe and the International Food Policy Research Institute (IFPRI) came together through the NutriSave program with a shared ambition: to move beyond assumptions and generate rigorous evidence on where, how much, and at whose hands fruits and vegetables are lost along the supply chain. Most importantly, we wanted to find which actors are best placed to reduce those losses.

Over the years, the research community has accumulated a reasonable body of evidence on losses in grains and legumes. But when it comes to perishables—the tomatoes, mangoes, and dark leafy greens that form the nutritional backbone of diets across sub-Saharan Africa—the actual data on losses are sparse, the supply chains are complex, and the business opportunities to address losses remain largely invisible.

Our approach: Following food both ways through the market

We started in Nairobi’s wholesale markets. Rather than simply observing what happened at the wholesale level, however, we traced products in both directions: backward to the traders supplying those markets, and forward to the mama mboga vendors who buy from wholesalers and sell directly to households. This two-directional tracing is, to our knowledge, unusual in food loss research—and it matters, because the most effective place to intervene depends on understanding the entire chain.

We also carefully structured our sample within larger markets, categorizing vendors by size—small, medium, and large—to understand whether scale shapes losses in predictable ways. And we made one further methodological choice that may seem insignificant but proved critical: we observed markets for a full week, not just a single day.

This matters, because if an entrepreneur is trying to build a business around purchasing surplus food at day’s end, they need to know not just how much is typically left over, but how much that quantity varies. A highly variable surplus is a very different commercial proposition from a predictable one.

Preliminary findings: Larger vendors carry the most risk—and represent the biggest opportunity

Our preliminary findings point in a consistent direction. Losses are concentrated among larger vendors — those with higher volumes are more likely to end the day with unsold produce, about double the losses among smaller vendors (17% versus 9% among tomatoes, for example). Losses typically occur when products are stored, either during market hours or overnight, rather than during transport.

This finding is encouraging, since helping a smaller number of actors find a solution for high losses reduces the cost of creating change at a reasonable scale. Smaller vendors and traders selling into the wholesale markets show considerably lower loss rates. The mama mbogas—often assumed to be operating at the margins—turn out to actually be quite efficient: they calibrate their stock closely to what they expect to sell.

Across crop types, tomatoes and mangoes show slightly higher losses than dark leafy greens, which aligns with what we might expect given their moisture content and fragility. But all three crops are perishables with short windows, and the patterns we are seeing suggest that the challenge is not primarily about the food itself: it is instead about the mismatch between supply and demand at scale.

Implications: Connecting surplus to solutions

The picture emerging from this data points to two broad categories of intervention, both of which require a clearer cost-benefit case than currently exists. The first is linkage: connecting larger vendors, who are most likely to face end-of-day surpluses of tomatoes or mangoes, to businesses with other uses for that produce.

This would redirect nutritious food that would otherwise be lost into nutrition programs, processing, or circular economy applications, creating value for vendors and communities alike. At present, it is not clear whether a consistent, daily supply of product might be available at the end of the day; this research project will help shed light on that point.

The second broad category is cold storage. If produce that would otherwise be lost on a hot day or overnight could be held safely until the following morning, losses at the large-vendor level could fall substantially. But cold storage is capital-intensive, and making the economics work requires understanding the frequency and magnitude of overnight losses. The fact that losses are higher among larger vendors with more capital is promising in this case; we are continuing to build the cost-benefit estimates needed to make either intervention viable at scale.

By Carolyne Maina, TechnoServe, and Alan de Brauw, Senior Research Fellow at the International Food Policy Research Institute (IFPRI).