Every serious conversation about food in Kenya starts at the farm gate. It usually ends there too.
In June, the Kenya National Bureau of Statistics put annual inflation at 6.4 per cent. Food and non-alcoholic beverages came in at 8.6 per cent. Food is once again rising faster than almost everything else Kenyans buy.
The standard response is production. Grow more. Store better. Build cold chains. That response is correct as far as it goes. Research published by WRI Africa in September 2025 found that Kenya loses up to 40 per cent of the food it produces each year, worth roughly KSh 72 billion. Halving that by 2030 could feed seven million people and return KSh 36 billion to the economy. Those are serious numbers and they deserve the attention they get.
But look at where that research stops. It follows maize, potato, fruit and fish through harvesting, drying, storage, processing and wholesale. It does not follow the tomato into a kitchen in Kilimani. The last mile of Kenya’s food system is a restaurant, a hotel kitchen, a school canteen, a hospital caterer, a butchery, a roadside kibanda. We have almost no national figures for what is lost there.
That gap matters more than its size suggests, because the kitchen is where food is at its most expensive. A kilo of kale lost in the field costs a farmer the farm-gate price. The same kilo lost behind a restaurant in Westlands has already absorbed transport, refrigeration, handling, labour, rent and tax. The loss is small measured in tonnes and large measured in shillings. In a food business where ingredients typically run 30 to 35 per cent of revenue, a few unmeasured points of waste is the difference between a business that survives the year and one that does not.
Plenty of them do not. The last full KNBS survey of micro, small and medium enterprises found that 2.2 million micro businesses shut in the five years to 2016, and that 46.3 per cent of them closed inside their first year. Ask an owner why and you will usually hear about the economy or about taxes. Both are real. But sit with the books of a kitchen that failed and you often find something duller. Nobody knew what anything cost. Purchases were never matched against sales. Theft, spoilage and over-portioning were invisible because nothing was counted or analysed.
This is where an unrelated piece of policy becomes interesting.
From 1 January 2026, the Kenya Revenue Authority validates the expenses declared in annual returns against eTIMS records. Costs that are not backed by a compliant electronic invoice are disallowed and taxed as income. Under the Tax Procedures (Electronic Tax Invoice) Regulations of 2024, this reaches every business with turnover above KSh 5 million, every VAT-registered business regardless of turnover, and every withholding agent.
Most commentary has treated this as a burden, and for a small kitchen it genuinely is one. But notice what it produces. For the first time, a very large number of Kenyan businesses will hold a dated, line-item, digital record of what they bought, from whom, and at what price. That is the raw material of cost control. Kenya is building the data layer almost by accident.
The problem is that the record only flows one way. It goes to KRA. It does not come back to the operator in any form they can use. A chef who can see that beef moved from Sh620 a kilo in March to Sh780 in June, while portion sizes stayed the same, is a chef who can do something about it. Today that information sits in a filing obligation rather than a management tool.
Three changes would close the gap.
First, extend the national food loss baseline past wholesale. Kenya already has a post-harvest loss reduction strategy. Retail and food service should be inside the measurement, not outside it. We cannot reduce what we have never counted.
Second, pair enforcement with capability. eTIMS onboarding is being run as a revenue exercise. Fund the other half of it. Basic training in stock control, portioning and costing, delivered through county business units and industry associations, would cost a fraction of the collection effort. The state is already requiring these businesses to digitise. It should help them get some productivity out of the exercise, not only tax.
Third, publish benchmarks. Operators cannot manage against numbers they have never seen. Simple public medians for food cost percentage and waste rate, broken down by segment, would tell an owner in Kisumu whether 41 per cent is a crisis or ordinary. Right now every operator is guessing alone.
None of this is glamorous. It will not attract climate finance the way a cold store does. But Kenyan kitchens are throwing away food that has already carried the cost of transport, packaging, refrigeration and wages. It is the most expensive food we lose, and the only part of the food system we do not measure.
