Serengeti Analytics Raises $18M to Put Crop-Yield Models in Farmers' Hands

Serengeti Analytics closed an $18 million Series A led by Rift Valley Ventures, with participation from two development finance institutions and the agricultural arm of a Tanzanian bank. The company has spent four years selling satellite-derived yield forecasts to processors and traders; the new money is meant to invert the customer.

The consumer product is deliberately unglamorous. A farmer registers a plot by walking its perimeter with a feature phone, and thereafter receives SMS advisories keyed to that specific field — planting windows, pest pressure warnings, and a yield estimate updated every ten days. Serengeti charges the equivalent of $1.40 a season and says 61,000 farmers signed up during a pilot in Tanzania and southern Kenya.

"Our enterprise customers pay us a great deal of money to know what the harvest will be," said chief executive Neema Lyimo. "The farmer is the person generating that harvest and the last one to find out. That is a strange market and we would like to fix it."

The economics are the hard part: a $1.40 product needs enormous scale to matter, and SMS delivery is not free. Serengeti's bet is that the enterprise contracts subsidize the consumer tier long enough for the consumer tier to become the more valuable data asset — a familiar strategy, applied to a market where nobody has yet made it work.