Field-service software,learned in the field

The roadmap questions could not be answered from Lagos. I travelled across 29 states to sit with the farmers and field technicians using the products, and the field kept correcting my assumptions. The biggest correction: many farmers didn't own smartphones. Not couldn't use. Didn't own.

So the product met them where they were. We introduced SMS intake: text what you need and when you need it to a number, and it lands in our database for dispatch. We introduced micro-payment plans so a farmer could buy a smartphone in instalments, and for those who couldn't, we supplied the phone and recovered the cost gradually from the harvest revenue we helped generate.

And because our users wrote to us in Pidgin, Yoruba, and Igbo, we translated. Manually, in-house, long before AI made that cheap. Localization was people.

AgriSquare, the greenfield community platform for agricultural enthusiasts, grew from 3,200 to 25,000 users in under six months against a twelve-month mandate. There was no growth hack. I went where the enthusiasts already were: the farms I was visiting anyway, local-government farmer databases we could reach by SMS, university agriculture departments, agricultural conferences. Small additions, steadily compounding.

Meathub was the beautiful app that didn't work. We built the software well and never solved cold-chain logistics, so the launch failed on delivery, not on design. Farmcrowdy itself eventually folded on operational grounds, even where the digital products were figured out. I keep that lesson close: software is usually the part you can control, and rarely the part that kills you.

Farmers, technicians, students, and professionals shared one platform and did not share a language, literally or otherwise. The platform grew large; the cross-group interactions never matched its size. Multi-persona products live or die on whether the personas actually need each other. I have carried that test into every multi-persona product since.