We understand a lot about why India’s farmer producer organisations struggle. What’s missing is the applied evidence to act on it well: data tied to outcomes, and gathered where the records go quiet.
There are close to 50,000 of these farmer-owned companies. To see where they actually stand, I pulled the public records on all 49,826. A few numbers stuck with me:
→ Only about 1 in 3 has even ₹5 lakh to do business with.
→ 43% never collected the share capital their own members signed up to pay.
→ 95% still only sell raw produce. Almost none process or add value.
→ The 2020 batch, whose five years of government support is ending right now, is the youngest and least funded of the lot.
We’ve described this for years. Every report lands on the same shortlist: credit, markets, professional management, member participation. But a description isn’t yet evidence you can act on.
Two gaps are the ones that should be supported now.
First, we can see the symptoms but haven’t tied them to outcomes. We know which companies look weak today, but almost no one has connected the traits of the ones that have already failed to the ones still standing. Without that correlation, money gets sprayed across all 50,000 instead of aimed at the ones it can actually save.
Second, the evidence that matters most isn’t in any record. Whether members actually trade through their own company, whether there’s a real buyer for the crop: you can only learn it in the field. Almost no one is gathering it at scale. What little exists is rarely outcome-linked, and rarely in the public domain.
So the missing work isn’t more description. It’s applied evidence gathering: connecting what we measure to what actually drives survival, and going to the farmers for what the records can’t show.
I put the full picture into a short, scrollable story: where FPOs really stand, and what it would take to fix them rather than just describe them.
https://fpo-story.surge.sh
If you work with FPOs, self-help groups, lenders or rural livelihoods, I’d value your read, and your inputs.
PS: AI agents were used in collating and analysing this data.