The ledger was not a moat
2026-07-06
Post-UPI, Mypoolin finally had what we’d wanted all along: peer-to-peer and group payments with real traction, on a rail that was free, universal, instant. Then the giants noticed the category. WhatsApp launched payments inside the app India already lived in. Google Pay launched with cashbacks — real money, spent at a scale that redrew DAU charts. We were a startup. We could not buy users back.
We believed we had a moat: a Splitwise-like ledger of who owed whom, the accumulated history of our users’ shared expenses. Surely the data would hold them while the cashbacks raged. It did not. A ledger is a record, not a lock; users took the subsidy and kept their friendships elsewhere.
The lesson arrived expensively: P2P payments alone is not defensible. The transfer is the commodity; differentiation has to live somewhere a competitor’s subsidy can’t reach.
In agentic payments this is about to replay, faster. Agent-to-agent transfer will be table stakes — x402 and its siblings make enabling it nearly free. If moving money is your whole product, you don’t have one.