The wrong thesis
Six months ago, we looked at Splitwise — global app, no UPI, growing wave of frustrated Indian users — and thought we had a clear opening. Add UPI integration. Brand it for India. Done.
Within the first month of building, we shipped UPI deep-link settle: tap a balance, GPay/PhonePe/Paytm opens with the amount pre-filled. It worked technically. Users tried it once. They didn't come back.
Why UPI alone wasn't enough
UPI is now infrastructure, not a feature. Every Indian fintech app has it or will within a year. PhonePe, Google Pay, Paytm, CRED, Splitkaro — they all have it. Building UPI integration is a week of engineering. It can be copied in a week.
Worse, UPI is a feature users only touch at the very end of an interaction — the settle moment. If everything before that moment doesn't fit how Indians share money, the UPI step doesn't save you. They'll go back to WhatsApp threads and Excel sheets.
What the actual moat looks like
The cultural context around money in India is the real wedge:
- Group sizes are bigger. Indian college trip groups regularly hit 8–12 people. Western apps optimize for 2–4. UI paradigms break at 10+ people.
- Splits are unequal by default. Vegetarians and non-veg, drinkers and teetotalers, who joined which activity. Equal-split-only apps fail constantly.
- "Adjust kar lungi" is a real dynamic. Friends in India often absorb small differences to keep the peace. Apps that force precision feel awkward.
- Recurring rent has a cycle. Indian flats have a rhythm — rent on the 1st, electricity end of month, water mid-month. The app should know this rhythm.
- Hinglish is the actual language. "Chhodo yaar, main adjust kar lungi" isn't a translation; it's a feeling. Apps that speak this voice connect harder.
The reframe
Once we stopped treating UPI as the moat and started treating it as table stakes, the question changed from "how do we add features?" to "how do we fit the dynamic?"
Three things worked:
- Friends don't sign up. The dynamic of "I'll send you a screenshot" vs "please install this app first" is huge. Removing it changed completion rates.
- Offline-first. Indian users use apps in metro tunnels, on Goa beaches, on Manali roads. "Works always" is more valuable than any feature.
- Three features done well. Equal/custom split, settle-up via UPI, group balance view. That's it. The simplicity itself is the moat. (See our post on overbuilding.)
The lesson for other indie founders
If you're building Indian fintech-lite and your wedge is "we have UPI", you don't have a wedge. UPI is the assumed baseline. The wedge is whatever cultural friction your incumbent doesn't see because they're a Western app. Find that. Build for that.
For us, it was Indian rent + UPI flatmates + offline. Different products will have different wedges. But none of them are payment rail integration alone.