The India paywall experiment
Last updated: August 2026
The most surprising row in our analytics: India has the highest purchase intent of any market we serve - and nearly the worst conversion. This post is the funnel, the mistake that caused it, and what we shipped.
The funnel, raw
| Step | India (90d) | United States |
|---|---|---|
| Saw the paywall | 2,147 | 1,791 |
| Tapped buy | 1,487 (69%) | 910 (51%) |
| Completed at the store sheet | ~70 (4.7%) | ~300 (33%) |
Read that middle row again: 69% of Indian users who saw the paywall tapped the buy button - the highest intent anywhere, ahead of the US. Then 19 out of 20 abandoned at the payment sheet. Intent was never the problem. The sheet was.
The mistake: parity pricing dressed as localization
Our store listing had per-country prices in 61 currencies, which looked localized. Converted back to dollars, nearly every market paid within about 20% of the US price - and some poorer markets paid more (Bangladesh's annual worked out around $107 against the US $89.99). Currency conversion is not price localization. A $90 annual against Indian purchasing power is roughly a $700 ask in US terms - the sheet-abandonment was users being rational.
What we shipped
- Weekly-first paywall - the annual stopped being the pre-selected default worldwide.
- Store-honest copy - the paywall now advertises exactly what the payment sheet will do, trial or no trial.
- A cheap one-time unlock as the dismissal offer - own your rating outright without a subscription.
- Purchasing-power price work per market instead of converted parity.
The metric that decides if it worked
Not installs, not paywall views - tap-to-completion at the sheet. India already supplies the taps. If completion moves from 1-in-20 toward the US 1-in-3, the same traffic multiplies revenue with zero new users. That is the experiment; results will get their own post.
Lessons if you sell globally
- High-volume, low-revenue markets are not "bad markets" - check where in the funnel they die before writing them off.
- The payment sheet is the real paywall. Everything before it is advertising.
- Convert your prices back to local purchasing power once a quarter. Parity creep is silent.
See what all the fuss is about
ChadMe is the app these numbers come from - AI face rating with per-trait scores. First rating is free.
FAQ
Why not just remove the paywall in low-income markets?
Serving AI ratings costs real compute per scan - free-for-everyone loses money at volume. Purchasing-power pricing plus a cheap one-time unlock is the honest middle.
Is India worth serving at all for a paid app?
On intent, absolutely - it out-taps every market we have. The revenue question is purely whether the price and payment path match the market, which is on the developer, not the users.