The CPI to CPA ratio is the share of installs that pay, seen from the cost side. We don’t talk about it enough.
CPM, IPM, CPI, CPA, pricing and price tolerance all shape the ratio.
The US matters and Tier 1 holds money. The question is which countries give your app the best CPI to CPA ratio. Pricing power and market saturation skew it hard. Some markets are so saturated that when you optimize for purchase, CPIs are too expensive to ever sell profitably. You’re staring at 40 to 50x jumps from CPI to CPA. Meanwhile, another country might have softer CPIs and enough buying intent to make the math work.
I’ve seen an account profit in a Tier 2 country because CPIs were much cheaper and the app still converted. Even if conversion is a bit lower, when CPI is 7.5x cheaper, you’re still way ahead on CPA. The math is simple. A $2 CPI with 3% of installs paying beats a $15 CPI with 4%. That’s a $67 CPA versus $375.
It’s the same app in two markets. CPA is CPI divided by the share of installs paying, so the cheaper installs win even with the lower conversion rate.
But cheap CPIs don’t automatically mean healthy margins. This is where pricing power comes in, and pricing data like Adapty’s State of In-App Subscriptions 2025 shows the other half. Annual plan prices sit well below Western Europe and North America in most Tier 2 regions, and lower still beyond them. RevenueCat’s State of Subscription Apps 2026 shows the same pattern from the revenue side, with revenue per install and trial conversion both varying by region.
| Market | CPI | Pricing | What happens |
|---|---|---|---|
| Saturated | Too expensive to sell profitably | Cannot make up for it | 40 to 50x from CPI to CPA |
| Middle zone | Still soft enough for scale | Has not collapsed yet | CPI to CPA to LTV alignment holds, so this is the goldmine |
| Trap | Cheap | Has fallen as far as CPI, or further | Cheap CPIs do not turn into healthy margins |
My own three way split of the markets above, not a dataset.
How far CPI falls against how far price falls decides whether a market is a goldmine or a trap.
The play is finding the middle zone, and that is not always where the lowest CPIs are. Creative testing is a separate question. I still prove creative winners in the market I plan to scale in.
There’s no master dataset for that, but maybe that’s the point. The apps that win are the ones mapping these ratios themselves. Once a market is live, I check it every month with a country by country ROAS audit that catches markets that miss target.