CPI is up 47% in Europe: the case for distribution beyond paid UA

If you run growth at a mobile studio, you already feel it in the dashboard: every install costs more than it did a year ago, and a smaller share of those installs is still around at the end of the month. The numbers behind that feeling are now hard to argue with. Blended gaming cost-per-install (CPI) sits around $0.56 in 2026, up roughly 30% year-on-year — and in Europe specifically, the jump is 47%. North America isn’t far behind at +31%. Meanwhile more than 95% of installs churn within 30 days.

That combination — paying more to acquire users who leave faster — is the quiet emergency in mobile growth right now. The playbook most studios still run, “spend more to grow,” was built for a world where the marginal install was cheap and the auction was efficient. That world is gone. This piece is about what replaces it: not a clever bidding hack, but a structural shift toward distribution channels you don’t have to rent by the click.

The math that used to work, and why it stopped

The paid-UA growth model has one load-bearing assumption: that the lifetime value of an acquired user comfortably exceeds the cost to acquire them, with enough margin to fund the next cohort. For most of the 2010s that held, because CPIs were low, auctions were thin, and you could lean on Google Play and the App Store to do the discovery work for free on top of your paid spend.

Three things broke that assumption at once.

Acquisition got more expensive. A ~30% blended CPI increase year-on-year isn’t a blip — it’s the cumulative effect of signal loss after ATT, more advertisers chasing the same finite attention, and platform auctions that now extract more of the surplus. In Europe the 47% figure reflects both fierce competition and tighter privacy enforcement that makes targeting less efficient.

Retention got worse. When 95%+ of installs are gone within 30 days, the denominator of your LTV math collapses. You’re not buying users; you’re buying a brief, expensive look. The few who stay have to subsidise everyone who didn’t — and at $0.56 a head, that subsidy is heavy.

The free discovery layer thinned out. The two dominant stores are more crowded and more pay-to-play than ever. Organic store discovery, the thing that quietly made paid UA economics work, is no longer a reliable tailwind. You’re increasingly paying for installs the store used to hand you.

Put those together and the equation inverts: for a growing number of titles, the next paid install is unprofitable on a fully-loaded basis. You can optimise creatives and bidding at the margins, but you cannot optimise your way out of a structurally broken unit economic.

”Spend more to grow” is now “spend more to stay flat”

Here’s the trap a lot of studios are in without naming it. Their growth plan is still denominated in UA budget. When the board asks how to grow next quarter, the answer is a bigger number in the same channels. But because CPI is rising faster than the budget, that bigger number buys roughly the same volume — or less. The budget grows; the audience doesn’t. You are now running faster to stand still, and calling it a growth strategy.

The honest reframe is that paid UA has quietly become a maintenance cost, not a growth engine — for many titles, anyway. It still has a role. Performance marketing is excellent at scaling something that already works, putting fuel on a fire that’s already lit. What it is no longer good at is being the only way you reach an audience. A studio whose entire reach depends on the two stores’ ad auctions has a single point of failure, and that point is getting more expensive every quarter.

Distribution is the lever paid UA can’t reach

The way out is not a better auction. It’s a different question. Instead of “how do I pay less per install in the same channels,” ask “what channels let me reach players without paying per click at all, or at a structurally lower cost?” That’s a distribution question, not a marketing one — and it’s where the leverage has moved.

There are four alternative channels that, layered together, change a studio’s reach economics. None of them is a silver bullet on its own. The point is the layering: a portfolio of channels that don’t all share the same cost curve as Google and Apple’s ad auctions.

Alternative app stores (post-DMA)

The DMA in the EU and similar moves elsewhere have forced open the iOS and Android distribution monopoly. Third-party marketplaces and sideloading are now legal realities, not gray-market hacks. The economics matter: many alt stores run a more favourable revenue share than the standard 30%, and several court games specifically with featuring you’d never get on a crowded first-party store. We covered the full landscape in our state of alternative app stores in 2026 — the short version is that there’s now a legitimate distribution surface outside the duopoly, and it doesn’t price discovery as a per-click auction.

OEM channels

When a Samsung, Xiaomi, or Huawei device ships, it doesn’t arrive empty. OEM storefronts, pre-install slots, and carrier-bundled placements put your title on the device before the user ever opens an ad. For studios that fit, this is some of the lowest-CPI distribution available, because the install isn’t priced through a real-time auction at all — it’s a negotiated channel relationship. We argued the indie case in detail in OEM partnerships are a quietly great channel for indies; the same logic scales up. If you’re deciding which OEM store to integrate first — Galaxy Store, Xiaomi GetApps, or Huawei AppGallery — regional reach and onboarding friction differ substantially; we compared the three head to head.

Carrier billing and emerging-market reach

In much of Asia, MENA, and LATAM, the binding constraint on monetisation isn’t acquisition cost — it’s payment friction. Direct carrier billing (DCB) lets users pay through their phone bill, with no card required. This unlocks paying audiences that the card-based stores simply can’t convert, and it reaches them in markets where CPI is a fraction of Western levels. Reach and conversion in one move. We covered why DCB out-converts cards and how to integrate it in a separate publisher’s guide, and mapped where it pays off across India, Africa and LATAM region by region.

Web-native distribution

For a growing set of titles, “distribution” no longer requires an app at all. WebGPU and WASM have closed enough of the performance gap that real games run in the browser, inside messaging platforms, and on web portals — with no store, no install friction, and no 30% cut. We unpacked this shift in web-native games are quietly reshaping mobile distribution. As a reach layer that bypasses the auction entirely, it’s one of the most underrated moves a studio can make in 2026.

What channel diversification actually does to the numbers

The reason to layer channels isn’t ideological. It’s that each one attacks a different part of the broken equation.

• Lower effective acquisition cost. OEM preloads, alt-store featuring, and organic web reach add installs that don’t carry a $0.56 auction price. Blended across the portfolio, your true cost-per-acquired-user drops even if your paid CPI doesn’t.
• Better-fit audiences, better retention. A user who finds you through a curated games-only store, or who was preloaded in a market where your genre over-indexes, tends to retain better than one swiped in from a generic ad auction. That directly attacks the 95% churn problem.
• Higher take-home per payer. A 12-20% alt-store or OEM revenue share versus a 30% standard cut is a permanent margin uplift on every transaction — and carrier billing converts payers the card rails would have lost entirely.
• Resilience. When your reach isn’t a single auction, a CPI spike in one channel doesn’t threaten the whole business. You stop being a price-taker.

None of this means switching paid UA off. It means demoting it from “the growth strategy” to “one channel in a stack,” and building the rest of the stack deliberately.

A pragmatic sequence for getting there

You don’t rebuild your distribution overnight, and you shouldn’t. A realistic path for a studio feeling the CPI squeeze:

1. Quantify the leak. Calculate your fully-loaded CPI and 30-day retention by channel and by market. Find the cohorts where the next paid install is already unprofitable. That’s your urgency, in numbers your board will recognise.
2. Pick one alternative channel that fits the title. Ad-monetised mid-core or hyper-casual on mid-range Android? Start with an OEM channel. IAP-driven and aimed at emerging markets? Lead with carrier billing. A title that runs well in a browser? Web-native. Match the channel to the game, not to the hype.
3. Soft-launch and measure against your paid benchmark. Run the alternative channel as a real cohort. Compare CPI, retention, and payer rate directly against your paid-UA numbers. Let the data, not the narrative, decide where to scale.
4. Layer, don’t replace. Add a second alternative channel once the first is operating. The goal is a portfolio where no single channel — paid UA included — is more than a manageable share of total reach.

The studios doing this well treat distribution as an operating discipline, not a one-off project. That’s deliberately how we built the IMH distribution practice and the studio-side engagement model: channel selection, integration, and ongoing channel ownership, so a growth team isn’t trying to learn five new partner ecosystems while also shipping the game.

The bottom line

A blended CPI of $0.56, up 30% year-on-year and 47% in Europe, against 95% 30-day churn, is not a temporary cost-of-doing-business. It’s a signal that the buy-growth model has run out of road for a large slice of the market. The studios that thrive over the next few years won’t be the ones who found a cleverer way to bid. They’ll be the ones who stopped renting all their reach by the click and built distribution they actually own.

If you want to map which alternative channels fit your specific titles, and what the integration and economics look like, that’s exactly the conversation our Founding Developer Program is built for. The CPI curve isn’t going to bend back. The distribution stack is the part you can still control.

FAQ

Why has mobile game CPI risen so much in 2026?

Blended gaming CPI is around $0.56 in 2026, up roughly 30% year-on-year, with Europe up 47% and North America up 31%. The drivers are signal loss after privacy changes like ATT, more advertisers competing for the same finite attention, platform auctions extracting more surplus, and a thinner free-discovery layer on the dominant stores. The increases are structural, not a temporary spike.

Should studios stop running paid user acquisition?

No. Paid UA remains excellent at scaling something that already works. The change is that it should be one channel in a stack rather than the entire growth strategy. With more than 95% of installs churning within 30 days and rising CPI, depending on store ad auctions alone is a single point of failure that gets more expensive every quarter.

What are the main alternatives to paid UA for mobile games?

Four channels, layered: alternative app stores opened up by the DMA and similar regulation, OEM channels (storefronts, pre-installs, carrier bundles), direct carrier billing for emerging-market reach and conversion, and web-native distribution that needs no store or install. Each attacks a different part of the broken unit economics; the leverage comes from combining them.

How does channel diversification improve unit economics?

It lowers effective acquisition cost by adding installs that don’t carry an auction price, improves retention through better-fit audiences, raises take-home per payer via lower revenue shares (often 12-20% versus 30%) and carrier-billing conversion, and adds resilience so a CPI spike in one channel doesn’t threaten the whole business.

Which alternative distribution channel should a studio try first?

Match the channel to the title. Ad-monetised mid-core or hyper-casual on mid-range Android usually fits an OEM channel first. IAP-driven titles aimed at emerging markets should lead with carrier billing. A title that runs well in a browser is a web-native candidate. Soft-launch it as a real cohort and compare CPI, retention, and payer rate against your existing paid-UA benchmark before scaling.