Apple App Store
Fewer downloads, far more money: a smaller audience worth more per head.Market Size: The Downloads-vs-Revenue Inversion
Apple's App Store generated $52.5 billion in gaming revenue in 2025, up just 0.6% year-over-year, according to Sensor Tower's State of Gaming 2026 report. That figure is gaming-only; the same storefront is cited anywhere from $42B to $117B depending on whether the measure is gaming-only, all app categories, or a forward projection. Within the gaming-only scope, it topped Google Play ($30 billion) and Steam ($11.7 billion) combined, per Sensor Tower's own comparison, despite logging only 7.8 billion downloads (down 5.7%) versus Google Play's 42.4 billion. Likely Third-party estimate
The downloads-revenue inversion
Total downloads in 2025, measured in billions.
Total gaming revenue in 2025, measured in billions of dollars.
The two panels use different scales because downloads and revenue are measured in different units, so what matters is which bar is taller on each side. Google Play generates roughly five times the downloads of the App Store but earns a little over half as much gaming revenue. All figures are sourced to Sensor Tower's State of Gaming 2026 report.
The average iOS gamer spent roughly $63 on apps in 2025 versus roughly $13 for the average Android user: a near 5x gap in per-user value that explains why Apple's smaller download base still outearns Android's much larger one.
Commission Structure
Apple charges 30% standard, dropping to 15% for developers enrolled in the Small Business Program (under $1M annual revenue) or on a subscriber's revenue after their first year. There is no revenue-scale tiering above the Small Business threshold the way Steam tiers by lifetime revenue; a developer crossing $1M moves immediately to the full 30% rate.
Google Play
The volume platform: most of the world's installs, a fraction of the money.Market Size
Google Play generated approximately $30 billion in gaming revenue in 2025 per Sensor Tower, on 42.4 billion downloads: accounting for 81% of all mobile game downloads globally against the App Store's 15%. Free-to-play titles represented 96% of total mobile downloads, rising to nearly 99% on Google Play specifically. Likely Third-party estimate
Commission Structure
Google charges 30% standard, 15% on the first $1 million in annual revenue per developer, then 30% above that threshold. Subscription revenue is charged at 15% from the first day regardless of the threshold; structurally similar to Apple's model but framed as a universal first-million discount rather than a small-business enrollment program.
Mobile Revenue Concentration
Mobile's winner-take-most problem is at least as extreme as Steam's: arguably worse.Sensor Tower's State of Mobile 2026 data found that 92.5% of in-app-purchase revenue and 79.8% of downloads were attributed to games from the top 1% of publishers in 2025. Eight individual games crossed $1 billion in IAP revenue in 2025 alone, a new record, led by Honor of Kings (~$1.7–2.4B) and Last War: Survival (~$1.6–2.2B). Likely · Third-party estimate
This concentration compounds at the payer level, not just the publisher level. Sensor Tower's Day 90 cohort data shows the top 5% of paying users generate 48% of total mobile game revenue, and in the United States specifically, just 0.02% of global installs (a narrow band of high-spending iOS users) account for roughly 20% of total global mobile gaming revenue. Likely Third-party estimate
Concentration runs two layers deep
The first bar shows which games win, measured by publisher; the second bar shows which players pay, measured within the games that won. A studio can reach the top 1% of publishers by revenue and still find that its entire monetization model depends on the small sliver of its own paying players shown in coral on the lower bar. Both figures are sourced to Sensor Tower's State of Mobile 2026 report.
Top three titles' share of category spend, by genre
Genre sets the revenue ceiling before the choice of platform does. The same studio, launching the same game on the same storefront, faces a structurally different revenue distribution depending on which genre category it competes in. This breakdown is sourced to Sensor Tower's State of Mobile 2026 genre analysis.
Discoverability: Editorial vs. Algorithmic
The same F2P economics, two different paths to visibility: very different UA costs.Both mobile storefronts share the core F2P monetization logic described in Part I (sale-after-engagement, LTV-driven UA math, Day-1/Day-7 retention as the critical checkpoint) but differ in both discovery mechanics and raw acquisition cost. The App Store retains meaningful editorial curation (featured placements, "Today" tab stories, category charts) alongside algorithmic ranking; Google Play leans more heavily on algorithmic ranking, user reviews, and retention signals.
That editorial-vs-algorithmic split shows up directly in acquisition cost. Per Liftoff's 2025 benchmarks, average cost-per-install for casual games runs approximately $1.41 on iOS versus $0.14 on Android: a roughly 10x platform differential. Casino and gambling-adjacent apps on iOS reach as high as $21 per install. Likely Third-party estimate
App Store vs. Google Play: side by side
| Dimension | Apple App Store | Google Play |
|---|---|---|
| Discovery driver | Editorial curation + charts | Algorithmic ranking + reviews |
| 2025 gaming revenue | $52.5B | ~$30B |
| 2025 downloads | 7.8B | 42.4B |
| Avg. casual CPI | ~$1.41 | ~$0.14 |
| Avg. annual user spend | $63 | $13 |
| Commission | 30% / 15% (SBP) | 30% / 15% (first $1M) |
Apple App Store
Google PlayThe Decision Matrix
Match the economy to the game, not the game to the cheapest commission.Bringing the full suite together. The choice between premium and mobile F2P, among premium storefronts, or between iOS and Android emphasis should follow from three things: the game's actual monetization design, its genre-specific concentration risk, and its realistic revenue ceiling. Not from whichever storefront's headline commission rate looks most attractive in isolation.
| Studio profile | Likely fit | Why |
|---|---|---|
| Small team, sub-$1M lifetime projection, sale-at-download design, and an owned demand channel driving its own traffic | Epic Games Store | 0% commission under $1M makes the platform work as checkout rather than discovery; without an owned channel the smaller audience is pure lost reach |
| Established premium IP, $10M+ projection, needs maximum reach | Steam | Audience scale outweighs the 30% headline rate once tiering kicks in above $10M |
| F2P design, strong early retention hooks, volume-first strategy | Google Play | 10x lower CPI and 81% of mobile download volume rewards a scale-first funnel |
| F2P design, premium production values, targeting whale-tier monetization | App Store | 5x higher average user spend and editorial features favor a smaller, higher-ARPU pitch |
| Console-native design, subscription tie-in potential | Xbox (Game Pass) | Game Pass's ~$5B revenue base offers a negotiated-terms alternative to pure unit sales |
Diagnostic Questions
Before greenlighting a platform strategy, five things for a whiteboard.- Every revenue figure in your model: which firm produced it, what scope does it measure, and does that firm count free-to-play, in-game spending, and back catalogue? App Store 2025 revenue alone has been cited anywhere from $42B to $117B depending on source and scope. Treat any uncited figure as unverified.
- Are you benchmarking against platform totals, or against new releases at your size? Only one of those describes your likely outcome.
- What genre are you building in, and what does that genre's specific concentration look like? A geolocation game and a strategy game face structurally different odds even on the identical platform.
- At your realistic revenue ceiling, does Epic's first-million exemption or Apple/Google's small-developer rate outweigh a larger platform's audience? For a game projecting under $1 million lifetime, that is a live calculation, not a rhetorical one.
- Does your UA budget account for the ~10x iOS-vs-Android CPI gap, or is it using a blended average that will misprice one platform or the other?