MobileAppCPIBenchmarks2026:ByVertical,Geo,andCreativeType

Cost-per-install across mobile apps has been climbing. Adjust's 2026 Gaming App Insights Report puts global mobile gaming CPI up 30 percent year over year. iOS and Android averages have widened, not narrowed. The market is more efficient and more expensive at the same time.

Most benchmark reports stop at the headline numbers. iOS $2.24. Android $1.12. A few vertical breakdowns. That's useful for one Slack screenshot and not much else.

This guide goes further. It covers the 2026 averages across the major verticals and geographies, but it also explains the four levers that actually move CPI (the same ones most reports leave out) and how to read whether your number is good, bad, or fine. We've used our own data where we have proprietary insight (health and wellness, primarily) and cited industry sources from Liftoff, Adjust, AppsFlyer, RevenueCat, Adapty, and others where we don't. Methodology is called out throughout.

If you're setting a UA budget, evaluating campaign performance, or trying to figure out whether your CPI is actually a problem, this is the reference.

Key takeaways

• Global average CPI in 2026 sits at $2.24 on iOS and $1.12 on Android (Searchlab, citing Adjust and AppsFlyer 2026 data).
• Mobile gaming CPI rose 30 percent year over year in 2026 to a blended global average of $0.56 (Adjust 2026).
• UGC creative lifts impression-to-install rate by an average of 152 percent (Liftoff 2025 Mobile Ad Creative Index, across 4.7 trillion impressions).
• iOS CPIs run 2 to 5 times Android in most verticals, driven by higher LTV users in iOS-dominant markets.
• UK CPIs run 20 to 40 percent lower than US across most genres, making the UK a meaningful efficiency play for English-language apps.
• For health and wellness apps, our own data shows Tier 1 market CPI typically lands between £2 and £7, with Day-7 ROAS in the 100 to 120 percent range for accounts with strong creative coverage.

The 2026 headline numbers

Searchlab's 2026 app marketing statistics compilation, drawing on Adjust and AppsFlyer source data, places worldwide CPI at $2.24 on iOS and $1.12 on Android. Western European averages are higher: $3.40 on iOS, $1.85 on Android.

Channel-level splits matter. TikTok Ads sits at $2.45 CPI globally, currently the cheapest paid acquisition channel for most app verticals. Influencer marketing is the most expensive at $4.20. Meta Ads, the dominant channel for most consumer apps, sits between the two depending on vertical and creative quality.

These averages are useful for orientation. They're misleading as a benchmark for any specific app. A $4 CPI is excellent for a high-LTV subscription app in North America and a disaster for a hyper-casual game targeting Southeast Asia. The number on its own tells you almost nothing.

The four variables that determine where your real benchmark sits are vertical, geography, platform, and creative type. We'll work through each.

Why CPI varies so much: the four levers

Vertical. Different app categories command different acquisition costs because they monetise differently. Financial apps command the highest CPIs across the public benchmark sets (Mapendo's 2025 report flags this clearly) because the LTV justifies the cost. Hyper-casual games sit at the other extreme, with CPIs often under $1 because their monetisation through ad revenue caps how much publishers can rationally bid. Health and wellness, eCommerce, and productivity apps sit in the middle band.

Geography. The spread is enormous. North American CPI runs roughly 5x to 20x Latin American CPI for the same vertical and platform. Western Europe sits between the two. Liftoff's mid-core gaming data shows LATAM CPI at $0.27 against North America's $5.45 for the same genre. UK CPIs specifically run 20 to 40 percent below US, per Amps33's 2026 unit economics analysis, which makes UK-focused launches a meaningful efficiency advantage for English-language apps.

Platform. iOS users cost more than Android users in nearly every vertical, often 2 to 5 times more. The reason is downstream economics. iOS users live disproportionately in higher-GDP markets (US, Japan, Western Europe), they spend more on average per app, and they convert to paid subscriptions at higher rates. Publishers know this and bid accordingly. Casual games show one of the widest gaps: Liftoff's 2025 data measured casual game CPI at $1.41 on iOS versus $0.14 on Android.

Creative type. This is the lever nobody benchmarks properly. Liftoff's 2025 Mobile Ad Creative Index found that introducing UGC to a campaign lifts impression-to-install rate by an average of 152 percent. By format, native ads run cheapest at $1.80 average, video at $2.68, interstitial at $2.66. Playable ads convert 8 times non-playable for top games and 16 times for other game advertisers. The creative format and the creative quality compound: a strong UGC video on Meta will outperform a polished branded video by margins large enough to halve effective CPI.

The next sections take each vertical in turn and stack our data against the public benchmarks.

Health and wellness CPI: the deep dive

This is where we have the strongest proprietary data. The Social Outline works with health, fitness, sleep, pregnancy, and wellness apps across UK, US, and European markets. The ranges we typically see across our client portfolio:

• Tier 1 markets (UK, US, Canada, Australia): £2 to £7 CPI
• Tier 2 markets (Western and Northern Europe): £1 to £4 CPI
• Tier 3 markets (LATAM, Southeast Asia): £0.30 to £2 CPI
• Day-7 ROAS for subscription apps: 100 to 120 percent for accounts with strong creative coverage

These ranges represent campaigns running on Meta with creative built around the TSO Creative Framework (psychological coverage across Valence Zone, Self-Concept Anchor, and Language Intensity). They are not theoretical or top-decile. They are operating ranges.

For context against industry data:

RevenueCat's State of Subscription Apps 2025 places median Year 1 LTV for Health and Fitness apps at $27.21. Hard paywall apps in the category reach $49.30 median. Install LTV (revenue per install over 12 months) lands at $1.21, leading all subscription app categories.

Adapty's State of In-App Subscriptions 2026 confirms the category's strength: install-to-trial conversion in North America runs at 14.5 percent against a 11.2 percent global average. Hard paywalls convert at 5.5x the rate of freemium and produce 2x the LTV per subscriber, though with a 1.7x higher refund rate.

Health and wellness sustains higher CPIs than most categories because the downstream economics justify it. A subscription app with a median Y1 LTV of $27.21 can absorb a $5 CPI on Day-0 because the LTV-to-CPI ratio comfortably clears the 1.5x minimum threshold recommended by FoxData's 2026 UA cost analysis.

Two things drive CPI volatility in this category. First, seasonality. New Year peaks push CPI up across fitness apps for 6 to 8 weeks each January as audience demand spikes and inventory tightens. Second, personalisation positioning. Apps that lean on "personalised plans" or "AI-powered routines" in their messaging tend to attract higher-intent traffic at higher CPIs, where apps positioned on generic transformation messaging see lower CPIs with weaker downstream conversion.

The practical read: if you're running a Tier 1 health and wellness app on Meta with healthy Day-7 ROAS at £4-5 CPI, you're in the strong-performance band. If you're at £8+ CPI in the same conditions, you're likely sitting on a creative coverage problem (which our creative fatigue guide covers in depth).

Mobile gaming CPI by genre

Gaming is the most data-rich vertical because Liftoff, Singular, GameAnalytics, Sensor Tower, and Unity all publish detailed benchmarks. The categorisation differs slightly between sources, but the patterns are consistent.

Hyper-casual games: CPI typically under $1 in Tier 1 markets. Liftoff's data places casual games at $1.41 iOS and $0.14 Android. These games rely almost entirely on ad revenue monetisation, which caps what publishers can rationally bid.

Casual games: Median CPI sits at $1.41 on iOS and $0.14 on Android per Liftoff's 2025 Casual Gaming Apps Report (covering Feb 2024 to Feb 2025, analysing 1.4 trillion ad impressions and 2.5 billion installs). Day-7 ROAS averages 7.6 to 7.8 percent across both platforms.

Mid-core games: Average CPI runs around $2, roughly double casual. Liftoff's mid-core data via Udonis breaks down to $0.73 Android and $3.65 iOS as 12-month averages. Day-7 ROAS sits at 4.3 percent globally, with North American mid-core hitting 4.5 percent. Shooter genre commands the highest CPI in the mid-core bracket at $7.47 average, but also the strongest D7 ROAS at 6 percent. RPG genres are the cheapest at $0.60 CPI but with weaker D7 ROAS at 1.7 percent.

Hardcore and strategy games: CPI ranges from $5 to over $30 depending on title and market. iOS hardcore CPI ranges from $2 to $5+ as a baseline, per megadigital.ai's 2026 analysis.

Casino games: The extreme end. Liftoff's data places iOS casino CPI at $21.03, the highest single category in the public benchmark set. This is driven by LTV: casino game users have some of the highest median LTV in mobile gaming, and publishers bid aggressively to acquire them.

Geographic spread within gaming is dramatic. Latin America mid-core CPI sits at $0.27 against North America's $5.45 for the same genre, a 20x gap. The strategic implication: cheap CPI markets are not always profitable markets. The 1.5x LTV-to-CPI ratio threshold has to clear in either market for the campaign to make sense.

eCommerce and shopping app CPI

eCommerce app CPI sits above gaming and below finance in the public benchmark sets. Mapendo's 2025 report places retail and eCommerce among the higher-CPI categories on average, with significant elevation during sales periods and holiday seasons.

Photo and video apps (a related shopping-adjacent category) show some of the widest CPI spreads in the public data. RevenueCat's State of Subscription Apps 2025 found Photo and Video top-quartile iOS CPI above $14, driven by competitive concentration. Most eCommerce app CPI sits lower than this but follows similar volatility patterns.

We don't have proprietary eCommerce CPI data at the depth we have for health and wellness, so our guidance here leans on the public benchmarks. If you're running a mobile-first eCommerce app on Meta and your CPI is above $6 in Tier 1 markets outside Q4 sale periods, it's worth auditing creative coverage. Most eCommerce app fatigue we've audited externally lives in the same psychological zone (High Positive, Ideal Self, Direct Response) that dominates the rest of the consumer app space.

The hidden factor: CPI by creative type

This is the lever almost every benchmark report leaves out, and it's the one your team can directly control.

UGC adds 152 percent impression-to-install lift. Liftoff's 2025 Mobile Ad Creative Index, drawing on 4.7 trillion impressions, 263 billion clicks, and 1.1 billion installs from January 2023 through May 2025, measured the average impression-to-install lift from introducing UGC into a campaign at 152 percent. This is a campaign-level halving of effective CPI from a single creative format choice.

Playable ads convert 8x to 16x non-playable formats for games. Same Liftoff dataset. Playables convert at 8 times non-playable rate for top-spending game advertisers and 16 times for other game advertisers. The catch is that playables only run at meaningful scale on a narrow set of channels (AppLovin and ironSource/Unity primarily), and they cost $5,000 to $15,000 per concept. For most consumer apps spending the majority of their budget on Meta, playables are not the answer. Video and UGC are.

Format-level CPI averages from older Liftoff data:

• Native ads: $1.80 average
• Banners: $2.14 average
• Playable: $2.38 average
• Interstitial: $2.66 average
• Video: $2.68 average

Native is cheapest but the conversion rate disadvantage cancels the savings for most game and subscription apps. Video wins on conversion despite the higher per-install cost.

The strategic point we make in our creative fatigue guide: creative type and creative coverage compound. A library of UGC videos that all sit in the same psychological zone (the same valence, the same self-concept hook, the same intensity) will fatigue at the rate of a single ad, regardless of how many variants exist. UGC alone isn't enough. UGC across psychological zones is.

This is why we benchmark CPI not just by vertical and geography, but by creative coverage. A health and wellness app running 20 UGC variants all anchored to "transformation" messaging will see a different CPI curve than one running 8 variants spread across "transformation", "permission", "guilt relief", and "identity reframe" messaging. The second library compounds. The first decays.

CPI by country: the tier breakdown

The three-tier model is useful for planning. Not perfect, but it gives you a working frame.

Tier 1 (US, UK, Canada, Australia, parts of Western Europe): Premium CPI. Highest LTV. Most competition. Most agencies and brands concentrate spend here, which sustains the high cost.

Tier 2 (rest of Western Europe, Northern Europe, Japan, South Korea): Moderate CPI. Strong LTV. Less saturation in some sub-categories. Often the best efficiency play for English-language apps that can localise reasonably.

Tier 3 (LATAM, Southeast Asia, India, Eastern Europe): Low CPI. Lower LTV. Volume play. Useful for apps with monetisation models that work at low ARPU (hyper-casual gaming, ad-supported social, basic utility apps).

The UK is a particular efficiency story. UK CPIs run 20 to 40 percent below US across most genres (Amps33 2026 analysis). For UK-headquartered apps or US apps with strong UK ICP, the UK is a meaningfully cheaper launch market than the US for similar LTV profiles. We see this consistently across our portfolio.

The strategic call: do not chase Tier 3 CPI just because the numbers look attractive. A $0.30 CPI in Indonesia is only useful if your LTV in Indonesia clears $0.45. For most subscription apps, it won't. Match the geo to where LTV justifies the CPI, not to where CPI looks cheap on the dashboard.

What to do if your CPI is above benchmark

Three diagnostic questions, in order:

Is the audience right? If you're targeting broad lookalikes on Meta with no recent audience refresh, the algorithm is showing your ad to a wider, less-qualified pool. Tightening to lookalike-1% on recent purchasers or installers usually drops CPI meaningfully within 7 to 10 days. If you're already tight on audiences and CPI is still above benchmark, the audience isn't the issue. Move on.

Is the creative right? This is where most "high CPI" problems actually live. Audit your last 20 active ads against the TSO Creative Framework. If 70 percent or more sit in one psychological zone (most likely High Positive, Ideal Self, Direct Response), you have a coverage problem. The library is fatiguing as one unit even if individual ads look fresh. The fix is psychological rotation, not visual refresh. Our creative fatigue guide covers the full diagnostic.

Is the offer right? If multiple creative zones are failing on the same underlying offer, the offer is the issue. The market no longer finds your promise compelling. This is the most expensive problem to fix because it requires rebuilding the value proposition, not just the ads. But it's also the rarest. Most "high CPI" problems are creative, not offer.

There are also cases where CPI being above benchmark is fine:

• High-LTV subscription apps with downstream economics that justify it
• Brand awareness campaigns where the immediate ROAS is less important than reach
• Newly launched apps where the first 30 days of any campaign run higher CPI while Meta calibrates

The right framing: CPI is a means, not an end. The number that matters is ROAS or LTV-to-CPI ratio. If your LTV-to-CPI ratio sits above 1.5x and your unit economics work, your CPI is fine regardless of how it compares to public benchmarks.

If you want a quick estimate of how many creatives you need per month to sustain efficient CPI given your spend and vertical, we built a free creative refresh calculator that does the math.

Frequently asked questions

Want this run for you?

The benchmarks are public knowledge. The system for hitting them consistently is the work.

We run performance creative for mobile apps spending £25k or more per month on Meta. Our model is built around psychological coverage rather than volume refresh, which is why our client portfolio sits in the strong-performance band of these benchmarks rather than the middle of the distribution. If that sounds like the kind of system you want running on your account, apply to work with us. We take a small number of mobile app clients per quarter.