Mobile UA Benchmarks 2026: CPI, ROAS, IPM & Retention by Vertical

Mobile user acquisition has rarely felt this complicated. Media costs keep climbing, privacy frameworks keep shifting, creative cycles keep shortening, and AI is rewriting how campaigns are optimized inside Meta, Google, TikTok, and Apple Ads. In that environment, knowing what "good" performance actually looks like is half the job. The other half is knowing how to read those numbers without making bad decisions.

This guide brings together the metrics that matter most for mobile UA in 2026: CPI, ROAS, IPM, and retention. We will look at how they behave across verticals, why they should never be read in isolation, and what UA teams can do to push them in the right direction.

Benchmarks are a planning tool first, a judgment tool second. If you are building a 2026 UA budget, forecasting payback, or pitching a growth target internally, you need a sense of where your numbers should land. Without that frame of reference, it is easy to either set ambitions too low (and underinvest) or chase performance that the market does not actually deliver.

That said, benchmarks are direction, not truth. The same app can hit very different CPI, ROAS, and retention numbers depending on geo, platform, monetization model, channel mix, and creative maturity. A casual game scaling in Tier 1 markets behaves nothing like the same title launching in LATAM. A fintech app monetizing through subscriptions behaves nothing like one running on transaction fees. Treat industry numbers as a compass, not a target.

CPI on its own is almost never enough. An install is just the entry point. What matters is what that user does next, how long they stay, and how much revenue they generate. That is why mobile UA in 2026 has to read CPI, ROAS, IPM, and retention together, as a single system.

CPI (Cost Per Install)

CPI is the most quoted metric in mobile UA, and the most misread. It tells you how much you paid to acquire one install. What it does not tell you is whether that install is worth anything. CPI is influenced by app category, country, platform (iOS still skews higher than Android in most markets), audience targeting, bidding strategy, and creative IPM. Two campaigns can land at the same CPI and produce wildly different outcomes downstream.

ROAS (Return On Ad Spend)

ROAS is the metric that links UA back to the business. It measures revenue generated per dollar spent, usually evaluated at fixed points: D1, D7, D30, D90, and sometimes D180. The right ROAS horizon depends on your monetization model. A hyper-casual game with IAA monetization needs to see ROAS recover fast. A subscription app with a long trial-to-paid cycle may not show meaningful ROAS until D60 or D90. Reading D7 ROAS for a subscription app and panicking is one of the most common mistakes in the space.

IPM (Installs Per Mille)

IPM measures installs generated per thousand impressions and reflects how strongly a creative resonates with its audience. It is the cleanest creative performance signal you have, because it isolates the ad from bid strategy and targeting. A high IPM creative will almost always lower CPI, because networks reward engagement. In 2026, with creative fatigue cycles compressing on Meta and TikTok, IPM has become a daily operational metric, not a quarterly review.

Retention (D1, D7, D30)

Retention measures the share of users still active after a fixed number of days. D1 reflects onboarding quality and immediate value. D7 reflects whether the product is forming a habit. D30 reflects long-term stickiness and is one of the strongest leading indicators of LTV. Retention also functions as a quality filter on traffic: if D1 looks fine but D7 collapses, the issue is rarely the ad. It is usually the audience match, the store listing, or the early product experience.

The ranges below are directional. They reflect the broad patterns reported by AppsFlyer, Adjust, Singular, Sensor Tower, and Liftoff in their 2026 publications. Real performance depends on geo, platform, monetization, and campaign setup.

Gaming remains the largest vertical by UA spend, with global investment estimated near twenty five billion dollars annually according to AppsFlyer and Newzoo. CPI varies enormously by sub-genre. Hyper-casual and casual titles still see the lowest CPI ranges in most regions, while mid-core and strategy games command higher install costs but typically deliver much stronger LTV. According to Adjust's Mobile App Trends 2026, gaming CPI rose meaningfully year over year in 2025, a trend that has continued into 2026. ROAS expectations follow the monetization model: IAA-heavy games need fast early ROAS, while IAP-heavy games can tolerate slower curves. Retention is genre-defined. Strong puzzle and casual titles often see D7 retention in the high teens to low twenties, while top mid-core games push higher. UA teams should focus on creative velocity and playable formats, which consistently outperform on IPM.

FinTech CPI sits in the higher half of the global benchmark, often two to four times gaming, driven by competitive auctions and a smaller qualified audience. Sessions in finance apps rose sharply in 2025 according to Adjust, which has improved engagement signals for advertisers. ROAS in fintech is typically measured against revenue events further down the funnel (verified account, first deposit, first transaction) rather than D7 revenue. Retention tends to be strong if onboarding is solid, since users who complete KYC and link an account are committed. UA teams should focus on funnel events, not pure install volume.

This category leans heavily on subscriptions. CPI is moderate to high depending on geo, and ROAS analysis must extend to at least D30 to reflect trial conversion. Seasonal demand spikes (January, post-holiday, late spring) reshape benchmarks materially. Retention is fragile in the first two weeks, then stabilizes once a habit forms. The vertical is one of the most creative-sensitive: lifestyle UGC and transformation narratives consistently outperform polished studio output.

E-commerce CPI sits in a relatively wide range because the category includes everything from value retailers to luxury. ROAS is measured against orders and lifetime customer value. D1 retention can look surprisingly low because many users install for a single purchase intent, but D30 cohorts that return tend to monetize well. Promo-led creatives, deal-of-the-day formats, and dynamic product ads drive most of the volume.

Subscription apps span fitness, mental health, dating, education, and a growing wave of GenAI-powered tools, which have been the fastest growing category on Android and one of the top growing categories on iOS according to Sensor Tower's State of Mobile 2026. CPI varies widely. The defining benchmark here is not D7 ROAS but trial-to-paid conversion and payback window. Best in class subscription apps target payback inside D90 to D180. Retention should be tracked separately for free users and paid users, with paid retention being the meaningful metric.

Travel CPI is highly seasonal and geo-dependent. ROAS often appears low in early days because the booking cycle is long, with users researching for weeks before converting. Retention is naturally lower than utility apps, but repeat install behavior is high. UA teams in travel should plan around event-driven campaigns and not expect flat performance month over month.

Education apps, especially language and skill-building, see moderate CPI but require careful ROAS modeling because most monetization is subscription-based. D1 retention is one of the best predictors of long-term success in this category. Creative testing matters a lot, and short before-and-after style UGC tends to outperform feature-led ads.

Utility CPI is often among the lowest in the market because intent is broad and creative messaging is simple. Retention is typically weaker, with many users installing for a one-time task. The best performing utility apps build retention through notifications, value reinforcement, and bundling with secondary use cases. ROAS depends heavily on whether the app is ad-supported, freemium, or subscription.

A cheap install that churns in twenty four hours is more expensive than a paid install that stays. Pure CPI optimization rewards the wrong behavior: it pushes campaigns toward broad, low-intent audiences and away from the qualified users that actually drive LTV. The right way to read CPI is alongside retention and ROAS by cohort.

Higher CPI is often justified when the audience is harder to reach but more likely to convert. Fintech, premium subscriptions, and mid-core gaming all live in this zone. The discipline is to balance CPI with quality signals and let unit economics, not vanity, decide budget allocation. Teams running structured user acquisition services typically build CPI guardrails by audience segment rather than at the campaign level, which protects against optimizing for the cheapest possible user.

D7 ROAS has become the default UA reporting metric, and for many app categories that is too short a window. Subscriptions, fintech, and mid-core games all monetize on longer cycles. Using D7 as a single judge of campaign quality systematically underestimates winning campaigns and overestimates aggressive IAA spend.

A more useful framework is to define a payback window per business model, then evaluate ROAS at multiple checkpoints inside that window. Cohort-level ROAS, not blended ROAS, is the standard in 2026. Blended numbers smooth out exactly the signal you need to act on. If your D7 ROAS looks weak but your D30 cohort projection points to payback inside the planned window, that is a campaign worth scaling, not killing.

IPM is the closest thing UA teams have to a direct creative quality score. A campaign with strong IPM gets more reach at lower CPI because networks like Meta and TikTok reward engagement. Liftoff's 2026 creative performance data showed playable ads generating IPM well above video and static formats, with video itself outperforming static. The takeaway is not that everyone should run playables, but that creative format choice has a measurable impact on acquisition cost.

What works in 2026 is volume plus iteration. The top mobile spenders are producing thousands of creative variants per quarter, and most concepts have a useful life of one to two weeks before fatigue sets in on Meta and TikTok. Building this kind of throughput is what separates apps that scale from those that plateau.

Dedicated creative production services exist precisely to keep this pipeline moving without sacrificing brand consistency, mixing video, UGC, statics, and playables based on what the data says is winning that week.

Strong retention is the cleanest evidence that your UA is bringing in the right users. Weak retention, especially a sharp drop between D1 and D7, usually points to one of three things: a mismatch between ad creative and product experience, a store listing that overpromises, or an onboarding flow that fails to deliver value fast enough.

Benchmarks vary widely. Top quartile casual games target D1 in the high thirties or low forties, while utility apps often live in the mid twenties. Subscription apps care more about trial-to-paid retention than raw D1. The rule is simple: compare yourself to your own vertical, not to the global average.

Aligning your store listing with your ads is one of the cheapest ways to lift retention. Users who install based on a clear, accurate expectation churn less. Strong ASO services work hand in hand with paid UA to make sure the message users see in the ad matches what they see on the listing and the first run experience. When ad messaging, store metadata, and onboarding all tell the same story, retention curves tighten visibly.

Benchmarks shift based on a predictable set of variables. Geography is the biggest single driver: Tier 1 markets cost more, convert harder, and retain better. iOS and Android remain structurally different, with iOS CPI typically higher but iOS LTV stronger in most monetized categories. Channel mix matters: TikTok, Meta, Google, Apple Ads, Mintegral, and Moloco all index differently by vertical. Creative quality and IPM compress CPI; weak creative inflates it.

App store conversion rate magnifies or wastes every dollar spent. Monetization model dictates which ROAS horizon is meaningful. Seasonality bends benchmarks across travel, retail, fitness, and education. Privacy and attribution coverage (SKAN, AdAttributionKit, ATT opt-in rates that sit near thirty eight percent globally per Adjust) determine what you can even measure.

And campaign maturity is often underrated: a campaign in week two and a campaign in week eight are not running on the same logic, even at the same spend.

Strong UA performance in 2026 comes from a few disciplined habits. Build campaigns around business goals (qualified users, subscription conversions, ROAS targets), not around install volume. Test creatives continuously and accept that most will fail; the goal is to find the winners faster. Segment by geo, audience, and funnel stage so optimization signals stay clean.

Treat ASO as part of UA, because store conversion rate compounds every paid click. Report on cohorts, not blended averages. Track retention and monetization side by side; one without the other is misleading.

Diversify channels: paid UA paired with influencer marketing often outperforms either channel alone, especially in lifestyle, gaming, and consumer subscription categories where trust signals matter more than reach. And use benchmarks to ask better questions, not to replace strategy.

Most app teams reach a point where in-house resources are not enough to scale efficiently. Common moments include moving past initial UA tests into structured growth, cutting waste from campaigns that are scaling but not improving, building a real creative testing system, lifting ROAS to hit a payback target, expanding into new markets with unfamiliar channel dynamics, diagnosing low quality installs or weak retention cohorts, and connecting UA, ASO, creatives, and analytics into one growth system instead of four disconnected workstreams.

This is where an experienced app growth partner adds value. Mobihunter works with mobile apps and games across gaming, fintech, e-commerce, subscription, and lifestyle categories, building performance-driven UA strategy that ties acquisition, creative, and store presence into one system. The goal is never installs for their own sake. It is unit economics that work.

Benchmarks are useful when they sharpen your judgment and dangerous when they replace it. CPI, ROAS, IPM, and retention each tell part of the story; the answer is always in how they interact. Winning mobile UA in 2026 will reward teams that combine smart strategy, fast creative testing, clean analytics, and disciplined execution.

If you are reviewing your 2026 UA performance and want a second opinion on whether your numbers reflect real growth or just spend, start by auditing them against your vertical and your business model. Identify the gaps that matter most, then build a plan that addresses them in sequence rather than all at once.

To see how Mobihunter helps mobile app and game teams scale acquisition with measurable performance, explore our user acquisition services or get in touch with the team about your current UA setup. A short conversation is often enough to surface the one or two changes that move the numbers most.