13 Mobile Game KPIs That Actually Matter in 2026

Most studios track thirty metrics and make decisions on three. The thirteen below are the ones that actually gate a decision: CPI, organic uplift and tutorial completion; D1, D7 and D30 retention plus DAU/MAU; ARPDAU, payer conversion, ARPPU and eCPM; and LTV against ROAS payback. Everything else is diagnostic detail you reach for once one of these thirteen moves.

Here they are, grouped by the question each one answers, with 2026 benchmarks.

Acquisition: Can You Buy Players Profitably?

1. CPI — Cost Per Install

What one install costs you in paid media.

2026 benchmarks: around $4.22 on iOS and $2.97 on Android on average (Game Growth Advisor). iOS runs 40–60% higher consistently, because those users monetise better. Efficient looks like under $0.50 for hyper-casual and under $3.00 for mid-core in Tier 1 markets.

How to read it: CPI on its own is not a performance metric, it's a price tag. A $6 CPI is excellent if LTV is $20 and catastrophic if LTV is $4. Studios that optimise CPI in isolation reliably buy cheap users who never pay.

2. Organic Uplift

Organic installs divided by paid installs. How much free distribution your paid spend generates through store ranking, word of mouth and creator content.

How to read it: this is the metric that separates campaigns that build a game from campaigns that rent users. A UA channel with a mediocre CPI but a 2x organic uplift is often cheaper in real terms than a channel with a great CPI and no halo at all. Creator-led campaigns tend to over-index here, because a video keeps working after the media spend stops — one reason we measure influencer UA on blended cost rather than channel CPI.

3. Tutorial Completion Rate

Share of new installs that finish the first-time user experience.

How to read it: the cheapest retention fix in games. If this sits below ~70%, no amount of UA optimisation will help — you're paying to fill a bucket with a hole in the first ninety seconds. Fix onboarding before you touch the media plan.

Engagement: Do They Come Back?

4–6. D1, D7 and D30 Retention

The share of players who return one, seven and thirty days after install. The single most predictive family of metrics in mobile games.

2026 industry medians: roughly D1 22–26%, D7 4–10%, D30 under 4%, depending on whose sample you take — the wider figures come from GameAnalytics' 2026 benchmarks, drawn from 16,000+ live games across nine regions and both platforms. D1 has fallen from about 27% a year earlier (Adjust Gaming App Insights, via industry reporting).

By genre the spread is much wider than the average suggests:

Source: Mistplay Mobile Game Retention Benchmarks, as reported by Segwise.

How to read the three together: D1 measures your onboarding. D7 measures whether the core loop holds. D30 measures whether you have a live game or a launch. A strong D1 with a weak D7 is a content problem, not a marketing one.

7. DAU/MAU — Stickiness

Daily actives divided by monthly actives. Effectively, how many days a month the average player shows up.

How to read it: 0.20 means six days a month. Session-based casual games live around 0.15–0.25; games built on daily habits — social casino, match-3 with daily events — push higher. Watch the direction, not the absolute: a falling DAU/MAU while installs hold steady means your live-ops calendar has stopped working.

Monetisation: Does the Economy Work?

8. ARPDAU — Average Revenue Per Daily Active User

Total daily revenue divided by daily actives. The heartbeat metric of a free-to-play economy.

2026 benchmarks: a health floor of above $0.15 for ad-monetised and above $0.30 for IAP-led games. Strong performers target far higher — roughly $0.80 casual, $1.50+ mid-core, $2.00+ hardcore (Juego Studio).

How to read it: never in isolation. ARPDAU rising while DAU falls usually means you've squeezed your whales harder, not built a better economy. Multiply it by DAU before you celebrate.

9. Conversion to Payer

Share of active players who ever spend money.

2026 benchmarks: above 2% is strong, top free-to-play titles reach 3–5%, social casino can exceed 8%.

How to read it: most games sit between 1% and 3%, so the lever is usually not "convert more people" but "deepen the ones who convert". Which is why the next metric matters as much.

10. ARPPU — Average Revenue Per Paying User

Revenue divided by paying users only.

How to read it: ARPPU and payer conversion move in opposite directions when you change price points, so watch them as a pair. A small paying cohort spending deeply routinely outperforms a large one spending shallowly. If ARPPU is climbing while conversion falls, you've made the game more expensive, not more compelling.

11. eCPM — Effective Cost Per Mille

What advertisers pay you per thousand ad impressions.

2026 benchmarks: rewarded video $15–25, roughly 3x interstitials; banners $1–3, contributing about 10% of ad revenue; the overall range runs $2 to $50+ depending on format and geography.

How to read it: eCPM is mostly a function of your geo mix and ad format, not your skill. Before optimising placements, check whether the drop is just a shift in where your installs came from.

Unit Economics: Should You Spend More?

12. LTV — Lifetime Value

Total revenue one acquired user generates over their lifetime.

How to read it: LTV must clear CPI by at least 1.5x to cover platform fees, overhead and attribution gaps. Model it at D7, D30 and D90 — a D7 LTV curve is what lets you make buying decisions this week instead of next quarter. And rebuild the model every quarter; a curve fitted to last year's cohorts will quietly lie to you.

13. ROAS Payback Window

How many days it takes to earn back what you spent acquiring a cohort.

How to read it: this is the metric your CFO actually cares about, because it's a cash-flow number, not a profitability one. A game with 200% D180 ROAS and a 150-day payback can still run you out of money. Most studios target payback inside 90 days; hyper-casual needs it far sooner.

The Mistake Behind Most KPI Dashboards

The industry shifted in 2026 from chasing the lowest CPI to balancing cost against lifetime value. That shift matters more than any single benchmark on this page, because it changes which number you optimise when two disagree.

Three habits separate teams that use KPIs from teams that merely collect them:

Segment before you conclude. A blended D1 across geos, channels and platforms is an average of things that behave nothing alike. The insight is always one level down.

Compare against your own cohorts first. Published medians tell you roughly where the field is. Your genre-matched cohort from two quarters ago tells you whether you're improving.

Measure incrementality, not attribution. Especially for creator campaigns, where the halo lands in organics and last-click quietly credits the wrong channel. It's the standard we hold ourselves to across our case studies and performance work.

Thirteen metrics, four questions: can you buy players, do they come back, does the economy work, should you spend more. Answer those honestly and the other twenty numbers on your dashboard become optional.