How does lifetime value build over time per acquisition cohort?
№ 071 · LTV curve by cohort
Definition
MetricCumulative realized revenue per install by month since acquisition
Unit$ per install, cumulative
Lines plot cumulative ARPU per cohort over days since install. Multiple cohort lines compared. Newer cohorts should sit higher. Build by tracking per-user cumulative revenue grouped by install cohort. Example: Jan 25 cohort $4.20 LTV at D90; Apr 25 cohort $5.80 at D90 (improving).
- LTV
- Lifetime value. Average gross profit per customer per period divided by that period churn rate, or summed over expected life.
- ARPU
- Average revenue per user. Revenue in a period divided by active users in that period.
Benchmarks
| Bottom 30% | Median | Top 30% |
|---|---|---|
| under $0.30 at 12 months, with the curve flat from roughly day 60 | $0.31 per install by day 60; ~$0.84 per install at 12 months | $1.20+ per install at 12 months (category leaders; North America runs ~2x the global curve); hard-paywall apps reach $3.09 per install by day 60 |
Subscription-led consumer apps, cumulative gross revenue per install, global mix, 2025 transaction data (RevenueCat, Adapty). Per-subscriber 12-month LTV is a different denominator and runs $45-47 in the top categories; never plot the two on one chart. · Install LTV vs subscriber LTV confusion is the main definitional trap on this card. Numbers are gross of store fees; net curves sit 15-30% lower. Ad-monetized consumer apps are underrepresented in these datasets, so ad-led LTV curves need internal baselines.
| Category | Bottom 30% | Median | Top 30% |
|---|---|---|---|
| Health & Fitness subscription apps | under $0.60 | $1.21 per install at 12 months | $2.40+ per install at 12 months (NA skew) |
| AI-powered apps | at or below the all-app median, common where the AI feature is bolted on | roughly 1.7-2x the all-app median at comparable windows | $1.44+ per install at 12 months; $0.63+ by day 60 |
When it looks bad
Cohort curves flatten early and newer cohorts flatten lower than older ones, so the fan of lines compresses toward the x-axis over time.
The January cohort reached $0.90 per install by month 12 while the June cohort flattened at $0.45 by month 4 and added under $0.05 in the following five months.
What to do about it
- Attach trials to annual plans for engaged users; trials on annual plans lift one-year LTV by roughly 35% vs direct purchase (Adapty 2026).
- Run structured paywall experiments on plan mix, trial length and placement rather than copy and color; the configuration gap is worth up to 636% on LTV (Adapty 2026).
- Attack the first-renewal leak with pre-renewal value messaging and a downgrade offer; first-renewal retention averages ~59% across categories, the single largest step-down in the curve (Adapty 2026).
- Rebuild UA bidding on predicted LTV from day-7 behavior instead of install volume, so the fan of cohort curves stops compressing as spend scales.
Sources
- RevenueCat Median revenue per install $0.31 by day 60; AI apps average $0.63+ revenuecat.com ↗
- RevenueCat Hard-paywall apps generate 8x higher revenue per install at day 60 than freemium ($3.09 vs $0.38) revenuecat.com ↗
- Adapty Average 12-month install LTV $0.84; AI apps $1.44; trials on annual plans lift one-year LTV ~35% vs direct purchase adapty.io ↗
- Adapty Health & Fitness leads install LTV at $1.21 globally; North America install LTV runs roughly 2x the global average across most categories adapty.io ↗
- Adapty 12-month subscriber LTV: Productivity $46.97, Utilities $46.30, Education $45.10 (per-subscriber, not per-install) adapty.io ↗
- Adapty Gap between best and worst paywall configurations is 636% on LTV; systematic testers see up to 40x revenue vs non-testers adapty.io ↗
- AppsFlyer Revenue accumulation shape: casual iOS hybrid models reach 55% of 90-day revenue by day 7; IAP-only apps reach 66% by day 30 appsflyer.com ↗