How to Calculate LTV for a Subscription App (2026)
A working LTV calculation method for subscription apps — three estimators (simple, cohort-based, decay-curve) and when each one is honest.
LTV — Lifetime Value — drives every important acquisition decision. How much can you bid for an install? How big a discount can you offer? How long can you afford a free trial?
Get LTV wrong by 50% and you'll either underspend (and grow slowly) or overspend (and run out of cash). Most indie devs get it wrong because they use the simple formula past the point where it lies.
This is the working method we use at ASOhack.
The simple formula (and why it lies)
LTV = ARPU × Avg subscriber length
= $7/mo × 8 months
= $56
This works when:
- Churn is roughly constant month-to-month.
- You have enough cohort history to know "8 months."
- Your product is in steady state (no recent paywall or pricing changes).
It lies when:
- You're a new app and don't know what "average length" is yet.
- Churn isn't constant — most apps churn hardest in months 1-2, then stabilize.
- You changed pricing recently and old cohort data doesn't reflect new behavior.
Three estimators, ranked by honesty
Method 1: simple (estimate)
LTV = ARPU × (1 / monthly churn)
If monthly churn is 10%, average subscriber length is 1/0.10 = 10 months.
Use when: you have <3 months of data. Worst-case estimator, but useful for sanity check.
Method 2: cohort-based (actual)
Pull each cohort and measure their actual cumulative revenue at month 1, 2, 3, etc. Average across cohorts.
Cohort Jan 2026:
M1: $200
M2: $350 (paid renewals)
M3: $470
M6: $650
M12: $810
You can plot revenue accumulation over time. The shape tells you LTV trajectory.
Use when: you have 6+ months of cohorts.
Method 3: decay-curve (best for projection)
Fit a decay curve (exponential or power law) to your cohort retention. Project forward to month 24, 36, etc.
Retention(month n) = a × n^(-b)
LTV = ARPU × Σ Retention(n) over n=1..N
Most analytics tools (RevenueCat, Apphud, etc.) do this automatically. Manually it's a regression in a spreadsheet.
Use when: you have 4+ months of retention data and want to estimate LTV beyond your observation window.
The Apple/Google tax mistake
Your LTV must use net revenue, not gross.
- App Store / Play standard: 30% take.
- App Store Small Business / Play first year: 15%.
- After year 1 of subscription on App Store: drops to 15%.
If your subscription is $9.99/month, net revenue is roughly:
- Year 1: $9.99 × 0.70 = $6.99/month
- Year 2+: $9.99 × 0.85 = $8.49/month
A subscriber who stays 18 months nets you roughly:
12 months × $6.99 + 6 months × $8.49 = $134.82
Not the $179.82 the simple formula would give.
LTV inputs you need to track
Three numbers, monthly:
- Monthly ARPU = total net revenue / paying subscribers.
- Monthly churn rate = subscribers who didn't renew this month / subscribers eligible to renew.
- Cohort size by acquisition month, for retention curves.
Most subscription tools (RevenueCat, Apphud, Adapty) give you these directly.
LTV by acquisition channel
LTV is not uniform across users. Organic > paid almost always. Paid LTV varies by channel:
| Channel | Typical LTV indexed to organic |
|---|---|
| Organic | 100% (baseline) |
| Apple Search Ads | 80-100% |
| Google App Campaigns | 60-80% |
| Meta | 50-70% |
| TikTok | 40-60% |
| Unity / IronSource (rewarded) | 20-40% |
Why this matters: a 100% blended LTV ROAS can hide a 200% organic ROAS subsidizing a 40% TikTok ROAS. Calculate per-channel, not blended, for spend decisions.
How to use LTV in real decisions
Decision 1: How much can I bid for an install?
Max bid CPI = (Net LTV × target ROAS) × Install → Paid conversion
If net LTV is $50, target ROAS is 100% (break-even), and 10% of installs convert to paid:
Max bid CPI = $50 × 1.0 × 0.10 = $5
You can pay up to $5 per install and still break even.
Decision 2: What discount can I offer?
If your LTV is $50 and you offer a 50% off first year ($30 instead of $60), the discount cost is ~$30. That's profitable to offer to acquisition channels with CPI ≤ $20 net of the discount.
See the intro pricing guide for the full strategy.
Decision 3: How long should my trial be?
Short trial (3 days): high trial start, lower trial-to-paid conversion. Long trial (14 days): lower trial start, higher trial-to-paid conversion.
The right answer depends on whether your aha moment fits in 3 days or needs 14. See trial conversion data.
The 90-day rule
Most indie devs don't have enough data for accurate LTV until they've been running for 6+ months. The 90-day rule:
- Day 0: use industry benchmarks. Trust them ±50%.
- Day 90: refine with your actual data. Trust ±25%.
- Day 365+: confident LTV estimate. Plan annual spend off this.
Until day 90, don't bet the company on a single LTV number.
Common mistakes
- Using gross revenue. Always net of Apple/Google take.
- Single LTV across channels. Per-channel always.
- Forgetting refunds. Apple/Google refund users at-will. Subtract from net revenue.
- Not segmenting by country. US LTV is often 3-5× tier-3 markets. Mixed cohorts mislead.
- Stale LTV after pricing changes. Recalculate after any paywall/price test.
Run the numbers
Use the Ad Analytics Calculator to plug in your inputs and see LTV + ROAS scenarios. Pair with the Ad Benchmark Analyzer to see where your unit economics sit vs category medians.
Related reading
Ready to Optimize Your App Store Listing?
Try our free ASO tools — no signup required.