LTV Modeling for Mobile Apps: A Practical Approach
LTV is the number that sets your allowable CAC. Model it wrong and every bid is wrong. Here’s a practical way to do it.
Key takeaways
- Model net LTV (after fees, refunds and churn), not gross revenue.
- A simple method: monthly margin per user × average lifetime (≈ 1 ÷ churn).
- LTV sets your allowable CAC; predict it early so you can act before cohorts mature.
LTV (lifetime value) is what a user is ultimately worth to you. It’s the number that decides how much you can pay to acquire them, so getting it roughly right matters more than getting it precisely late.
Net, not gross
Always model LTV net of store fees, refunds and churn. Gross revenue LTV flatters every decision and quietly puts you underwater. The number you bid against has to be the money you actually keep.
A simple, usable method
Monthly margin per user multiplied by average lifetime, where lifetime is approximately one divided by monthly churn. It’s not perfect, but it’s honest and fast, and it’s enough to set bids while you build something more sophisticated.
Predicting LTV early
Real LTV takes months to confirm, which is too slow for a media account. So you build an early predictor (often from D3 to D7 behavior) and back-test how well it tracks mature cohorts, then act on the prediction.
How LTV feeds CAC
Your allowable CAC is a function of LTV and your payback target. Lift LTV and you can bid higher at the same payback, which is why funnel and paywall work is really CAC work in disguise.
Turn your LTV into an allowable CAC and payback.
Open the calculator