LTV for mobile games is the revenue an average player brings in over their whole time with your game. It is the number every user acquisition decision leans on, because it tells you how much you can afford to pay for an install. Get it wrong and you either overspend on players who never pay back or underspend and let competitors take the traffic.

This post covers how we build LTV from cohorts, how to handle ad revenue and in-app purchases together, and how to turn the number into a mobile game user acquisition strategy you can act on.

What LTV means for a mobile game

Lifetime value is total revenue per install over a defined window. For games the window matters more than the label. A “lifetime” figure is really a forecast, because most players have not finished playing when you need to make the decision.

So in practice we work with two numbers:

  • Observed LTV: what a cohort has actually earned so far, for example by day 7 or day 30.
  • Predicted LTV: what we expect the same cohort to earn by day 180 or day 365, based on how earlier cohorts behaved.

Observed LTV is honest but slow. Predicted LTV is fast but only as good as the assumptions behind it. A healthy process uses both, and checks the prediction against reality as cohorts age.

Start with cohorts, not averages

A cohort is a group of players who installed on the same day, from the same source, in the same country. Always calculate LTV per cohort. A blended average across every install hides the thing you most need to see: that players from one network or one country behave very differently from another.

For each cohort, track revenue per install on day 1, 3, 7, 14, 30 and beyond. Plot it as a curve. Early on the curve climbs steeply, then flattens as players churn. The shape of that curve is your game’s monetisation fingerprint.

Two practical rules help:

  • Keep cohorts large enough to be meaningful. A handful of installs tells you nothing.
  • Separate paid installs from organic ones. Organic players often look better, and mixing them in flatters your paid results.

The simple formula and when it works

The quickest method multiplies average daily revenue per active user (ARPDAU) by the number of days a player is expected to stay. It is useful before launch, when you only have benchmarks from similar games and need a rough check on whether the economics could work.

Once you have real data, for example in a soft launch, the better method is to model the retention curve. You take retention at several points (day 1, 7, 14, 30), fit a curve through them, and project it forward. Multiply the projected active players on each day by ARPDAU and add it up. That gives a predicted LTV for the install, not just a guess at lifespan.

We cover how to gather those first readings safely in our guide to soft launching mobile games.

Combining ad revenue and in-app purchases

Most games we work with earn from both ads and purchases. For our own titles the mix is roughly 80 percent ads and 20 percent in-app purchases, though yours will differ by genre. The two streams behave differently, so track them separately before adding them together.

Ad revenue is earned by almost every player. It rises smoothly with sessions, and it depends on your ad placements, the player’s country and the demand in your mediation setup. It is predictable, which makes it good for early forecasts.

In-app purchase revenue comes from a small share of players, and a few of them spend a lot. It is lumpy. A cohort can look weak on day 3 and then jump when a few buyers appear on day 10. That makes it harder to predict from a short window.

Our approach is to model each stream on its own curve, then sum them. If you blend them too early, a change in ad fill or a pricing test will quietly distort the forecast, and you will not know which side moved. The game design choices behind both streams are covered in our post on monetisation design.

Turning LTV into a bid and a payback target

LTV only matters once it changes what you do. There are three uses.

Setting the maximum you can pay. If the predicted LTV of a cohort is lower than what you pay for its installs, you are losing money on that source. If it is higher, you have room to scale. The margin you need depends on your payback window and your appetite for risk.

Choosing a payback window. Payback is the day your cohort’s cumulative revenue passes what you spent on it. Short windows, such as day 7, protect cash. Longer windows, such as day 60 or beyond, let you bid for players who pay back slowly but are worth more. A studio with limited cash should lean short. One with a steady stream of revenue can afford to be patient.

Comparing sources. Because LTV is per cohort, you can rank networks and countries by what the players are actually worth, not by how cheap the install was. A cheap install that never pays is the most expensive one you can buy.

The link to return on ad spend is direct: ROAS is simply a cohort’s revenue divided by what you spent on it, at a given day. We explain how to judge it in what is a good ROAS for mobile games.

A mobile game user acquisition strategy built on LTV

A strong strategy starts with the economics and works outward. This is the order we follow:

  1. Measure LTV by cohort for your existing players, split into ads and purchases.
  2. Choose a payback window that fits your cash position.
  3. Set a bid ceiling per source from the predicted LTV and that window.
  4. Test new networks and countries with small, comparable cohorts.
  5. Re-check predictions against actual results every week and adjust.

Step four is where many studios go wrong. If you compare a new network’s day 1 result against an old network’s day 30 result, the new one will always look bad. Compare like for like, at the same day, and wait until the cohort has matured enough to trust.

For the wider picture of how this fits into planning, see what user acquisition is for mobile games and our guide to mobile game UA budget planning.

Common mistakes with LTV

  • Trusting one number. LTV is a range, not a point. Hold it loosely, especially for young games.
  • Ignoring country mix. A cohort that is mostly high-value countries will look better than the same game in a broader launch. Mix shifts can look like performance changes.
  • Using stale curves. Retention and monetisation drift after updates, new content and ad changes. A curve from six months ago may not describe today’s players.
  • Counting revenue you do not keep. Store fees and other costs reduce what a purchase is worth to you. Use the net figure when comparing against what you spend.
  • Forgetting seasonality. Players spend and engage differently across the year, so check forecasts against the same period last year where you can.

How we use LTV day to day

Across 40 months of aggregated results, our blended ROAS has stayed above break-even. That does not come from one clever number. It comes from checking predicted LTV against real cohorts week after week, trimming sources that fall behind, and putting more into the ones that hold up. Applovin leads our mix, followed by Google Ads, Mintegral and Unity, and each is judged on the LTV of the players it brings, not on install cost alone.

If you take one thing from this post, make it this: calculate LTV by cohort, split ads from purchases, and always compare sources at the same day. Everything else in your acquisition plan gets easier once that is in place.