Lifetime value (LTV) is the total revenue you expect from a customer over the whole time they stay with you. For a subscription app, a simple estimate is LTV = net monthly revenue per subscriber ÷ monthly churn rate. For a mobile game, teams usually track cumulative LTV, the revenue an install cohort has generated by day N divided by the number of installs, and forecast the full lifetime from it.
Which one you reach for depends on the business. Either way, the number only matters once you set it against what you pay to acquire a user.
What is the basic LTV formula?
For a business with steady recurring revenue, the customer lifetime value formula is
LTV = monthly average revenue per paying subscriber (ARPU) × average customer lifetime in months
When churn and revenue per subscriber stay roughly steady, the average lifetime is about 1 ÷ monthly churn rate, so the formula is often written as
LTV = net monthly revenue per paying subscriber ÷ monthly subscriber churn rate
Take an example app with net revenue of $9.99 per subscriber per month that loses 8% of its subscribers each month. The average lifetime is 1 ÷ 0.08, which is 12.5 months, and the LTV is $9.99 × 12.5, which is about $125.
Use the revenue you actually keep. Platform fees, refunds and applicable taxes come out before the money reaches you, so define net revenue, deduct them the same way every time and say so next to the number. If you want LTV as profit, apply your gross margin to that net figure.
| Version | Formula | When to use it |
|---|---|---|
| Simple subscription LTV | Net ARPU per month ÷ monthly churn | Stable monthly plans, early planning |
| Margin LTV | Net ARPU × gross margin ÷ monthly subscriber churn | Comparing LTV with subscriber acquisition cost as profit |
| Cumulative cohort LTV | Cohort revenue to day N ÷ installs in the cohort | Games, apps with in-app purchases or ads |
| Predicted LTV | Early cohort LTV × multiplier from mature cohorts | Judging a campaign before its cohort matures |
Our guide to calculating churn rate covers the churn half of the formula, including why monthly churn should not be multiplied by 12.
Why does the simple formula fail for mobile games?
When you use the simple formula, you treat average revenue and churn as roughly steady over time. In many free-to-play games neither is. Most players never pay at all, nearly all the revenue comes from a small share of payers, and players leave fastest in the first days after install. Average everyone together and you lose sight of all of that.
In an example cohort of 10,000 installs, 300 players make a purchase in the first 30 days, and those 300 account for all of the $11,000 in purchase revenue. Average revenue per install is $1.10, while average revenue per paying player is about $36.67. If you apply the monthly formula here, you blend the 9,700 players who never paid with the 300 who did into one average and treat revenue and churn as steady from month to month.
So game teams measure LTV by install cohort. Group players by install date, split each cohort by campaign, channel or platform where it matters, and track how much revenue each cohort has produced after 1, 7, 30 and 90 days.
How do you calculate cumulative cohort LTV?
The formula is
LTV on day N = total revenue from the cohort from install through day N ÷ number of installs in the cohort
Divide by all installs, including players who never paid, so the figure is comparable with the cost per install you paid to acquire the cohort.
Here is an example cohort of 10,000 installs acquired at $1.20 each.
| Day | Cumulative revenue | Cumulative LTV per install | Return on ad spend |
|---|---|---|---|
| D1 | $1,800 | $0.18 | 15% |
| D7 | $5,200 | $0.52 | 43% |
| D30 | $11,000 | $1.10 | 92% |
| D90 | $17,500 | $1.75 | 146% |
By day 30 the cohort has generated revenue equal to 92% of its acquisition cost, and by day 90 it has generated 146%, so revenue payback falls somewhere between the two. Teams usually track the exact payback day, the first day cumulative LTV reaches the cost per install, because finance uses it to see how long acquisition spending stays unrecovered.
How do you calculate LTV by acquisition channel?
Run the same calculation for each channel or campaign, because the cheapest installs are often not the most valuable ones. Here is an example with two channels measured at the same cohort age.
| Channel | Cost per install | D30 LTV per install | D30 return on ad spend |
|---|---|---|---|
| Channel A | $0.90 | $0.70 | 78% |
| Channel B | $1.60 | $1.70 | 106% |
Channel A looks cheaper on cost per install, yet Channel B has already paid back by day 30. A team looking only at cost per install would shift budget toward Channel A, even though Channel B has the stronger D30 return.
Two checks matter here. Compare cohorts at the same age, so a 30-day-old cohort is never measured against a 90-day-old one. And split cohorts by acquisition source, because one blended number can hide a channel that never pays back.
How do you predict LTV before a cohort matures?
Waiting 90 days to judge a campaign is too slow for most acquisition teams. The common approach is to predict later LTV from early LTV using ratios from your own mature cohorts.
- Take cohorts that are at least 90 days old and similar in source, platform and market.
- For each one, divide D90 LTV by D7 LTV to get a multiplier. In the example cohort above, $1.75 ÷ $0.52 is about 3.4.
- Apply the range of multipliers to a new cohort's D7 LTV. If your mature cohorts show multipliers between 3.0 and 3.7, and a new cohort's D7 LTV is $0.60, its predicted D90 LTV is between $1.80 and $2.22.
Before you rely on a multiplier, test it on cohorts that have already reached day 90 and record how far the predictions were off. Report the range with every prediction, and mark which cohorts are still immature. A handful of large purchases in a small cohort can make an early LTV look far better than it will end up, so check the prediction again as the cohort ages and re-estimate the multipliers as more cohorts mature or when pricing, product or acquisition mix changes.
How do you include ad revenue in LTV?
Many games and free apps earn from ads as well as purchases. Calculate the two parts separately, then add them.
Ad LTV on day N = cohort ad revenue through day N ÷ installs in the cohort
If you track ad revenue per daily active user, calculate it for the cohort as ad revenue ÷ total active user-days over the same window, then multiply by the average active days per install. In an example cohort where that figure is $0.04 and the average install was active on 9 of its first 30 days, ad LTV at D30 is $0.04 × 9, which is $0.36. Add that to purchase LTV at D30 for the full figure. Leaving ad revenue out understates LTV and can make a profitable ad-supported cohort look as if it never pays back. Our customer story on following every install from ad spend to ad revenue shows what that full view looks like in practice.
What about LTV for subscription apps?
Subscription apps can use the simple formula, with two adjustments.
Separate plan types. Monthly and annual plans behave very differently. RevenueCat's State of Subscription Apps 2025 reports that nearly 30% of annual subscriptions are canceled in the first month, and that most apps with cheap annual plans keep up to 36% of users after a year, while high-priced monthly plans keep just 6.7%. A single blended LTV hides how differently the two plan types behave. For annual plans, take the net revenue for each renewal year, multiply it by the share of subscribers expected to reach that renewal, and add the years up over a horizon you state.
Start the clock at the trial. If your app offers a free trial, measure LTV per trial start and per first paid conversion as two separate numbers. Compare each one with an acquisition cost on the same basis, so LTV per trial start goes against cost per trial start and LTV per paid conversion goes against cost per paid conversion.
For example, if 1,000 people start a trial, 40% convert to a paid monthly plan, and LTV per paying subscriber is $125, then LTV per trial start is $125 × 0.40, which is $50. If acquiring a trial start costs $35, LTV is about 1.4 times acquisition cost, and the $125 arrives over an average of 12.5 months. That leaves little room once you account for margin and the cash tied up while you wait, even though $125 per subscriber looks comfortable.
How do you compare LTV with acquisition cost?
The usual checks are the ratio and the payback day.
- LTV to acquisition cost ratio. Match the unit on both sides, per install against cost per install and per subscriber against subscriber acquisition cost. State whether LTV uses net revenue or gross profit, and calculate it per channel.
- Revenue payback day. The first day cumulative net revenue per install reaches the cost per install. In the example cohort it falls between day 30 and day 90. For profit payback, apply your margin first.
Check the two together. A strong ratio helps little if the money comes back slowly, and a fast payback on a thin ratio limits how far you can scale. Agree on target values for both with your finance team before a campaign launches, then use them when you decide to scale, hold or stop a channel. Century Games used this kind of view to shift budget toward high-LTV channels based on real-time revenue feedback.
How to calculate LTV in Agentic Engine
In Agentic Engine, the Retention Analysis model covers the LTV and ROI questions. Our product documentation lists examples such as what the 30-day LTV of new users has been over a past period and on which day the ROI of each ad channel breaks even. You choose a first event that marks a new user, such as registration, and a return event such as login, then turn on Also Show to calculate metrics such as LTV and ROI for the users who completed the return event. That population differs from all installs, so confirm the denominator before you compare the result with cost per install.
For the revenue side, the payment analysis best practice starts with Events Analysis for the total count, number of paying users and total amount of payments, then uses Distribution Analysis to group users by total payment amount and Interval Analysis to see how long players take to go from registration to their first payment.
Common LTV mistakes
- Using gross revenue. Store fees and refunds come out before you see the money. Calculate LTV from net revenue.
- Dividing by payers. Divide cohort revenue by installs when you compare it with cost per install.
- Mixing cohort ages. Compare D30 with D30. Older cohorts have had more time to earn revenue.
- Blending channels. When you average all channels together, you can miss the one that never pays back.
- Treating predictions as facts. Show the range and revisit it as the cohort matures.
Frequently asked questions
What is LTV?
LTV, or customer lifetime value, is the total revenue a customer brings in over the whole time they stay. Teams use it to decide how much they can spend to acquire a customer.
How do you calculate LTV for a subscription app?
Divide net average revenue per user per month by the monthly churn rate. At $9.99 a month and 8% monthly churn, LTV is about $125. Calculate monthly and annual plans separately.
How do you calculate LTV for a mobile game?
Use cumulative cohort LTV. Add up the revenue an install cohort has generated by day N and divide it by the number of installs. Include ad revenue if the game shows ads.
What is the difference between LTV and ARPU?
ARPU is average revenue per user over a single period, such as a day or a month. LTV adds up revenue over a user's whole lifetime, so it depends on both ARPU and how long users stay.
How can I increase LTV?
Work on retention, the share of users who pay, or how much payers spend. Change one thing at a time and compare it with a group of users who do not get the change. For a real cohort example, see our guide on how to increase customer lifetime value.
Where to go next
LTV depends on how long users stay, so retention is the other half of the picture. See how to calculate retention rate for apps and games for D1, D7 and D30. To connect LTV back to the ads that brought users in, see creative testing that follows users from clicks to revenue.
