How to calculate LTV for mobile user acquisition without a lifetime of data
Originally published August 2019. Some details have changed since: iOS tracking now requires ATT permission, and privacy-preserving ad attribution has limited, delayed postbacks. That can reduce campaign-level attribution detail; your own app can still record customer revenue and retention.
You do not need a lifetime of data to estimate lifetime value (LTV). A simple spreadsheet model is often good enough to get started, and most apps do not need a sophisticated one, which usually needs resources many teams do not have.
Where is the method explained?
In a video, How to Calculate Mobile Lifetime Value (LTV) Without a Lifetime of Data, published 15 August 2019 on the AppMasters YouTube channel and about 12 minutes long. The spreadsheet demonstration is in the linked video. This page preserves the original recommendation to start simply; it does not reproduce the calculation steps.
The point of the video, in the same words as the original post: there are more sophisticated ways to calculate LTV, but this simple approach is often enough to get started, and a simple model is very often good enough.
A small planning example
This hypothetical example is separate from the 2019 video. Suppose 1,000 acquired users have generated $2,000 of net receipts by day 30: observed day-30 value is $2 per acquired user, including non-payers. If comparable, mature cohorts earned 60% of their day-180 net receipts by day 30, a starting forecast is $2 / 0.60 = $3.33 per user by day 180. The extra $1.33 is forecast, not cash already earned. Use a range if the 60% relationship varies, and do not assume it transfers to a new price, country or acquisition mix. Deduct the costs needed for your contribution-margin target before translating value into an affordable acquisition cost.
How do you know your LTV estimate is any good?
Any LTV forecast built on early data is a prediction, so check it against what actually happened. Compare the forecast for a cohort with the revenue that cohort really produced, and adjust. We cover that in how backtesting can protect you from unrealistic LTV expectations.
What do you do with the number?
Use it to decide how much you can pay for a user and how long you can wait to get your money back, but be clear about which revenue you count. See how to measure your mobile marketing ROAS: 3 approaches and how to set a mobile user acquisition budget.
Related reading
Frequently asked questions
How do you calculate LTV with limited data? The video builds a simple spreadsheet model from the data you have rather than waiting for a lifetime of it. Check the forecast against real cohorts as they mature.
Is a simple LTV model good enough? The original post recommends a simple model as a starting point. Validate its forecasts against your own cohorts before using it to guide spend.
Why does iOS make LTV harder to estimate? ATT and privacy-preserving ad attribution can limit how precisely you assign customers to campaigns. That is different from losing your own customer revenue history: model the revenue you can observe and be explicit about attribution gaps.
Published
