Quick answer
LTV (Lifetime Value) is the value you expect to earn from a customer over time. For operating decisions, the number is usually most useful when it is based on gross profit, not just revenue. The real question is not only how to calculate LTV, but how much confidence you should have in that estimate before you use it to justify CAC, payback, or growth plans.
A common shortcut
LTV ~= (ARPA * gross margin) / churn rate (with consistent time units).
When the shortcut is good enough
- The business has relatively stable churn and limited expansion noise.
- You need a quick planning estimate before you build a heavier cohort model.
- ARPA, gross margin, and churn all come from the same segment and the same time unit.
When the shortcut breaks
- Expansion revenue is significant and simple churn formulas understate value timing.
- Churn changes over time, especially when early churn is much higher than later churn.
- Different segments behave differently and blended averages hide what is really happening.
The assumptions that matter most
- Use gross margin, not operating margin, when the number is meant to support CAC and payback decisions.
- Match monthly ARPA with monthly churn, or annual ARPA with annual churn.
- Be explicit about logo churn vs revenue churn because the two can tell very different stories.
Customer lifetime is often the next question
- If you want to understand how long customers actually survive, go next to Customer lifetime.
- If you need to see how fragile the estimate is, go next to LTV sensitivity.
- If the shortcut feels too thin, go next to Cohort LTV forecasting.
What to pair with LTV
- Use CAC and payback when the question is whether acquisition still works in cash terms.
- Use LTV:CAC when the question is long-term sustainability, but only after you align definitions.
- Use NRR/GRR and cohort retention when expansion or contraction changes the lifetime story.
Common mistakes
- Using revenue LTV while comparing it to fully-loaded CAC.
- Mixing monthly churn with annual ARPA or other unit mismatches.
- Letting one low blended churn number hide segment instability.
- Treating a shortcut estimate as precise when the business clearly needs cohorts.