Customer lifetime value estimates how much revenue a single customer produces across their whole relationship with you. In subscription businesses it is derived from average revenue per account and churn rather than observed directly, because most customers have not left yet.
The intuition is simple. If 3% of customers leave each month, the average customer stays about 1 ÷ 0.03, or roughly 33 months. Multiply that lifetime by what an average customer pays per month and you have their lifetime value.
Two conventions exist and they answer different questions. The revenue form, ARPA ÷ churn, estimates the revenue a customer will produce. The margin form, ARPA x gross margin ÷ churn, estimates the profit, and is the one to use when comparing against acquisition cost. Neither is wrong; using one where the other belongs is.
LTV is an estimate built on an assumption that churn stays where it is. It is most useful as a relative measure, comparing segments or tracking your own trend, and least useful quoted to two decimal places as if it were an observation.
LTV = ARPA ÷ customer churn rate
The margin form, for comparing against CAC
LTV = ARPA x gross margin ÷ customer churn rate
Given
A point of churn is worth more than a lot of pricing work here. At 4.1% instead of 3.1%, the same ARPA gives an LTV of $3,724, a quarter of the value gone.
Comparing revenue-based LTV to a fully loaded acquisition cost compares a gross number to a net one and flatters the ratio by whatever your gross margin is not.
LTV divides by churn, so a quiet month makes lifetime value explode. Average over several months before dividing.
A blended figure across self-serve and enterprise describes no real customer. Segment first, then divide.
LTV is a projection that assumes today's churn holds for the next several years. It is a planning input, not an asset.
Kometrics computes LTV as ARPA divided by the average customer churn rate of the six preceding months, matching ChartMogul so the two can be reconciled directly. The month being measured never contributes its own churn to that average, which is the detail most reimplementations miss: the lag is deliberate, and it stops one noisy month from swinging the whole series.
The result is a revenue LTV, not a margin one, because Kometrics reads your billing data and has no way of knowing your cost of goods. If you are comparing against acquisition cost, multiply by your gross margin first. The free LTV calculator on this site takes margin as an input for exactly that reason.
ARPA is your recurring revenue divided by the number of paying accounts: what the average customer is worth per month. It is also called ARPU or ARPC, with account, user and customer used loosely to mean the same thing in most SaaS reporting.
Churn rateChurn rate is the share of your customers, or of your revenue, that you lost during a period. Which of the two you mean changes the number substantially, so the qualifier matters as much as the figure.
Cohort retentionCohort retention groups customers by the month they started paying and follows each group forward, so you can see how long customers stay and how their spend develops, without newer customers masking older ones.
MRRMRR is the normalized monthly value of every active subscription you have right now. It is a run rate, not a cash figure: it answers what a month of your current book is worth, not what landed in the bank.
Kometrics connects to Stripe, Paddle, Creem and Asaas and computes LTV continuously, with every movement behind it. Free under $1,000 MRR.
Start free