What it means
A customer who buys once is worth the contribution on that purchase. A customer who buys every month for five years is worth sixty months of contribution, less the cost of keeping them, discounted for the wait.
Lifetime value is the second calculation, and it changes how a business thinks about customers: not as transactions but as relationships with a present value, which can be bought (through acquisition), grown (through retention and expansion) and lost (through churn). The simplest version multiplies contribution per period by expected lifetime, where lifetime is one divided by the churn rate: a customer contributing $50 a month with 4% monthly churn has an expected lifetime of 25 months and a lifetime value of $1,250.
The formula assumes constant contribution and churn, no discounting and no retention cost, and it is the version most often quoted and most often too high. Several refinements make it more honest.
Discounting matters because contribution received in year five is worth less than contribution received now, and at a 10% discount rate a customer with 20% annual churn has a lifetime value roughly 30% below the undiscounted figure. Retention cost (account management, loyalty programmes, service and retention marketing) belongs in the calculation, and contribution rather than revenue should be used, since revenue-based lifetime value overstates the figure by the cost of goods.
Churn should be observed from actual cohorts, ideally by segment, rather than assumed as a target. Expansion matters too: customers who buy more over time (upgrades, cross-sales, price rises) have lifetime values above the constant-contribution formula, captured by net revenue retention.
Cohort curves help because churn is usually high early and lower later, so a survival curve fitted to actual data gives a better lifetime than a single rate. The uses follow.
Acquisition: the maximum sustainable CAC is a fraction of lifetime value (a third is the common rule, leaving margin for overhead, error and profit), and channels are funded up to that limit. Retention: the value of reducing churn is the increase in lifetime value across the base, which usually dwarfs what the retention programme costs.
Segmentation shows which customers justify better service and targeted offers (low-value segments may not justify acquisition at all), product and pricing changes can be tested against the trade-off between a higher price and higher churn, and a business's customer base can be valued as the sum of its customers' lifetime values, which is how subscription businesses are assessed by investors and acquirers. The cautions are as important as the uses: lifetime value is a forecast whose assumptions must be observed, tested and updated, since a figure built on a churn rate the business has never achieved is a hope, and it is an average that hides variation, because a few high-value customers pull the mean above the median customer.
It should also be discounted, capped in horizon (five years is common, since predictions a decade hence are worthless) and reconciled against the actual contribution of cohorts acquired three years ago.
In practice
Real-world examples
Example 1
A mobile operator calculates a discounted lifetime value of $900 per contract customer and sets its maximum handset subsidy accordingly.
Example 2
An online retailer finds that customers who buy from two categories in their first three months have a lifetime value three times those who buy from one, and redesigns its post-purchase marketing to cross-sell.
Example 3
An insurer values its policy book as the sum of its policyholders' expected future contributions, discounted, and reports the figure as embedded value.
Think of it
CLV is the total value a customer brings over their entire relationship-their long-term worth to you.
Formula
Calculation
Simple CLV = Contribution per period x Expected lifetime in periods, where Expected lifetime = 1 / Churn rate per period
Equivalent: CLV = Contribution per period / Churn rate
Discounted CLV = Contribution per period x [Retention rate / (1 + Discount rate minus Retention rate)] (for annual periods, constant retention)
Net CLV = Discounted lifetime contribution minus Retention and service costs over the lifetime minus CAC
Cohort CLV = Sum over periods of (Cohort's actual or forecast contribution in period / Cohort's starting size), discounted
Maximum sustainable CAC = CLV / Target ratio (commonly 3)
Worked example. A subscription software company: average revenue per customer $150 a month; gross margin 78% (contribution $117 a month, $1,404 a year); monthly churn 2.5%; retention cost (customer success, loyalty, retention marketing) $15 per customer per month; discount rate 12% a year; CAC $2,100.
Simple CLV = $117 / 0.025 = $4,680 (expected lifetime 40 months). LTV to CAC = 2.2.
Discounted CLV (annual basis): retention rate 74% a year (0.975 to the power 12); annual contribution $1,404. CLV = $1,404 x [0.74 / (1 + 0.12 minus 0.74)] = $1,404 x 1.947 = $2,734. Discounting and the annual approximation cut the figure by over 40%.
Net CLV: retention cost $15 x 40 months = $600 (undiscounted); discounted about $470. Net CLV = $2,734 minus $470 = $2,264; less CAC $2,100: $164. The company makes $164 of present value per customer after everything. LTV (net of retention cost) to CAC = 1.1. The growth model is marginal.
Cohort check: the cohort acquired 36 months ago (1,000 customers) has produced actual contribution of $2,450,000 to date, $2,450 per starting customer, with about 430 customers still active; the model's undiscounted prediction for 36 months was $2,800 per customer. The model is 14% optimistic; the churn curve shows 45% of the cohort left in the first year (worse than the 26% the constant rate implied) and 12% a year thereafter (better).
Segment analysis: enterprise customers (15% of the base) have monthly churn of 0.8%, contribution $600 a month and discounted CLV of about $31,000 against a CAC of $9,000: a ratio of 3.4. Small business customers acquired through paid search (40% of the base) have churn of 4%, contribution $60 and discounted CLV of about $870 against a CAC of $1,800: a ratio of 0.5. The blended figures hid a profitable segment and a loss-making one.
Decisions: acquisition spend is redirected from small-business paid search towards enterprise and towards small-business channels with lower CAC (content, partners); a retention programme targeting first-year churn is funded; the entry-level price is raised 15%, which the model predicts will raise monthly churn by 0.2 points and still lift CLV by about 10%; and the maximum CAC by segment is set at a third of discounted net CLV.
Valuation: the customer base of 12,000 has a discounted net CLV (before CAC, which has already been spent) of about $2,264 per customer on average, or about $27,000,000 in total, a figure the board compares with the acquisition offers it receives and with the enterprise value implied by its revenue multiple.Case study
Seen in the real world
This is a fictional, illustrative scenario. A consumer subscription company presented investors with a lifetime value of $420 per subscriber against a CAC of $95, a ratio of 4.4. The lifetime value was revenue-based ($30 a month over 14 months), used a churn rate from the company's best cohort, and was undiscounted. An investor's analyst rebuilt it: contribution rather than revenue ($30 less $12 of product and fulfilment cost: $18); churn from the blended cohort data (9% monthly, lifetime 11 months rather than 14); a 15% discount rate; retention costs of $2 a month; and a two-year cap.
The result was about $150. The ratio was 1.6, and after the company's overhead, each subscriber lost money. The investor declined the round; the company raised elsewhere on the original figures, spent the money on acquisition at $95 per subscriber, and ran out of cash eighteen months later. The analyst's note had made a point that has become standard in the sector: lifetime value is only as good as its worst assumption, and the assumptions most often wrong are revenue for contribution, target churn for actual, and no discount rate.
Watch out
Common mistakes
- Calculating lifetime value on revenue rather than contribution, which overstates it by the cost of serving the customer.
- Using an assumed or best-cohort churn rate rather than the observed blended rate, and leaving the value undiscounted and uncapped.
- Managing to an average lifetime value that hides segments with very different economics, some of which should not be acquired at all.
Questions
People also ask
What is a good lifetime value?
One that is at least three times the cost of acquiring the customer, calculated on contribution, observed churn and a discount rate, and confirmed by the actual contribution of past cohorts.
How is lifetime value used in decisions?
To set the maximum acquisition cost by channel and segment, to value retention improvements, to prioritise high-value customers, to test pricing changes, and to value the customer base.
Should lifetime value be discounted?
Yes, at the company's cost of capital, and its horizon should be capped (typically five years) because predictions of customer behaviour far ahead are unreliable. Undiscounted, uncapped figures are systematically too high.