CLV, or customer lifetime value, is the total net value a business expects to earn from one customer over the entire relationship, usually expressed as contribution margin over the expected lifetime, discounted to present value. In performance marketing, CLV sets the upper limit on what a customer is worth acquiring, which makes it the counterweight to CAC and the basis for bidding, budgeting and channel decisions.
CLV appears in two forms. Historical CLV sums what past customers actually generated, while predictive CLV estimates what a new or current customer will generate from purchase frequency, order value, margin and retention. The second form is what marketers use when they decide how much to spend today.
How customer lifetime value is calculated
The simplest useful formula multiplies average order value × purchase frequency × gross margin × expected lifetime. A customer who spends 60 dollars four times a year at a 40 percent margin for three years yields a CLV of about 288 dollars in margin, not 720 dollars in revenue. Using revenue instead of margin is the most frequent and most expensive error.
For subscription businesses, a cleaner version divides monthly margin per customer by the monthly churn rate: 30 dollars of margin and 5 percent churn give an expected lifetime of 20 months and a CLV of 600 dollars. More rigorous models add a discount rate, commonly 8 to 15 percent a year, because money earned in year three is worth less than money earned now.
The table lists the inputs of a standard CLV model and the typical ranges that shape the result.
| Input | What it represents | Typical range or source |
|---|---|---|
| Average order value | Revenue per transaction | From order data, by segment |
| Purchase frequency | Orders per customer per year | 1-2 for durables, 6-20 for consumables |
| Gross or contribution margin | Share of revenue kept after variable costs | 20-80% depending on category |
| Retention or churn rate | Probability the customer stays active | Cohort analysis, monthly or yearly |
| Discount rate | Time value of future profit | 8-15% a year in planning models |
| Time horizon | Period over which value is counted | Often capped at 2-5 years |
Historical, predictive and segment-level CLV
Different methods suit different data maturity. Cohort-based historical CLV is transparent and easy to audit but describes the past. Probabilistic models, such as the BG/NBD model for purchase frequency combined with a Gamma-Gamma model for spend, estimate future behavior for non-contractual businesses where customers leave silently.
| Approach | How it works | Strength | Weakness |
|---|---|---|---|
| Simple formula | Averages multiplied together | Fast, intuitive | Hides variance between customers |
| Cohort historical | Cumulative margin per acquisition cohort | Grounded in real data | Slow to reflect change |
| Probabilistic (BG/NBD) | Models purchase and dropout probability | Works without explicit churn | Needs clean transaction history |
| Machine learning | Predicts value from early behavior signals | Captures many features | Opaque, needs large datasets |
Averages conceal the most useful insight: value is highly concentrated. In many businesses a minority of customers produces most of the lifetime margin, so segment-level CLV by channel, first product or campaign is far more actionable than a single company-wide figure.
CLV in relation to CAC, ROAS and conversion rate
CLV becomes a decision tool only when it is compared with acquisition cost. The CLV to CAC ratio shows how many dollars of lifetime margin each dollar of acquisition spend returns; a ratio of about 3:1 is a widely used rule of thumb for a healthy growth business, though the right level depends on cash position and payback needs.
| Metric | Time frame | Question it answers |
|---|---|---|
| CLV | Whole customer relationship | What is a customer worth? |
| CAC | Acquisition moment | What did it cost to win the customer? |
| ROAS | Usually first order or attribution window | How much revenue did ads return? |
| Conversion rate | Single session or funnel step | What share of visitors acted? |
A campaign with a weak first-order ROAS can still be highly profitable when the customers it attracts have a high CLV. Conversely, a promotion with a strong ROAS may recruit discount-driven buyers who never return, which lowers CLV across the cohort.
Using CLV to steer acquisition budgets
In practice, CLV is used to set maximum allowable acquisition cost per channel or segment and to feed value-based bidding. Ad platforms that accept conversion values can optimize toward predicted lifetime value instead of first-order revenue when the business passes those values back.
| Decision | How CLV is applied | Practical caution |
|---|---|---|
| Bid and CPA ceilings | Allowable cost set below segment CLV | Leave margin for overhead and error |
| Channel budget mix | Shift spend toward high-CLV sources | Check that the effect is causal |
| Retention investment | Compare cost of saving a customer with value kept | Target customers at genuine risk |
| Payback planning | Time until cumulative margin exceeds CAC | Cash flow can limit growth even at high CLV |
The payback period deserves equal attention. A CLV of 600 dollars against a CAC of 200 dollars looks excellent, but if the margin arrives over 30 months, the business must finance that gap; many teams therefore cap the CLV horizon at 12 to 24 months for bidding decisions.
Pitfalls and misconceptions
CLV is a forecast, and forecasts drift. The most common problems come from optimistic assumptions that quietly inflate acceptable acquisition costs.
- Revenue instead of margin: inflates CLV by the inverse of the margin rate, often two to five times.
- Infinite or very long horizons: assume retention that newer cohorts may not reach.
- One average for all customers: hides the gap between loyal buyers and one-time bargain hunters.
- Ignoring returns and service costs: refunds, support and shipping subsidies reduce real value.
- Stale models: pricing, product or market changes make last year’s CLV unreliable within months.
A sound practice is to recalculate CLV by cohort every quarter, compare predictions with actual results, and treat the number as a range rather than a fixed truth.
Put simply
CLV is the answer to the question: how much is a customer worth to us over all the years they stay? A coffee shop regular who buys a latte every weekday is worth far more than the price of one cup.
Knowing this number helps a business decide how much it can sensibly spend to win a new customer. If a customer brings in more over time than it costs to attract them, the spending pays off, even when the first purchase alone does not.
In practice
A pet food subscription company acquires customers through social ads at a CAC of 70 dollars. The first box earns only 12 dollars of margin, so first-order ROAS looks poor and the finance team questions the channel.
The marketing analyst builds cohort CLV: customers keep subscribing for an average of 14 months with 11 dollars of monthly margin, giving a 12-month CLV of about 130 dollars. However, customers acquired through a 50 percent first-box discount churn twice as fast, with a 12-month CLV near 65 dollars, below their CAC.
The company ends the deep discount, moves budget to audiences resembling its longest-staying subscribers and passes predicted values to the ad platform. Within two quarters, blended CAC rises slightly to 75 dollars, but 12-month CLV climbs to about 150 dollars, and the channel reaches a CLV to CAC ratio of 2:1 within the first year.