Customer lifetime value (CLV)

How to calculate customer lifetime value

Multiply three numbers for the same group of customers and the same period:

CLV = average order value × orders per customer × gross margin

Customer lifetime value is also called lifetime value (LTV) or CLTV. The name changes; the method matters more. Some tools report revenue per customer, some report margin, and some report a forecast, all under the same label. Before you compare two figures, check which one each is.

Use margin, not revenue. A customer who spends $76 with you over a year did not leave you $76: at a 50% gross margin, the goods cost you $38 of it. Revenue-based CLV makes every customer look twice as valuable as they are at a 50% margin, and the error is larger for low-margin stores. See average order value for what to count as revenue, and gross margin for what to count as cost.

Read it by cohort

A cohort is every customer whose first order fell in the same month. Follow one cohort forward and add up what it has bought. Take 200 customers whose first order was in January 2025, an average order value of $40 and a gross margin of 50%:

Divide by all 200 customers in every row, including those who never came back. Dividing by the customers who bought again makes the figure rise as people leave. At 12 months the formula gives the same answer as the table: $40 × 1.9 × 0.5 = $38. Revenue per customer at 12 months is $15,200 ÷ 200 = $76, twice the margin figure.

In Tanqory, each customer profile shows the order count and total spent, and you can export any customer list as a CSV file with orders and total spent. Total spent is revenue, so multiply it by your gross margin before you treat it as CLV.

Historical and predicted lifetime value

Historical CLV is what customers have already bought, as in the table. It is a fact, and it only grows. Its weakness is that young cohorts have had little time to buy, so their figure is low because they are new, not because they are worse.

Predicted CLV adds an estimate of what customers will still buy. The simplest model assumes each order is followed by another with a fixed chance. If that chance is 60%, a customer places 1 ÷ (1 − 0.6) = 2.5 orders on average. At $20 of margin per order, predicted CLV is $20 × 2.5 = $50. The January cohort has reached $38 at 12 months, so the model expects $12 more over the rest of the relationship.

Every prediction rests on its assumption. Here it is that customers keep coming back at the same rate forever, which few do. Check each prediction against the historical figure as the cohort ages, and lower the rate when the cohort falls short.

Choosing a time window

"Lifetime" has no end date, so a usable figure needs one. Pick a window that fits how often your customers buy:

• Short buying cycles (coffee, pet food, skin care): 12 months shows several repeat orders.
• Long buying cycles (furniture, bicycles, mattresses): 24 or 36 months, because a second order may be years away.
• New stores: use whatever window your oldest cohort has completed, and say which it is.

Compare cohorts only at the same age. A cohort at 6 months cannot be compared with one at 12 months. Keep the window fixed once chosen, so this year compares with last year.

Customer lifetime value and acquisition cost

CLV is most often used as a check on what you spend to win a customer. Customer acquisition cost is marketing spend divided by the new customers it brought in. If the January cohort cost $5,000 in advertising, acquisition cost is $5,000 ÷ 200 = $25.

The first order left $20, less than the $25 it cost to win the customer. By 3 months the cohort had earned $26 each, so the spend was paid back between the first order and the third month.

Gross margin leaves out the costs of each order. If packing and shipping cost you $7 an order, the cohort's 1.9 orders cost 1.9 × $7 = $13.30 per customer, which leaves $38 − $13.30 = $24.70. By that count the January customers have not paid back their $25 after 12 months. Revenue-based CLV hides all of this: $76 ÷ $25 = 3.04 looks safe.

We do not give a target ratio. The ratio you need depends on your fixed costs, which costs you subtract per order, whether acquisition cost counts all marketing or paid ads only, and which window you use. A ratio above 1 on contribution margin means the customers paid for themselves; what is left has to cover rent, wages and everything else.

What to count

Two stores with the same customers can report different CLVs. Pick one rule and keep it:

• Customers. A guest who checks out with two email addresses becomes two customers, each with fewer orders. Merge duplicates before you count orders per customer.
• Refunds and returns. Subtract refunded orders, or a customer who returns everything counts as valuable.
• Discounts. Use what the customer paid after discounts. A $50 order with a $10 code is a $40 order.
• Tax and shipping charges. Leave both out, as for average order value.

When a few customers carry the average

CLV is a mean, so a few very large customers can move it a long way. Take 100 customers: 98 left $50 of margin each in 12 months, and 2 that buy for their own shops left $3,000 each. Total margin is 98 × $50 + 2 × $3,000 = $4,900 + $6,000 = $10,900, so mean CLV is $10,900 ÷ 100 = $109. The median customer left $50.

If you set your acquisition spend from $109, you overpay for nearly every customer you win. Report the median beside the mean, or put large and trade buyers in their own segment and calculate CLV for each group.

Ways stores raise customer lifetime value

Each part of the formula is a lever, and each lever has a cost:

• Raise average order value. Bundles and free shipping thresholds raise each order. They can also lower conversion rate, so watch revenue per visitor.
• A message after the first order. An email when the product is likely running out brings back customers who would otherwise drift away. It costs the sending, and the unsubscribes if you send too often.
• Subscriptions and refills. A repeating order raises orders per customer directly. The cost is the subscription discount, paid on orders many customers would have placed anyway.
• A discount code for a second order. It moves the cohort from 1 order toward 2. The cost is the discount on every second order, including those that would have happened without it.
• Raise gross margin. A higher price or a cheaper supplier raises margin on every order. The cost is the orders you lose to the higher price, or the returns if quality drops.

Questions

Is customer lifetime value the same as LTV?

Yes. CLV, LTV and CLTV are names for the same idea: what a customer is worth over the whole time they buy from you. Tools differ in whether they report revenue or margin and whether they include a forecast, so check the method behind the label, not the abbreviation.

Should customer lifetime value use revenue or gross margin?

Use gross margin. Revenue includes what the goods cost you, so revenue-based CLV overstates what a customer leaves you and makes high acquisition costs look safe. At a 50% gross margin, $76 of revenue per customer is $38 of margin.

What is a good CLV to acquisition cost ratio?

There is no single good ratio. It depends on your fixed costs, which per-order costs you subtract, how you count acquisition cost and the window you use. Check instead that a cohort's margin, after per-order costs, pays back what it cost to win, and how many months that takes.

How long is a customer's lifetime?

As long as you choose to measure, because no customer announces they have stopped. A common choice is 12 months when customers buy often and 24 or 36 months when they buy rarely. Keep the same window so cohorts and years compare.

Can a new store calculate customer lifetime value?

Yes, as historical CLV over the time it has. A store open six months can report 3-month CLV for its first cohorts and compare each new cohort at the same age. A prediction over years from six months of data rests almost entirely on its assumption.