Ecommerce LTV: how to calculate it and use it to buy ads better

How do I calculate the lifetime value of my customers, and how do I use it to decide how much I can spend to acquire one?

Lifetime value is the number that tells you how much a customer is worth over the whole relationship, and therefore how much you can afford to pay to win one. It is also the metric most often calculated in a way that makes it look larger than it is. This guide gives two ways to calculate it with your own store's data, one worked example, and how to turn it into a ceiling for acquisition spend.

What LTV means for someone buying ads

LTV (customer lifetime value) is the value a customer produces over their relationship with your store. For ad decisions, three choices make it usable:

• Gross profit, not revenue. Compared with an acquisition cost, revenue LTV ignores the cost of the products and approves acquisition that loses money.
• A fixed horizon, not a lifetime. A 12-month LTV is a number you can measure and act on; an open-ended lifetime is a forecast.
• Settled orders only. Cancelled and refunded orders never produced value.

Two ways to calculate it

The quick formula

LTV ≈ average order value × orders per customer over the horizon × gross margin.

It is fast and it is usually too high, because the averages come from your whole customer base, where your oldest and most loyal customers carry the most weight. Fed with store-wide averages of $1,000 per order and 3 orders per customer at a 35% margin, it says $1,050.

The cohort method

Take the customers whose first order fell in one month. Add up the gross profit of every settled order they placed, from the first one to the end of the horizon, and divide by the number of customers in the cohort. It uses only what these customers actually did, which is why it is the one to trust.

A worked example

An illustration with round numbers, from the same store as our new-customer CAC example: 700 new customers in September, a first order of $1,050 on average, repeat orders of $900 on average and a 35% gross margin.

The cohort's 12-month gross-profit LTV is about $541, half of what the quick formula said. Both used the same margin. The difference is that the cohort only counts what these 700 people did, while the store-wide averages mix in years of the most loyal buyers.

Turn it into a ceiling for acquisition

Maximum CAC = LTV ÷ your target LTV:CAC ratio. A common rule of thumb is 3:1, meaning a customer should return three times their acquisition cost in gross profit over the horizon. Treat it as a starting point, not a law.

In the example, 3:1 on an LTV of $541 sets a ceiling of about $180. The store's new-customer CAC is $214, so its ratio is about 2.5:1. The first order's gross profit, $367.50, is already above the CAC, so each customer pays back their acquisition on the first purchase; what the ratio says is that the profit left after that is thinner than the target.

Two more ways to use it:

• By acquisition channel. Group cohorts by the channel your attribution credits for their first order. A channel with a higher CAC can still be the better one if its customers come back more.
• In bidding. Push the platforms towards customers worth more over time, not only towards the first order. Google's customer acquisition goal, for example, has a mode that prioritizes bidding towards new customers (Google Ads Help); how hard to push is a decision your LTV should make, not a guess.

Mistakes that distort LTV

• Revenue instead of gross profit. The most common one, and it inflates LTV by the whole cost of the products.
• Refunds left in. An order refunded in full produced no value, and a customer whose only order was refunded is not a customer.
• Split or merged identities. When the same person buys as a guest and later with an account, a system that cannot link the two sees two customers with one order each, which lowers LTV. The opposite error, merging different people into one, inflates it. We cover both in why platforms disagree about who is a new customer.
• Young cohorts read as finished. A cohort that is two months old has a two-month LTV. Label it that way; multiplying it up to a year assumes a repeat pattern you have not seen yet.
• Only the average. A few very large buyers can pull the mean far above the typical customer. Look at the median next to it.

LTV and CAC are two halves of one decision, and they only agree with each other when both come from the same customer records. Read this next to how to calculate new-customer CAC and, for the channel view, ROAS, POAS and MER. Adray keeps each order's new-or-returning status from the store's full order history, which is the starting point for grouping customers into first-order cohorts.