Home › Blog · Attribution · 23 October 2026 · 6 min read

Customer lifetime value by acquisition channel

How to measure customer lifetime value by acquisition channel with monthly cohorts, repeat rate and payback, and how to avoid small-sample traps.

Customer lifetime value by channel is the total revenue that customers acquired through each channel bring in over time, counted per customer. It shows whether a channel that looks expensive on the first order pays for itself through repeat purchases, and whether a cheap one brings customers who never return.

Why the first order is not the whole story

Most ad reports stop at the first purchase. They compare what you spent with what the first order was worth. That is the right view for a business where people buy once.

Many businesses are different. A customer who orders coffee, skincare or pet food may come back several times a year. For those businesses, two channels with the same cost per first order can have very different value. One brings bargain hunters who buy once during a sale. The other brings people who reorder every month.

Lifetime value, often shortened to LTV, measures that difference. Splitting it by acquisition channel tells you where your best customers come from.

The building blocks

Acquisition channel

This is the channel that first brought the customer to you. Each customer has exactly one, and it never changes.

Cohort

A cohort is a group of customers who made their first purchase in the same period, usually the same calendar month. The January cohort is everyone whose first order was in January.

Cohorts make comparisons fair. A customer acquired two years ago has had far more time to reorder than one acquired last month. Comparing their totals directly would flatter the older one. Within a cohort, everyone has had the same amount of time.

LTV at a given age

Take all the revenue a cohort has produced up to a certain age, such as three or six months after the first order, and divide it by the number of customers in the cohort. "LTV at six months" is a precise, checkable figure. A lifetime figure with no time limit is a forecast, and forecasts are easy to get wrong.

Repeat rate

This is the share of a cohort that has ordered at least twice by a given age. It tells you whether LTV is growing because many customers come back or because a few order a lot.

Payback

Payback is the time it takes for the gross margin from a customer to cover what you spent to acquire them. Gross margin is what remains of the revenue after the cost of the goods and of delivering them. The acquisition cost is your ad spend on the channel divided by the new customers it brought.

Why group by first touch

A customer's later orders often arrive through a different route than the first. They type your address directly, click a link in your newsletter, or search for your brand name. If you credited each order to its own last touch, almost all repeat revenue would go to direct traffic and e-mail. The ad that found the customer would get credit for one order only.

Grouping by first touch keeps every later order attached to the channel that started the relationship. That answers the question lifetime value is meant to answer: what is a customer from this channel worth?

It rests on your first-touch record being reliable, which depends on consistent UTM parameters. Our article on first-touch and last-touch attribution covers the two models.

A worked example

Suppose a shop looks at its January cohort for two channels. All figures are hypothetical. Gross margin is 50 percent.

MetaGoogle
Ad spend in January3,000 USD1,600 USD
New customers10040
Acquisition cost per customer30 USD40 USD
Revenue from first orders5,000 USD2,400 USD
Cumulative revenue at 3 months6,500 USD3,600 USD
Cumulative revenue at 6 months8,000 USD4,800 USD
LTV at first order50 USD60 USD
LTV at 3 months65 USD90 USD
LTV at 6 months80 USD120 USD
Customers who ordered again within 6 months25 of 100 (25%)18 of 40 (45%)

On acquisition cost alone, Meta wins: 30 USD per customer against 40 USD. A report that stops at the first order would favor moving budget to Meta.

Over six months the picture changes. A Google customer has brought in 120 USD and a Meta customer 80 USD. Dividing by acquisition cost, Google returns 120 ÷ 40 = 3.0 times its cost in revenue and Meta returns 80 ÷ 30 = 2.67 times.

Now payback, using the 50 percent margin:

  • Meta. Margin per customer is 25 USD at the first order and 32.50 USD at three months. The 30 USD acquisition cost is covered some time before month three.
  • Google. Margin per customer is 30 USD at the first order and 45 USD at three months. The 40 USD acquisition cost is also covered before month three.

Both channels pay back within three months. After that, each Google customer keeps adding more. At six months the margin per customer is 60 USD for Google against 40 USD for Meta. In this example both channels deserve budget, and Google deserves more than its higher acquisition cost suggested.

Pitfalls with small samples

LTV by channel divides your customers three ways: by channel, by month and by age. The groups get small quickly, and small groups mislead.

  • One large buyer distorts the average. Suppose a cohort has five customers and one of them places a 2,000 USD order. The average says this channel brings very valuable customers. Really it brought one. Look at how many customers are behind each number before you trust it.
  • Young cohorts are incomplete. Last month's cohort has had one month to reorder. Compare cohorts at the same age, and treat the newest ones as provisional.
  • Seasonal cohorts behave differently. Customers acquired during a big sale often reorder less than customers acquired at full price. A weak November cohort may say more about the promotion than about the channel.
  • Repeat rate needs a base. Two of four customers reordering is 50 percent. It is also two people. Percentages from small counts swing widely from month to month.
  • Customers must be recognized. A returning customer who checks out with a different e-mail address looks like a new one. Repeat rates are usually a little understated for this reason.

When a cohort is too small to read, combine months into quarters, or look at the channel across all cohorts at the same age. A coarser number built on more customers is more useful than a fine one built on a handful.

How to use it

  1. Start with LTV at three and six months per channel. Those ages arrive soon enough to act on.
  2. Compare each channel's LTV with its acquisition cost, using margin and revenue separately.
  3. Set the first-order return you are willing to accept for each channel based on how its customers have actually behaved. See ROAS vs blended ROAS for the break-even calculation.
  4. Revisit the numbers each quarter as cohorts age.

Where Tagbreeze fits

Tagbreeze records the first touch of each purchase and reports customer value by channel and by monthly cohort, so you can read tables like the one above from your own orders.

In short

  • LTV by channel is revenue per customer over time, grouped by the channel that first acquired them.
  • Use monthly cohorts and compare them at the same age.
  • Read repeat rate and payback beside LTV to understand what is driving it.
  • Check the number of customers behind every figure. Small groups produce confident-looking numbers that mean little.

Know what your ads really bring in.

It takes a few minutes. Nothing is changed on your site until you install the tag.

Add your site
Hands on a laptop beside parcels ready to ship