The LTV:CAC ratio divides what a customer is worth by what winning them cost: average value per customer within a set number of days, over marketing spend divided by new customers. Measure the value without VAT, after refunds, and in margin: CM1 after product costs, CM2 after shipping and payment fees too. On CM2, 1× is break-even.
The LTV:CAC ratio divides what a customer is worth by what it cost to win them: lifetime value over customer acquisition cost. For an online store, measure the value without VAT and after refunds, and in margin as well as in revenue, because only margin can pay back an ad. On the figures below one customer reads 2,37× on revenue and 1,20× on CM2: the same customer, a very different verdict. This guide gives the formula, works it through three ways, and sets out the rules nouz follows on its LTV tab, which measures customer lifetime value from a store's own orders.
In short
- LTV:CAC is the average value per customer within a window of days, divided by CAC, the marketing spend divided by the new customers it won.
- Value has three readings: revenue without VAT, CM1 after product costs, and CM2 after shipping, packing and payment fees too. Only CM2 can pay for an ad.
- Refunds come off on the day they happen: their value from all three readings, their return costs from CM2.
- On CM2, 1× is break-even: the customer has paid back what winning them cost, and nothing yet towards rent or salaries.
The formula
LTV:CAC is two averages over the same group of customers. LTV is everything those customers were worth within a set number of days of their first order, less their refunds, divided by how many of them there are. CAC, the cost of winning a customer, is the marketing spend of the days they were won on, divided by the same customers. The ratio is the first over the second, and it only means something when both halves describe the same people.
| Figure | Formula |
|---|---|
| LTV | (value of their orders − their refunds) within the window ÷ customers |
| CAC | marketing spend on the days they were won ÷ the same customers |
| LTV:CAC | LTV ÷ CAC |
Without cohort data the value comes out of averages: orders per customer times what one order is worth. Orders per customer are the first order plus the repeat orders of the customers who come back. At a 32% repeat purchase rate, the share of new customers who order again, and 2,2 more orders each, a customer places 1,70 orders on average. That shortcut is fine for a first estimate. The honest version is cohort analysis: follow the customers of one month forward instead of averaging every customer the store ever had, because a store that grew fast last quarter is full of customers who have not had time to come back yet.
One customer, three ratios
Take the figures the LTV calculator opens with. An order is worth €85,00 without VAT. Its products cost €32,30, which leaves €52,70 of CM1, the margin after goods. Shipping and packing cost €6,40 and the payment fee €1,62, which leaves €44,68 of CM2. A customer places 1,70 orders in their first year, and winning one costs €55,00.
| Measure | LTV | LTV:CAC | LTV after refunds | LTV:CAC after refunds |
|---|---|---|---|---|
| Revenue | €144,84 | 2,63× | €130,36 | 2,37× |
| CM1 | €89,80 | 1,63× | €80,82 | 1,47× |
| CM2 | €76,13 | 1,38× | €65,89 | 1,20× |
Read the three ratios together. On revenue the customer is worth 2,63× what they cost to win; on CM1, 1,63×; on CM2, the money left after everything the orders themselves cost, 1,38×. All three are true, and only the last says how much of the acquisition the customer has actually paid back. The first order alone does not: before refunds it covers 81% of the CAC, and the repeat orders do the rest. CM1 and CM2 are the first two rungs of the contribution margin ladder, set out in CM1, CM2 and CM3 explained.
Refunds come off, on the day they happen
A refund is part of what a customer was worth. Say a 10% share of these orders is refunded in full and the goods come back to stock. Each refund takes its value without VAT off all three readings. The product cost comes back to CM1 and CM2, because the goods can be sold again. The return label and handling, €7,45 a return, come off CM2. The outbound parcel and the payment fee stay spent; no refund gives them back. After refunds the customer is worth €130,36 in revenue, €80,82 in CM1 and €65,89 in CM2, and the CM2 ratio falls from 1,38× to 1,20×. The first order now covers 70% of the CAC.
Two details decide how much a refund moves the ratio. It counts on its own day, not on the order's, so a customer's value can fall from one month to the next when a return lands late. And a product that never comes back keeps its cost: a refund for a parcel lost in transit takes the revenue off and gives no product cost back, so on a store whose refunds are mostly goodwill, CM1 falls by the whole refund.
CAC from the days with spend
CAC has a trap of its own: the spend and the customers have to come from the same days. Many stores connect their ad accounts after months of selling, so the first synced days hold orders and new customers but no spend. Divide the spend you have by every new customer you have and CAC comes out too low, which makes the ratio look better than it is. nouz counts only customers won on days it holds marketing spend for, from the first day with any, and divides the spend of exactly those days by exactly those customers.
Marketing spend means the whole marketing line of the statement: the ad platforms' spend, CSV imports, Shopify's Shop Campaigns where Shopify reports them, and any marketing spend entered by hand, such as an agency fee or an influencer budget. Nothing is attributed to a channel, so no platform can claim a customer twice. How each version of CAC is built, blended against paid, is the customer acquisition cost formula guide.
Which customers count, and over how many days
The ratio is only fair when every customer it averages has had the same time to be worth something. A customer won yesterday has one order; a customer won a year ago has had a year to come back. So the LTV tab reads value within a window of days from each customer's first order: the longest of 365, 180, 90 or 30 days that enough customers have lived through, at least 10 of them and at least 10% of the customers counted. It names the window it used, and it averages only over the customers old enough for it, so CAC is read from the spend of the days those customers were won on.
Which customers count is a choice of its own, made in the tab's own dates, Customers won: the customers whose first order falls in the days you pick. Their value keeps counting after those days end, up to today. Pick the last 90 days to judge this quarter's acquisition, or All time, the default, to see every customer since your first synced day. Guest orders, with no customer to follow, are left out. The help centre's answer on how to read the LTV tab goes through it card by card.
What is a good LTV:CAC ratio?
On CM2, anything above 1× within the window means the customer has paid back what winning them cost out of what their orders left. That is the floor, not the goal: the overhead, rent and salaries among it, still has to come out of what is left above it, so a store with a heavy fixed block needs more. How far above 1× a store needs to be depends on its cash as much as on its margin. A store buys its stock before the second order arrives, so a customer who pays back in month eleven needs eleven months of financing behind them. The payback period, the month the cumulative contribution crosses CAC, is the second number to read beside the ratio, and the CAC payback calculator estimates it from your own averages.
The 3:1 that is often quoted is a rule of thumb from subscription software, usually measured on revenue or on gross margin over a whole lifetime. On the customer above, after refunds, a revenue ratio of 3,00× is 1,52× on CM2, and the revenue ratio at which CM2 merely breaks even is 1,98×. Translate a benchmark into your own margin before you hold yourself to it.
Five ways the ratio flatters
- Revenue in place of margin: a revenue ratio well above 1× can lose money on every customer when products, parcels and fees take more than half of each order.
- VAT left in: the value has to be without VAT, or a German store credits every customer with 19% that belongs to the tax office.
- Customers older than the spend: dividing today's spend by customers won before the ad accounts were connected makes CAC look cheap.
- One average across every cohort: customers won last month sit beside customers who have had two years to come back.
- Refunds left in: a customer who returned half of what they bought is worth half of it, and a ratio built on order totals never notices.
Where nouz shows it
On Insights, the LTV tab carries an LTV:CAC card with three ratios side by side: Revenue LTV:CAC, CM1 LTV:CAC, also called LTGP:CAC, and CM2 LTV:CAC, also called contribution LTV:CAC, with 1× as the break-even. The card names its window and the days its customers were won on. The cohort table has an LTV:CAC view that sets each first-order month against its own CAC, and the value chart draws CAC as the line the average customer's value has to cross. Until enough customers are old enough for the shortest window, the card names the day it will appear; without marketing spend for the dates, it says so.