What is a good CAC, LTV:CAC, and payback period?

CAC in this article is customer acquisition cost. What you spend to get a first order. Not a medical “CAC score.” If you arrived from a calcium or blood-test result, you are on the wrong page.

For fashion and beauty ecommerce, the judgment questions travel together. What is a good CAC? What is a good LTV:CAC? What is a good payback period? A 3-month payback can be plausible for a beauty brand that replenishes and tight for a fashion brand whose second order is a rumour. A 3:1 LTV:CAC is SaaS folklore, not a law of jackets or serums.

This piece defines the three numbers in operator English and shows why replenishment and second-order behaviour change the clock. The missing part is your months-to-contribute by first product and channel. Averages hide the cohort.

What CAC means in ecommerce

CAC is acquisition cost divided by new customers (or new-customer orders, if that is how you run the P&L). The fight is always the numerator.

Decide, and write it down:

paid media only, or paid plus agency fees, tools, and influencer
new customers only, or blended with returning (blended is not CAC)
over which date range, and whether a sale sat inside it

CPA (cost per acquisition or cost per order) is a short aside. Teams use CPA for any converting event and CAC for a new customer. If your “CAC” includes returning purchasers, you are looking at a blended efficiency number and calling it by the wrong name. That will make LTV:CAC look healthier than the acquisition engine is.

Fashion and beauty both break naive CAC. A shopper who buys /products/linen-shirt at 30% off and never returns is a completed acquisition with a short, often ugly, contribution story. A shopper who buys /products/vitamin-c-serum at full price and repeats in week ten is a different asset. Same dashboard row if you only watch blended CAC.

Treat CAC as the cost of a first relationship, then immediately ask what that relationship is worth. A “good CAC” without a product and channel next to it is a slogan.

Judgment, not a league table. If first-order contribution after returns cannot cover a large share of CAC, you are funding a second order you have not proven. That can be a deliberate bet. It should not be an accident you discover in month six.

LTV:CAC is a ratio, not a finish line

LTV (or CLV) is a forecast of what a customer will contribute over a window you choose. The ratio LTV:CAC asks whether that forecast is large enough relative to what you paid to start.

Three problems show up immediately.

The window. Twelve-month LTV on beauty replenishment can be meaningful if the repurchase cycle is real. Twelve-month LTV on fashion can be a hope dressed as a model. Ninety-day LTV is more honest for many apparel brands and will look “worse.” Honesty is the point.

The definition of value. Revenue LTV is vanity if returns, discounts and shipping eat the order. Contribution LTV is the number that can sit next to CAC. If your LTV is “average order value × purchase frequency × a margin you rounded,” you have a slide, not a ratio.

The mix. Customers acquired on branded search and email have a different LTV than customers acquired on cold paid with a first-order code. One blended LTV:CAC will flatter the cold channel with the warm channel’s repeat.

The 3:1 rule you see in investor threads is a SaaS-shaped habit. Ecommerce inventory, returns and markdown do not behave like software. A fashion brand at 1.5:1 on contribution LTV over a year may be healthier than a brand at 4:1 on revenue LTV that ignored returns. A beauty brand under 3:1 on a hero SKU with a proven eight-to-twelve-week repurchase can still be a good engine.

Use the ratio to ask “are we buying customers who pay back.” Do not chase a universal multiple.

Payback period: 3 vs 6 vs 12 months

Payback is how long until contribution from that customer (or cohort) covers CAC. It is a cash and risk number. LTV:CAC can look fine on a three-year story while payback is a year you cannot fund.

Judgment ranges operators actually argue about:

About 3 months. Often discussed as “good” for ecommerce. Plausible when the first order covers a serious share of CAC, or when a beauty replenishment lands inside the window. Tight for fashion if the first order is discounted and the second order is slow.
About 6 months. Common and sometimes healthy for brands with a real second purchase and enough cash to wait. Still a problem if you are inventory-heavy and the style will markdown before month six.
About 12 months. A long wait in fashion and beauty unless you have the balance sheet and a retention story you can show in cohorts, not in a model.

“Is a 3-month CAC payback good for a UK beauty brand that replenishes?” is the right shape of question. The answer depends on whether replenishment actually happens on that first SKU, at a price that still contributes, without a discount that resets the clock.

Payback is not the discounted-payback-period exam question. It is “when does this cohort stop being a receivable we hope for.”

Beauty replenishment changes the clock

Beauty has a physical clock. A 30ml serum, a moisturiser, a cleanser. Many of them have a use-up window you can estimate in weeks, not in marketing poetry.

That does not make every beauty customer a subscriber. It means:

a first SKU that is the daily driver is a different CAC bet than a kit or a gift-with-purchase
a 3-month payback can be reasonable if the second order is the same SKU, not a new acquisition
a “good” CAC on a heavily gifted first order can be a bad CAC if the second order never happens at full price

Shade adds fashion-like uncertainty. Colour cosmetics can return, sit unused, or fail to replenish even when skincare does. Do not apply a replenishment story to the whole beauty catalogue.

If you cannot name the median days-to-second-order for your top three first products, you do not yet have a beauty payback model. You have a blended CAC and a hope that “beauty retains.”

Fashion second order is a different problem

Fashion’s problem is often not “will they replenish the same unit.” It is “will they buy anything else before we have to forget them.”

A second order on a jacket customer might be 90 days later, a season later, or never. Size, returns and markdown sit on both orders. A payback story that assumes a second dress at full price is a story. Measure the actual second-order rate by first product and by whether the first order was on sale.

What that means in practice:

CAC must do more work on the first order in fashion than in replenishing beauty
a 3-month payback often requires a stronger first-order contribution, not a loyalty programme slide
“good LTV” that lives in year two is not a number you can spend against this quarter unless cash and inventory agree

None of this says fashion cannot retain. It says the mechanism is assortment, fit, brand and whether you trained them to wait for /collections/ to go on sale. Not a 60-day empty-bottle reminder.

First product and channel hide inside the average

Averages hide the cohort. That is the whole diagnostic hint, without handing over a method.

Two customers can share a store-level CAC and have opposite paybacks:

cold paid + first-order 20% off + fashion novelty SKU
branded or email + full-price beauty hero SKU

The dashboard will let you manage the blend. The P&L is the split.

Prioritise, even in a spreadsheet you already have:

CAC and first-order contribution by channel
second-order rate and days-to-second-order by first product
payback for those cells, not for “the brand”

If one cell is paying back in 10 weeks and another is never paying back, you do not have a CAC problem. You have a mix problem that looks like CAC.

Returns belong in the same grid. A fashion first order that converts and then returns 40% of the time has a higher true CAC than the ads manager showed. Beauty is usually kinder here, except where shade and “try seven foundations” behaviour shows up.

CAC vs CPA: keep the labels clean

CPA is what you paid per result the platform counted. CAC is what you paid per new customer you are willing to put in a cohort. When someone asks “what is a good CAC” and answers with a Meta CPA, they have changed the question.

Use CPA to operate the auction. Use CAC to operate the P&L. If the two diverge, the platform is either counting returning customers, counting view-throughs you do not want, or missing costs you put in the CAC numerator.

You do not need a new formula page. You need the labels to stay still for a quarter so payback means the same thing in May as it did in March.

What's missing

Definitions and category clocks can tell you whether a 3-month payback is even a sensible target, and why beauty replenishment and fashion second order should not share a rule. They cannot tell you your months-to-contribute by first product and channel. That split is the investigation. The average will not do it.

If you want that investigation, Explore an opportunity with TwoKai.