Customer Lifetime Value You Can Spend Against
A customer lifetime value figure earns its keep in exactly one way: it tells you what you can afford to pay for the next customer. If the number on your dashboard cannot survive being put next to an acquisition cost and a date, it is not a metric — it is a morale exercise.
Three things quietly stop most LTV numbers from being spendable. They are built on revenue instead of contribution, they are divided by the customers who came back rather than everyone you paid for, and the word “lifetime” leaves them with no end date, so the money they promise never has to arrive. Fix those three and you get a figure you can hand to whoever runs the ad account.
What the number has to answer
Strip away the definitions and lifetime value exists to set a ceiling. Below that ceiling, acquiring another customer makes you money; above it, it costs you money. Everything else — the cohort maths, the windows, the segment splits — is machinery for making that one comparison honest. Our glossary definition of customer lifetime value covers the short version; this is the part that decides whether you can act on it.
That framing also tells you immediately which version of the number is wrong for the job. “Our customers are worth a lot eventually” cannot be compared to a bill you pay this month. “A customer acquired in March returns this much contribution by June” can.
Break one: the number counts revenue, not what you keep
Revenue-based LTV is the most common version because it is the easiest to pull — total sales divided by customers, done. It also overstates your acquisition ceiling by however much it costs to fulfil an order, which for most stores is the bulk of it.
Between the price a customer pays and the money that stays in the business sit a predictable stack of lines. Cost of goods first, then the payment processing that comes off every single order, then the fulfilment and last-mile shipping you subsidise, then the discounts that were needed to win the order, then the share of orders that come back as returns or refunds. What survives all of that is contribution — and contribution is the only version of lifetime value you can responsibly spend against.
The gap between the two versions is not a rounding difference, and it widens with every repeat order because the deductions repeat too. This is the same distinction that decides whether an ad campaign is actually profitable — we walked through it in detail in the piece on break-even ROAS, and the arithmetic is identical here.
Break two: the denominator drops everyone who never came back
Ask for lifetime value and most reporting hands you profit from repeat customers divided by the number of repeat customers. That is a real number, just not the one that sets an acquisition ceiling: it has deleted every customer who ordered once and left, and you paid to acquire those too.
The correct denominator is the whole acquisition cohort: everyone you paid to bring in during a given period, including the majority who never ordered a second time. Dividing by the survivors tells you what a good customer is worth. Dividing by everyone you acquired tells you what a customer is worth on average, which is the only figure that can sit opposite a blended acquisition cost.
If this feels familiar, it should. It is the same denominator problem that makes a single ecommerce conversion rate unusable: the metric is fine, but nobody agreed on what sits under the division line, so two teams quote two numbers and both are right. Write the denominator into the definition of the metric and the argument goes away.
| Denominator | What it tells you | What it must not be used for |
|---|---|---|
| Customers with two or more orders | What a retained customer is worth once you have one. | Setting an acquisition budget — it excludes the people you paid for and lost. |
| Everyone acquired in the cohort | Average contribution per customer bought. | Judging your loyalty programme — one-time buyers drag it down by design. |
| Everyone acquired, split by channel | Which acquisition source buys customers worth keeping. | Channels with too little volume to be read as anything but noise. |
Break three: “lifetime” has no date on it
An open-ended lifetime value assumes every order a customer might ever place, which is a forecast wearing the costume of a measurement. It cannot be budgeted against: advertising is paid for monthly and paid back over years.
The fix is a time box. Pick a window, measure contribution from an acquisition cohort inside that window only, and state the window every time you quote the number. Ninety days, one hundred and eighty days and one year are the usual choices, and the right one is set by your repeat purchase interval: consumables and beauty refills reveal their pattern quickly, furniture and high-ticket fashion do not. If you sell something bought twice a year, a ninety-day window will tell you your customers are worthless.
Time-boxing also makes the number move. A March cohort can be compared to a June cohort at the same age, which is the cleanest read you will get on whether retention work is doing anything; an all-time average dilutes every improvement in years of history.
How to calculate customer lifetime value so it survives a budget meeting
Four inputs, one window, no averages of averages.
| Input | What it must be | Common mistake |
|---|---|---|
| Cohort | Every customer whose first order fell in one acquisition period. | Using “all customers”, which mixes cohorts of different ages. |
| Window | A fixed number of days measured from each customer’s first order. | Measuring to today, so older cohorts look better purely for being older. |
| Value per order | Contribution after goods, processing, fulfilment, discounts and returns. | Using order revenue, or gross margin that ignores per-order costs. |
| Divisor | The count of customers acquired in the cohort, repeat or not. | Dividing by repeat customers only. |
Multiply the average number of orders a cohort places inside the window by contribution per order, then divide by everyone in the cohort. If your analytics has a cohort report, it will do the grouping for you; if it does not, an order export with two columns — the customer’s first-order date and each order’s date — is enough to build it in a spreadsheet. Our free customer lifetime value calculator runs the same arithmetic if you would rather sanity-check a figure than build the model.
Then do the only thing the number was built for: divide it by what it costs to acquire one customer. Above one, the window pays for itself. Below one, you are financing growth out of something other than the customers.
The identity problem nobody accounts for
Every method above assumes one human equals one customer record. In practice a repeat buyer fragments: a work email in January and a personal one in April, one order through your store and one through a marketplace listing. In the markets we build for most, an order also arrives over WhatsApp and gets keyed in by hand, and cash-on-delivery orders attach to a phone number rather than an address book entry.
Every one of those splits turns a second order into a new first order, so lifetime value looks flat while retention is genuinely improving. Before concluding that retention is broken, match customer records on phone number as well as email and check how many “new” customers share an address with an existing one. It is usually the cheapest measurement fix available.
What actually moves the number
Lifetime value has exactly two levers — how much a customer contributes per order, and how many orders they place inside your window — and discounting moves the first one the wrong way to buy the second.
The second order is where the leverage sits, because the gap between one order and two is the widest gap in the whole curve. That is a job for the weeks immediately after the first delivery, not for a quarterly campaign: the post-purchase sequence, a reason to come back that is not a blanket discount code, and a reorder path short enough that it does not need one. A Shopify loyalty app can systemise that, and a referral programme makes existing customers cheaper to buy the next ones with, but both are amplifiers — they multiply a second-order rate that already exists rather than creating one.
The first lever is the same work as raising average order value without losing margin — bundles that trade on convenience rather than price, thresholds set from your own margin — and contribution earned that way compounds across every order in the window.
Frequently asked questions
What is a good customer lifetime value?
There is no absolute figure, and any benchmark quoted without a window and a denominator is unusable. The only meaningful form is a ratio: contribution per acquired customer inside your window against what it costs to acquire one — compared to your own previous cohorts, not to an industry average.
Should lifetime value use revenue or profit?
Contribution — revenue less cost of goods, payment processing, fulfilment, discounts and returns. Revenue LTV consistently overstates your acquisition ceiling, and the overstatement grows with each repeat order because the costs repeat alongside the revenue.
How long should the window be?
Long enough to contain a typical second order and short enough that you can act on it. Match it to your repeat purchase interval rather than copying a number: fast-moving consumables read clearly at ninety days, considered purchases usually need a year.
Can a store under a year old calculate lifetime value?
Yes, with a shorter window and honest labelling. A ninety-day cohort figure from a young store is a real measurement; extrapolating it into an annual “lifetime” figure is not.
The takeaway
Customer lifetime value stops being decoration the moment it carries three things: contribution rather than revenue, a denominator that includes every customer you paid for, and a window with a date on it. Rebuild the number that way, compare it to acquisition cost at the same age, and it starts answering the question it was always supposed to answer — how much the next customer is worth buying.
If you would rather have someone read your store’s numbers with you, the free Shopify store audit covers the tracking and checkout setup that has to be right before any of this is measurable.
Current store performance