KPIHelm — Your KPIs. Our insight. Your future.
All articles

How much can I afford to pay for a new customer on Shopify?

No more than the margin that customer brings you — the first order plus what they buy again — within a window you can finance. Measure it per channel, and count new customers from your store, not from the ad platforms.

Two guides in this series have now stopped at the same sentence: paying more for a customer than the first order returns is legitimate — if it was a decision, and you know your repeat rate. This one is about that number, and about the one it decides: what you can actually afford to pay for a customer. The reason it needs its own guide is that acquisition cost and ROAS answer questions on different clocks. ROAS asks what an ad returned this month. Acquisition cost asks what a customer costs to buy and how long they take to pay you back — and a store can hold a perfectly steady ROAS while the cost of every new customer climbs underneath it. If what an order leaves you is still unclear, start with the break-even guide; the margin it produces is an input here.

Step 1: Cost per customer is not one divided by ROAS

Acquisition cost is ad spend divided by new customers. The common shortcut — spend divided by orders — is a different number wearing the same name, and it fails in a specific direction: every repeat order makes acquisition look cheaper. The better your retention, the more flattered the figure, which is exactly backwards for a metric you use to decide how hard to push growth.

Take new-customer counts from your store, not from what the ad platforms report. They are counting claimed conversions, not first-time buyers, and the gap between those two is a subject of its own. Shopify already knows who had never ordered before.

Keep two versions, and never average them:

  • Blended acquisition cost — all ad spend divided by all new customers, whatever brought them. This is the business-level number: what growth costs you in total, including the customers who arrived without an ad.
  • Paid acquisition cost per channel — a channel's spend divided by the new customers your store attributes to that channel. This is the number a budget decision needs.

The first is almost always lower, because organic, referral and returning-visitor sales dilute it. Quoting the blended figure while making a channel decision is the most comfortable mistake in this whole guide.

Step 2: A rising acquisition cost hides inside a flat ROAS

The two ratios are separated by average order value and by the mix of new against repeat buyers, so one can move while the other doesn't. Put numbers on it:

  • Month A: 10,000 spent, 30,000 in sales → 3.0x. 200 new customers → 50 per customer.
  • Month B: 10,000 spent, 30,000 in sales → 3.0x. A bundle lifted average order value, so the same revenue came from fewer buyers: 150 new customers → 67 per customer.

Identical ROAS, acquisition a third more expensive, and nothing in the ads dashboard moved. It runs the other way too: a discount raises order count and drops ROAS while making each new customer cheaper. Read either ratio alone and you get a confident answer to a question you didn't ask.

Step 3: Measure the repeat rate by cohort, or it will flatter you

"About 30% of our customers order again" is a number worth distrusting. It pools customers who have had two years to come back with customers who bought last week — so it drifts upward as your store ages, and it cannot tell you whether the people you are buying now behave like the people you bought last year.

The honest version groups customers by the month of their first order and gives every group the same fixed window:

  1. Group new customers by the month they first bought.
  2. Pick one window — 90 days is a reasonable default — and apply it to every cohort, including the young ones.
  3. Within that window, add up what those customers left you: the margin on the first order plus the margin on anything they bought again. Margin, not revenue — the figure from the break-even guide.
  4. Divide by the number of customers in the cohort. That is contribution per new customer at 90 days.
  5. Put it next to that cohort's acquisition cost.

Contribution of 60 against a paid cost of 50 means the channel paid itself back inside the window. Against 70, it didn't — you are financing the difference, and whether that is an investment or a leak depends entirely on whether you decided to do it.

Step 4: The window has to be one you can finance

Two stores with the same eventual customer value are not in the same position if one recovers its acquisition cost in six weeks and the other in nine months. You pay the ad platform now; the customer pays you back later. A long payback window is a loan you are writing to yourself, and the length is set by your cash position, not by whichever number makes the campaign look affordable.

Three honest limits belong here, because each one turns this arithmetic into wishful thinking when ignored:

  • Some categories genuinely have no second order. If people buy once, there is no repeat rate to lean on and the first order has to carry the whole cost. That is a constraint, not a failure.
  • A repeat rate propped up by discounts buys a number, not a customer. If the second order only happens at 30% off, check whether it was profitable at all before counting it as payback.
  • A cohort's future is a prediction. Use the window you have actually observed. An extrapolated lifetime value is a forecast with your ad budget riding on it.

Step 5: What to read, and what to act on

Once a month, per channel: new customers from the store, spend divided by those customers, contribution per new customer at your fixed window, and the ratio between the last two. Then the blended figure for the business-level question.

Read the trend, not the level. There is no universally good acquisition cost — there is only yours, against your margin and your payback window, compared with its own history. A cost that rose 30% over a quarter is a finding whatever its absolute value; a cost that is high and stable in a category with strong repeat purchase may be entirely fine.

When it moves, look at what actually changed before you touch budgets: the mix of channels, the offer, the audience, or the market itself. Then do what every guide in this series ends with — change one thing, and judge it by how much new data has accumulated rather than how many days have passed.

KPIHelm watches this drift for you. Rising acquisition cost is one of the signals it reads per platform, against that platform's own history, so a cost creeping up on one channel is named while it is still a trend rather than a quarter you explain afterwards. Install it free on the Shopify App Store or see a sample report.