CPA Calculator

How it is calculated

Cost per acquisition is spend divided by conversions: CPA = spend ÷ conversions. Broken down the chain, CPA = CPC ÷ conversion rate.

That second form is the useful one, because it shows the two levers: the price of a click and the proportion that convert.

The profitable threshold is the contribution per order, spend more than that acquiring a customer and the order loses money.

The same thing with real figures

$4,000 of spend producing 147 conversions is a $27.21 CPA.

Via the chain: a $0.95 CPC divided by a 3.5% conversion rate is $27.14. The small difference is rounding.

Against $31.90 of contribution per order, that leaves $4.76 of profit per acquisition, thin but positive.

Improve conversion to 4.5% and CPA falls to $21.11, leaving $10.79. Improve CPC to $0.75 instead and CPA falls to $21.43, leaving $10.47. Either lever produces roughly the same result, and the conversion route also improves every other channel's economics at once.

The catch

CPA counts a conversion as a conversion regardless of what it was worth. An account acquiring cheap customers who buy small baskets can show an excellent CPA and generate less contribution than one with a worse CPA and higher order values.

It also ignores repeat purchase entirely. For a business with genuine retention, the right threshold is contribution over the customer's lifetime rather than on the first order.

Applying it

Compare CPA against contribution per order, not against a benchmark. The benchmark describes someone else's margin structure and says nothing about whether your number works.

Then segment it by campaign type, because prospecting and retargeting have completely different CPAs and blending them hides which is doing the acquisition.

Where the real leverage sits

Conversion rate is the most powerful term in the equation, because it divides. Moving from 3.5% to 4.5% is a 29% relative improvement and cuts CPA by 22%, and it does so on every channel simultaneously rather than only on the one being optimised.

It is also the term most within your control: page speed, clarity of the offer, trust signals, checkout friction and mobile experience are all decisions rather than auction outcomes.

Advertisers who spend their time in the bidding interface and none on the landing page are optimising the term they control least. The unglamorous work of removing a checkout step or halving page load time routinely does more for cost per acquisition than a quarter of campaign restructuring.

Track cost per acquisition for first-time customers separately from the blended figure. Blended CPA falls automatically as a business accumulates returning customers, which makes acquisition look like it is improving when nothing about it has changed.

New-customer CPA is the figure that measures whether the advertising is working, and it is usually meaningfully higher than the blended number quoted in reports.

Most platforms can report new-customer conversions natively once the data is passed to them, which makes the split a configuration task rather than an analysis one.

It is worth distinguishing the platform's reported cost per acquisition from the true one. The platform counts conversions it can attribute and divides by its own spend, which excludes agency fees, creative production and any tooling. A blended figure including those is usually 15% to 30% higher, and it is the number that has to be compared against contribution margin rather than the one on the dashboard.