How to set a CPA target you can actually afford

"What should my cost per acquisition be?" has an actual answer, and it comes from your margins rather than from an industry benchmark.

Benchmarks are useless here. Your competitor's viable CPA depends on their margin, their repeat rate and their overheads, none of which you know.

The basic calculation

Start with contribution margin, not revenue.

Average order value − cost of goods − payment fees − shipping − fulfilment = contribution per order.

If you sell at R1,200, the product costs R450, payment processing is R35, and shipping and packaging come to R110, your contribution is R605.

That R605 is the absolute ceiling for acquiring a customer at break-even on the first order. Everything below it is profit, before overheads.

Where most people stop, and shouldn't

Two adjustments change the number materially.

Repeat purchase. If 30% of customers buy again within a year at the same margin, your true contribution per acquired customer is higher than one order. But be conservative — use actual cohort data, not hope. If you don't have the data yet, don't credit yourself with it.

Overheads. Your contribution margin has to cover rent, salaries, software and everything else before anything is profit. If overheads run R80,000 a month and you do 200 orders, that's R400 per order gone.

Revised: R605 contribution − R400 overhead allocation = R205 genuinely available per order.

That's a very different target from R605, and it's the one that matters.

Setting the working target

Three numbers, not one:

• Break-even CPA — the ceiling. Above this you lose money.
• Target CPA — what you're aiming for. Typically 60–75% of break-even.
• Scale CPA — what you'll tolerate temporarily to grow volume, if cash flow allows.

Having all three lets you make decisions instead of panicking. A campaign at R190 against a R205 break-even isn't failing; it's thin. A campaign at R260 is losing money and needs action today.

Lead generation is harder and more commonly wrong

For lead gen you need the full chain:

Lead → qualified lead → sale.

If 100 leads produce 40 qualified and 8 closes at R6,000 profit each, that's R48,000 from 100 leads, so R480 per lead at break-even.

The critical part: if your close rate drops from 8% to 5% because lead quality fell, your affordable CPA drops by nearly 40% — and the platform has no idea. It's still cheerfully delivering cheap leads.

This is exactly why feeding qualified-lead data back into the platform matters, rather than optimising toward raw form fills.

Three ways this goes wrong

Using revenue instead of margin. "We can pay R400, we sell for R1,200" ignores everything in between.

Ignoring the payback period. A CPA that's viable over 12 months can still bankrupt you if you can't fund the gap for 12 months. Cash flow constrains what margin permits.

Setting a target from a benchmark. Industry averages describe businesses that aren't yours.

Revisit it quarterly

Costs change, prices change, repeat rates change. A target set eighteen months ago is describing a business you no longer run.

Once the target is right, the tracking has to be right too, or you're measuring against fiction. Start with why your conversion tracking is probably wrong.