How cost per conversion is calculated
Take the cost of whatever you are looking at — a keyword, an ad group, a campaign, the account — over a date range, and divide it by the conversions recorded against it in the same range. Google Ads shows this as Cost / conv.; Meta Ads Manager shows a cost per result for whichever result the campaign was set up to chase.
The arithmetic is trivial. The meaning is not, because the divisor is a reported number rather than an observed one. Everything that shapes the conversions column — which actions you chose to count, whether repeats count again, how long the conversion window stays open, how much of the total is modelled — flows straight into this figure. A campaign can halve its cost per conversion by counting an easier action, without a single extra customer.
Why cost per conversion matters
It is the number that turns advertising into a decision. If you know what a lead or an order is worth to the business, you know whether the price you are paying for one is sustainable, and that single comparison settles most arguments about budget, bids and which campaigns deserve to live.
It also behaves predictably as you scale, which makes it a good early warning. Spend usually buys cheaper conversions first, so a campaign pushed harder will see this figure rise. Watching when it crosses your limit tells you where growth stops being profitable, long before the bank balance does.
Where cost per conversion goes wrong
The most damaging error is treating it as cost per customer. If the tracked conversion is an enquiry, this is the cost of an enquiry, and only a share of enquiries become customers. Multiply that gap by your close rate before you compare the figure with what a customer is worth, or you will happily fund campaigns that lose money.
Small numbers are the other trap. On a low-volume campaign, one conversion arriving or not arriving swings the figure violently, and people react to noise as though it were a trend. It is also unstable near the end of a period: conversions arriving late are credited back to the day of the click, so recent figures improve after the fact, and long consideration cycles make the effect worse.
What to do about it
Set a ceiling before you look at the report, worked back from what the sale is worth and how many enquiries it takes to make one. Compare against that ceiling rather than against another business, another industry or a figure someone quoted you, because none of those share your margins or your counting rules.
Then read it in context. Judge it over a window long enough for your sales cycle, alongside conversion volume and conversion rate, and check whether a cheaper campaign is genuinely more efficient or simply counting something easier. If the cost is above target, remember the landing page moves it just as surely as the bid does.
