Cost Per Lead vs Cost Per Acquisition, and Why It Matters
You have probably been reporting one and having it heard as the other. Almost everybody has, at some point, and the reason it matters is not pedantry.
Cost per lead counts people who raised a hand. Cost per acquisition counts people who bought. Between the two sits every unqualified inquiry, every deal that did not close, and every quotation nobody accepted.
The person reading your report is usually thinking about the second number while you are reporting the first. That gap is not discovered gradually. It is discovered all at once, in a meeting, when somebody divides the spend by the actual sales.
The two numbers, defined
Cost per lead: spend divided by leads. A lead is somebody who took an identifying action, meaning a form, a call, a chat, a demo request.
Cost per acquisition: spend divided by customers. Somebody who bought.
The ratio between them is your lead-to-customer rate, and if you do not know that rate, you do not actually know what your marketing costs. That is the practical finding of this page, and it is a number that lives in the sales system rather than in the ads platform.
Why the confusion is so common
The platforms encourage it. Ad platforms report conversions, and a conversion is whatever you told them to count. In most lead-generation accounts, that is a form submission, which is a lead. The platform then labels the resulting figure cost per conversion or cost per acquisition, using the vocabulary of a purchase for a metric counting an inquiry.
The vocabulary is genuinely inconsistent across the industry. Cost per acquisition sometimes means cost per customer and sometimes means cost per action, which is a different thing entirely.
And the two systems do not talk. Leads live in your ad platform, customers live in a sales system, and joining them requires somebody to do it deliberately. Where that join has not been built, the ad platform's number is the only number available, and it becomes the reported one by default rather than by decision.
The failure this produces
Follow it through, because it is worth seeing as a sequence.
Marketing reports a cost per lead and calls it cost per acquisition. Everybody reads it as the cost of a customer. The number looks good.
Then someone in finance divides total marketing spend by actual new customers and gets a different number. Not slightly different: different by whatever your lead-to-customer rate is.
Nobody lied. The label was wrong, the label was inherited from a platform, and the correction lands at the worst possible moment, in a meeting nobody prepared for, with the marketing person looking either careless or evasive.
Fixing the labeling before that happens costs about an hour.
Getting your own number
Straightforward and worth doing today.
1. Count the leads from a specific month. From your ad platform or your form system.
2. Count how many of those leads became customers. From the sales system, and from that same month's cohort rather than from the same month's sales, which is a different and misleading number if your sales cycle is longer than a few weeks.
3. Divide. That is your lead-to-customer rate.
4. Then compute both metrics. Spend divided by leads, and spend divided by customers.
5. Look at the gap. That gap is the difference between what you have been reporting and what the business thinks you have been reporting.
Note the cohort timing carefully. If your sales cycle is long, the customers who bought this month came from leads generated earlier, and dividing this month's spend by this month's customers produces a number that describes neither.
Why this site publishes no benchmark
You will find pages giving a typical cost per lead by industry.
Those figures describe an unstated set of businesses with unstated products, prices, sales processes and definitions of a lead. A cost per lead is only interpretable alongside what a customer is worth and how many leads become one, and a figure quoted without both is a number with a decimal point and no content.
The comparison that matters is against your own value per customer, not against somebody else's average. If a customer is worth a lot and your sales team closes a high share of leads, a high cost per lead can be excellent. If neither is true, a low one can still be unaffordable.
Working out what a lead is worth to your business, before setting any target, has enough substance to be its own article on this site, and it is the thing that makes either metric meaningful.
What to report
Report both, labeled precisely.
"Cost per lead" and "cost per customer" are unambiguous. "Cost per acquisition" is not, and the safest move is to stop using it in reporting to people outside marketing.
Show the lead-to-customer rate alongside them. It is the bridge between the two numbers and it is frequently the most decision-relevant thing on the page: a rising cost per lead with an improving close rate is a different situation from a rising cost per lead with a flat one, and only one of those is a problem.
And keep the definitions written down. What counts as a lead, what counts as a customer, what is included in spend. This site covers the wider version of that discipline separately, in its article on what belongs on a dashboard a non-marketer reads, and the definitions are the part that makes a series comparable across a year.
Where the numbers come from, and the checks worth running
Leads should reconcile with the inbox. If your reporting says forty and the sales team received thirty-one, the difference is either duplicates, test submissions, or a conversion action counting something other than a completed submission. Our article on what a conversion should actually be measuring covers that, and the reconciliation is the check that catches it before somebody else does.
Customers should come from the sales system, not from an estimate.
And spend should be defined once. Media only, or media plus agency, or media plus everything. Whichever you choose, keep it constant, because a metric that improved because the definition of spend narrowed has not improved.
The one-line version
Cost per lead is spend divided by hands raised. Cost per customer is spend divided by people who bought. The ratio between them is your close rate, and if you do not know it, you do not know what your marketing costs.
Label them precisely, report both, and never let a platform's vocabulary decide what a number in your report is called. That is the same discipline our article on what wasted spend actually looks like in an account applies at the campaign level: a number that looks efficient at the wrong thing is worse than a number that looks bad at the right one.