Two numbers get called "acquisition cost," and they are not the same. One is what you pay to get a lead: a form fill, a call, a message. The other is what you pay to get a customer, someone who actually bought. Most benchmark articles blur the two, so you can end up comparing your cost per customer to someone else's cost per lead and deciding your business is either a disaster or a bargain.
Lead cost and customer cost, side by side
The table shows where each figure sits in the path from ad to sale. If a source doesn't say which row it measures, treat its number with caution.
| Figure | What you divide | What it tells you |
|---|---|---|
| Cost per click | Ad spend by clicks | How expensive attention is in your market |
| Cost per lead | Spend by inquiries | How well your ads and page attract interest |
| Cost per customer (CAC) | Spend by new paying customers | Whether growth pays for itself |
Why the gap between them can be huge
Many inquiries never become buyers. Some are price shoppers, some are outside your service area, and some never answer the callback. The closer your sales process is to a handshake and a quote, the wider the gap between a lead and a customer. A cheap lead that rarely closes can cost you more per customer than an expensive lead that usually does.
Work out your own figure first
Take everything you spent to win customers over a period: ad spend, agency or freelancer fees, software, and the share of anyone's salary devoted to selling. Divide by the number of genuinely new customers in the same period, not repeat purchases. Use a period long enough to cover your normal sales cycle, or the ratio will swing wildly.
If a customer came from a referral but also clicked an ad, decide in advance how you will count them and stay consistent. A rough, consistent number beats a precise one that keeps changing its rules.
Using outside figures without being fooled
Treat a published benchmark as a sanity check on order of magnitude, not a target. Before you compare, confirm that it measures customers rather than leads, that it covers a business with a similar price and repeat-purchase pattern, and that the source explains how it was collected. If it fails any of these, set it aside. For the lead-cost version of this problem, see our piece on [why industry averages for cost per lead mislead](/guides/average-cost-per-lead-by-industry).
Why the published data is so thin
Searches for this phrase are rare: about 40 a month across five related queries, and the main phrase itself ranged from 10 a month at its quietest to 20 at its busiest in the past year. Yet advertisers bid as much as $13.46 for the top spot on a click. Few people are looking, and those who are tend to be serious. You also won't find a deep pool of trustworthy public benchmarks, so your own records are the best source you have.
Turn the number into a decision
Compare your cost per customer with the gross profit a typical customer brings you. If you spend less than that profit to win them, growth is paying for itself and you can consider spending more. If you spend more, look first at the weakest step: leads that don't close, slow follow-up, or tracking that undercounts real sales. Fix the leak before raising spend.
What to do this week
Pull your last full period of new customers and total acquisition spend, and write both figures down. Then note where each customer first came from, even if the answer is "not sure." The "not sure" pile shows how far to trust the final number.
When the numbers refuse to reconcile
If ad platform reports, your analytics and your sales records tell three different stories, the problem is usually tracking, not marketing. An outside analyst can audit your conversion setup and tell you which figures to trust. That's one option, and a worthwhile one when your own records can't settle the question. If they can, you may not need anyone.