Guide
Customer acquisition cost: how to work out your own
The formula, what it hides, and why industry-average CAC is the least useful number in marketing. Includes a calculator that uses your figures.
Customer acquisition cost is everything you spent to win customers in a period, divided by the number of new customers you won in that same period. Spend $4,000 and win 20 customers, and your CAC is $200. The formula is the easy part. What makes the number useful — or misleading — is what you put in the top, and what you compare the answer against.
Most pages answering this question hand you a table of industry averages. This one argues you should ignore those, and gives you your own number instead.
Work out yours
Use one period — a month is usual — and use the same definitions each time. The two figures below the result are the ones that decide whether your CAC is a problem.
Nothing here is sent anywhere. The figures stay in this page and are gone when you close it.
What goes in the top of the fraction
This is where most CAC numbers go wrong, and always in the same direction: too low. Anything left out makes marketing look more efficient than it is, which is comfortable right up until you scale the spend and the economics do not hold.
| In or out | Why | |
|---|---|---|
| Ad spend | Yes, all of it | Google, Meta, LinkedIn, sponsorships, print |
| Agency and freelancer fees | Yes | The retainer, the setup fee, the creative invoices |
| Salaries | The share that worked on acquisition | A founder spending half their week selling is half a salary |
| Tools | The ones acquisition needs | Ad platforms, CRM, landing pages. Not your accounting software |
| Discounts and referral bounties | Yes | A 20% first-order discount is acquisition spend wearing a different name |
| Renewals and upsells | No | They are what a customer is WORTH, and counting them here flatters the number twice |
| Rent, admin, product development | No | You would pay these with no marketing at all |
What most people report
Ad spend ÷ new customers. $2,400 of Google Ads, 20 customers, CAC $120.
What it actually cost
Ad spend, the freelancer, the landing page tool and half a founder’s week ÷ the same 20 customers. CAC $310.
Both are arithmetic. Only the second survives being scaled, which is when the difference stops being a reporting preference and starts being a cash problem.
Why the industry average is the wrong benchmark
"Average CAC by industry" tables are the most-read pages on this subject and among the least useful, for three reasons that are rarely stated on them.
- The definitions differ. One company counts salaries, another does not. One counts a trial signup as a customer, another waits for payment. Averaging those produces a number that describes no method at all.
- The spread swamps the average. Within any sector, CAC varies by an order of magnitude with price, margin and sales cycle. A $40/month product and a $40,000 contract can sit in the same row of the same table.
- It cannot tell you what to do. Learning that you are above average gives you no action. Learning that your CAC rose 40% while your customer value stayed flat gives you several.
The useful comparison is always against yourself: this month against last, this channel against that one, measured the same way.
The two numbers that make CAC mean something
Value to CAC
What a customer is worth over their life, divided by what they cost to win. Around 3:1 or better means there is room to spend more and the growth is worth funding. Between 1:1 and 3:1 works but leaves nothing spare for a bad quarter. Below 1:1 you are buying customers at a loss, and scaling makes it worse rather than better.
Use gross profit, not revenue. Revenue-to-CAC is the number that makes an unprofitable business look fundable.
Payback period
How many months of a customer's contribution it takes to earn back what you spent winning them. This is the cash question, and it is the one that kills companies whose ratio looked fine: a 4:1 business with a 24-month payback is profitable on paper and running out of money.
Under 12 months is comfortable for most businesses. Over 18 means every new customer is funded from cash you already have, so growth is limited by your bank balance rather than by demand.
Three ways CAC gets reported wrong
- Platform CPA reported as CAC. Ad accounts report the cost of a conversion event, not of a paying customer, and they exclude every cost that did not run through them. It is always the smaller number.
- Blended CAC used to judge a channel.Total spend over total customers hides the fact that word of mouth is carrying an expensive channel. Both figures are worth having; only one answers "should we keep paying for this".
- Attribution lag ignored. Spend in March that wins a customer in May makes March look expensive and May look brilliant. The longer your sales cycle, the wider the window has to be before the monthly number means anything.
Getting this without a spreadsheet
The reason most businesses do not track CAC monthly is not that the arithmetic is hard — it is that the inputs live in four places. Spend sits in Google Ads and Meta, conversions in Analytics, customers in whatever you sell through, and the salary share in somebody's head.
Rallik connects those accounts and keeps the cost of a result beside the decision that caused it, so the question "is this channel still worth it" is answered with your own numbers rather than a benchmark. It will also tell you when it cannot measure something — which, for anything involving attribution, is more often than most dashboards will admit.
Related: how to decide where the budget goes once you know what each channel costs you, and whether you need a person or a system to make that call every month.
Common questions
- How do you calculate customer acquisition cost?
- Divide everything you spent to win customers in a period by the number of NEW customers won in that same period. Spend includes ads, agency and freelancer fees, acquisition tools, first-order discounts and referral bounties, plus the share of any salary that went into sales or marketing. It excludes rent, product development and anything you would pay with no marketing at all. If you spent $4,000 and won 20 new customers, your CAC is $200.
- What is a good customer acquisition cost?
- There is no good CAC in isolation — $200 is excellent for a customer worth $3,000 and ruinous for one worth $180. Judge it against two things: the value-to-CAC ratio, where roughly 3:1 or better means you have room to spend more, and the payback period, which is how many months of a customer's contribution it takes to earn back what you paid to win them. Under 12 months of payback is comfortable for most businesses; over 18 means growth is being funded out of cash you may not have.
- Is the average CAC for my industry useful?
- Almost never. Industry averages are computed across businesses with different margins, sales cycles, prices and definitions of what counts as spend, and most published tables do not say which definition they used. Two companies in the same sector can differ tenfold on CAC and both be healthy. The number worth knowing is your own, measured the same way every month so the direction means something.
- What is the difference between CAC and CPA?
- CPA — cost per acquisition — usually means the cost of one conversion event inside an ad platform: a lead, a signup, an add-to-cart. CAC is the cost of one paying customer across everything you spent, including the work that never touched an ad account. Platform CPA is always the smaller and more flattering number, and reporting it as CAC is the commonest way marketing spend looks more efficient than it is.
- How often should I recalculate CAC?
- Monthly, and always over the same window. CAC is most useful as a direction rather than a level: a number that has moved from $180 to $260 over three months tells you something urgent, while a single month's figure mostly tells you how lumpy your sales are. Watch out for lag — money spent in one month often wins a customer in the next, so a fast-changing budget makes a single month's CAC unreliable.
- Should CAC include salaries?
- Yes, in proportion to the work. Leaving salaries out is what makes agency case studies look impossible to match: their client's CAC excluded the two people managing them. If a founder spends half their week on sales, half that salary belongs in the calculation — and if that makes the number uncomfortable, the number was already uncomfortable and you were not looking at it.