How CAC is calculated
Take everything you spent to acquire customers over a period and divide it by the number of new paying customers won in that period. The arithmetic is trivial. The argument is always about what belongs in the top half.
A defensible version includes media spend, agency or freelancer fees, creative production, the acquisition tools you subscribe to, and the salary of anyone whose job is winning new business. It excludes the cost of serving customers you already have, product development, and general overheads. What matters far more than getting the boundary theoretically perfect is writing it down and never quietly changing it, because a CAC that improves because the definition shifted is worse than useless.
CAC is not the same as cost per acquisition. CPA counts a conversion event — a form, a call, a sign-up. CAC counts a customer who paid. Where the sales cycle is long, the two also fall in different months, so a period that looks efficient on CPA can look expensive on CAC simply because the deals had not closed yet.
Why CAC matters
On its own, CAC means nothing. Dear or cheap is decided by what a customer is worth, which is why it is only ever read next to lifetime value and gross margin. A business selling a high-margin service that customers renew can afford an acquisition cost that would destroy a low-margin retailer selling once.
It is also the number that decides whether growth is available. If CAC stays flat as you increase spend, there is more demand to buy. If it climbs steeply, you have reached the edge of the audience that wants what you sell, and the next gain has to come from conversion rate, pricing or a new market rather than from budget.
Common mistakes with CAC
Counting only media spend is the usual one, and it flatters every channel that needs a person to run it. Mismatched time windows are next: spend from this month divided by customers won this month, when those customers were reached weeks earlier.
Then there is counting existing customers as new. Repeat orders, renewals and people who were coming back anyway all push CAC down without any acquisition having happened. Finally, a single site-wide figure hides everything useful; a strong channel and a failing one average into a number that describes neither.
How to act on it
Start by defining what counts as a customer and what costs are included, and put both in writing where the team can see them. Then split CAC by channel and by new versus returning, because the whole point is deciding where the next rupee or dollar goes.
Judge the result against margin and against how long the customer takes to pay that cost back, not against a figure from another industry. If CAC is rising, look at conversion rate and offer before you look at bids — it is usually cheaper to convert more of the traffic you already pay for than to buy more of it.
