What is customer acquisition cost (CAC)?

This is a business figure, not a platform metric. The numerator holds everything the company spent on winning customers, including the people who do that work, and the denominator counts customers, not orders and not form submissions. Someone who buys three times in the period enters only once.

This is where the most common confusion in digital marketing appears: taking the cost per conversion from Google Ads as if it were CAC. The platform calculates its average CPA by dividing the cost of conversions by the number of conversions, and that calculation has three limits. It only includes spend on that platform, so it leaves out salaries, tools and fees. It counts conversions, which depending on the counting setting may be recorded once or every time after an interaction. And it does not distinguish whether the person who converted was a new customer or had already bought before.

The system itself acknowledges that distance by another route. For its new customer acquisition goal it needs to know who is new, and it resolves that with a list built on up to 540 days of recorded purchases, with Customer Match lists or with a specific parameter in the conversion tag. New customer and conversion are, for the platform too, two different things.

Unlike CPA, CAC answers a management question: what it costs to grow the customer base by one unit. That is why it is calculated per period and per segment, and always read alongside what that customer will contribute later.

The base formula fits on one line.

CAC = gasto de captación / clientes nuevos

The serious discussion sits in the numerator. The narrow version counts only paid media and is easy to obtain; the complete version includes the cost of the entire commercial function and is the one that shows whether the business model holds.

gasto de captación =
    inversión en medios
  + salarios de marketing y ventas
  + comisiones variables
  + herramientas y software
  + honorarios de agencia
  + producción de contenidos

Left out are the costs of serving and retaining people who are already customers: support, account management, loyalty programmes. Pulling those in turns CAC into a catch-all bucket and destroys its usefulness. The difference against the platform figure shows best when placed side by side.

CPA (Google Ads) =
    coste de las conversiones
    / conversiones

CAC = todo el gasto de captación
      / clientes nuevos

A worked example, with invented figures to illustrate the method. A company spends 18,000 euros on media in a month, 9,000 on marketing and sales salaries and 1,200 on tools, and wins 240 new customers. CAC is 28,200 / 240 = 117.50 euros. If Google Ads recorded 460 conversions that same month at a cost per conversion of 39 euros, neither figure is wrong: they count different things and serve different decisions.

The timing gap deserves a rule of its own. In long sales cycles, January spend produces customers in March, and dividing those two months by each other yields a meaningless CAC. The usual way out assigns the spend to the cohort of customers it generated, even if the result takes longer to settle.

CAC decides whether the business model holds up. It is compared against what a customer contributes across the relationship and against the gross margin on that amount. If winning a customer costs more than the customer will contribute, growth accelerates losses instead of reducing them, and no adjustment of ad creatives fixes that.

The second decision concerns cash: how many months the margin generated by a customer needs to pay back what it cost to win them. That payback period sets the pace at which the company can invest without draining the bank, and it is more actionable than the ratio of lifetime value to CAC, for which reference proportions circulate without published methodology.

The third is the split across channels. A channel with low CAC and customers who do not return can be worse business than one with high CAC and loyal customers, so CAC per channel only makes sense read together with the later behaviour of those customers. It also feeds the conversation about pricing: raising the average price widens the CAC the company can afford, and sometimes that is the fastest lever.

• Write down which cost lines belong in the numerator and what counts as a new customer, and review that definition once a year rather than changing it on the fly.
• Calculate a narrow CAC from media only and a complete one with salaries and tools. The first serves campaign optimisation, the second budget decisions.
• Assign spend to the cohort of customers it generated whenever the sales cycle runs longer than a month. Dividing spend and customers from the same calendar month distorts long-cycle businesses.
• Separate new customer CAC from the cost of reactivating an inactive customer. These are operations with very different costs and returns.
• Always compare CAC against the margin the customer leaves, not against their revenue. The invoiced amount is not available money.
• When you use a figure from an advertising platform, store the exact column name and its counting setting next to it, so that later you know what you were looking at.

• Presenting the Google Ads cost per conversion as CAC. It ignores staff and tooling costs, and it counts conversions that may come from existing customers.
• Counting all orders of the period in the denominator instead of new customers, which artificially lowers the figure in businesses with heavy repeat purchase.
• Adding retention and support costs to the numerator, so that CAC stops measuring acquisition and starts measuring total commercial spend.
• Attributing the whole result to the last click and deciding budget on that picture, especially when some channels only appear early in the journey.
• Comparing your own CAC against reference figures whose industry, period and included cost lines are unknown.