Customer acquisition cost is everything you spend on sales and marketing in a period, divided by the new paying customers that period produced. For a B2B SaaS company, working out your CAC takes about a minute. The arguments start with what counts as spend, and with what the number is meant to tell you once you have it. On its own, a CAC figure says very little. Read it next to your payback period and it says whether growth is funding itself or eating the runway.

The calculation most founders get wrong is the top half

The formula has two halves and founders usually stint the top one. CAC equals total sales and marketing cost, divided by new customers won, with both sides taken from the same period. Costs you spent in March that produce customers in June belong in a window long enough to hold both, which is why a quarterly calculation behaves better than a monthly one when your sales cycle runs past a few weeks.

The top half means fully loaded. Ad spend, yes. Also the salaries of everyone whose job is winning customers, commissions, the tools that stack runs on, content production, and what you pay agencies and contractors. Skip any of those and you have built a number that flatters you.

Founder time is the line item almost nobody books. At seed stage the founder is the sales team, the demo presenter and half the marketing department. Leave your own hours out and CAC looks close to free, right up until you hire the first salesperson and the metric appears to explode. Nothing exploded. The cost was always there. It was just unpaid and uncounted.

Here is one quarter at a small SaaS company, with invented round numbers so the arithmetic is easy to follow. Ad spend $18,000. Sales and marketing salaries $21,000. Tools $2,400. A design contractor and an agency invoice together $6,000. The total is $47,400, and the quarter produced 26 new paying customers. CAC is $1,823.

Count the ad spend alone and the same quarter produces a CAC of $692. Same company, same customers, same quarter. One of those two numbers would survive a board meeting, and it is not the smaller one.

The same quarter counted two ways: ad spend alone, then fully loaded.

Two labels keep this honest in a deck. Blended CAC uses all spend and all new customers, however they arrived. Paid CAC uses paid spend and only the customers paid channels brought in. Both are legitimate. Trouble starts when a blended figure gets compared with a benchmark built from paid spend, or the other way round, because most published benchmarks never say which one they used.

A good CAC is one that pays back fast enough

Whether $1,823 is a good CAC depends on what each customer returns and how fast. The check is the payback period: CAC divided by monthly revenue per customer, multiplied by gross margin. Run it on the example. A customer paying $199 a month at an 80 percent gross margin returns about $159 of gross profit a month, so a $1,823 CAC pays back in roughly 11.5 months.

Set that against the field. The 2026 Aleph and Benchmarkit benchmarks, built from full year 2025 results at 342 B2B SaaS and AI companies, put the median CAC payback at 16 months, with the top quartile recovering in 6 months or fewer. Under 12 months is the line most operators aim at. Enterprise motions run longer, usually 18 to 24 months, and get away with it because the contracts are larger and customers stay for years. A payback past your runway is the dangerous version, whatever the benchmark says.

The companion check is the ratio of lifetime value to CAC. The guideline most teams use is 3 to 1, and it traces back to David Skok’s SaaS Metrics 2.0 work rather than to any single survey. Treat it as a guardrail, not a trophy. A ratio well above 5 to 1 normally means you are underinvesting in growth, buying too little of a good thing.

The ratio also moves from the other side. Retention work stretches the lifetime inside LTV without CAC changing at all, which is why a customer who stays 36 months instead of 18 quietly doubles the value of spend you have already made.

CAC also has a sister metric on the revenue side. Revenue per visitor reads the same funnel from the traffic end, and the two numbers together show whether a funnel problem sits in what you pay for visitors or in what those visitors return.

The denominator is decided on your landing page

Read the formula again and notice where the argument usually goes. Teams fight over the numerator: the budget, the channels, the agency invoice. The denominator gets a shrug, as if the number of customers a given spend produces were weather.

It is not weather. Say two companies spend the same $18,000 and buy the same 2,000 visitors. One landing page turns 2 percent of them into demos. The other turns 4 percent. Forty demos against eighty, on identical spend, and every figure downstream of that page inherits the gap: cost per demo, cost per trial, and finally CAC itself.

Same spend, same visitors. The conversion rate decides the cost per demo.

We watch this happen in client work. Chat360 doubled demo bookings in two months without spending more on ads, after their page was rebuilt around the questions buyers were actually asking. The spend stood still, the output doubled, and their cost per demo halved at that step of the funnel. WotNot shows the same pattern one step later: pricing page conversions rose 14 percent on traffic the company was already getting, so more of the same visitors became trials.

None of this requires the ads to get cheaper or the market to get kinder. Budgets get reviewed every month in most companies. The page the budget points at gets reviewed when someone finds the time. CAC does not care which of the two you fixed. It only counts what came out.

Three ways the number gets flattered

Counting trials or demos as customers. A trial is a cost you have already paid, not a customer you have won. Put trials in the denominator and CAC drops on paper while the bank balance disagrees.

Quoting a media only figure as CAC. Paid CAC is a useful channel metric when it wears its label. Presented bare in a board deck, it hides the salaries that did most of the acquiring.

Benchmarking against the wrong segment. Your $1,800 CAC looks alarming next to a self serve tool acquiring at $200, and looks cheap next to an enterprise motion acquiring at $11,000. Neither comparison carries any information about your company. Segment, motion and price decide what good looks like.

Frequently asked questions

Should CAC include salaries?

Yes, in the version you make decisions with. At this stage salaries are usually the largest acquisition cost you have, and leaving them out understates CAC by enough to change decisions. Keep a separate paid CAC for channel calls if you find it useful, label it clearly, and never let the two figures share a slide without their labels.

How often should I calculate CAC?

Quarterly suits most early stage teams. Monthly figures swing with the sales cycle, since a deal that took 90 days to close lands in a single month and distorts it. A quarter smooths that out without hiding a trend for long. If your sales cycle runs longer than a quarter, lengthen the window to match it.

What is the difference between CAC and CPA?

CPA is cost per action: a click, a lead or a demo, measured inside one channel. CAC is cost per paying customer across everything you spent to win them. CPA tells you whether a channel works. CAC tells you whether the business does.

Before the next budget conversation, pull one quarter and run the fully loaded version of this. If payback lands well past 16 months, look at the denominator before you cut the numerator. For most B2B SaaS teams the cheapest CAC reduction on offer is the page their traffic already lands on, and it is usually the one nobody has touched in a year. If you want a second pair of eyes on where your funnel leaks, talk to us.