The short answer
A CAC baseline is the cost of acquiring one new customer, measured over a fixed period before a new message reaches the market. A-square aligns on the spend, the customer definition and the period with the client before anyone runs the numbers, so the figure holds up when results are judged.
Why does the baseline have to come first?
A number reconstructed after the new message is live will be argued about. So, it is crucial to agree on the definitions before anyone has run the numbers, and to lock the number before the message can move it.
Acquisition keeps getting more expensive. In 2024 the median B2B SaaS company in Benchmarkit's survey spent $2.00 on sales and marketing to win $1.00 of annual recurring revenue from new customers, up 14% from 2023 (Benchmarkit, 2025 SaaS Performance Metrics report). At that price, a positioning project that cannot show its effect on acquisition cost is a cost with no measurable return.
What are the four steps?
- Fix the period. Use the quarter before the new message goes live. A quarter balances recency against the impact of individual deals: it is recent enough to reflect current pricing and team, and long enough that no single deal swings the number. If you win only a handful of customers a quarter, use two quarters.
- Agree what counts as spend. Paid media, agency fees, software and the fully loaded cost of the sales and marketing team. Write the list down and use it every time CAC is measured.
- Agree what counts as a customer. Closed-won in the CRM, a paid conversion or a signed order form. New customers only, not upsells or renewals. One written definition that catches every new customer.
- Match spend to the customers it won. B2B deals often close months after the first touch, so this quarter's customers were won with earlier spend. Move the spend window back by your typical sales cycle. If most customers buy within a month, use the same quarter for both.
Divide spend by customers. With a three-month sales cycle and a launch on 1 July, count new customers won from April to June and spend from January to March. If that spend was $180,000 and the company won 24 new customers:
Baseline CAC = $180,000 ÷ 24 = $7,500
That number, its two windows and its two definitions are the baseline.
Where does the number come from?
Take spend from finance and customers from the CRM or billing system. Finance is usually the only record that holds salaries, agency fees and software alongside media. A CRM such as HubSpot holds the deals but rarely the full cost. A number the founder supplies is acceptable if the inputs come with it. What is not acceptable is a number nobody can trace.
What makes a baseline meaningless?
- Changing a definition midway.
- Dividing a quarter's spend by the same quarter's customers when deals take months to close.
- Using a window that contains an unusual event, such as a conference or a price change, without noting it.
- Reading CAC without average deal size. A shift towards larger customers raises CAC even when it pays off.
How do you know the work moved it?
A lower CAC after launch is not proof on its own. CAC also moves with the season, competitor bidding, pricing and hiring, and a baseline taken at a spike tends to look beaten whatever the work does.
Three checks guard against this. Compare each later quarter with the same quarter a year earlier, not only with the baseline. Treat any change smaller than the usual quarter-to-quarter swing as noise. Log unusual events in every window and consider confounding factors when judging impact.
What happens after it is locked?
On A-square engagements the baseline CAC is one of six things locked in week eight, at the end of positioning and before the new message goes live. The other five are the message, its category, the AI questions Memory Score tracks, the website pages it scores and the content it checks against the message.
From week nine, Memory Score is shown every week next to a rolling CAC, worked out over the same window length and on the same rules as the baseline. A single week holds too few deals to read on its own. CAC also lags the message: with a three-month sales cycle, it starts to reflect the new message after about three months and fully after about six.
In the GTM Partnership, half the fee depends on agreed outcomes, one of which is lower CAC. The definitions, window, lag, thresholds, attribution rules and the date each outcome is read are all agreed in week one, before anyone runs the numbers. The CAC threshold is set as a percentage change from the baseline and applied to the number locked in week eight.
