How do you calculate customer acquisition cost correctly?

Customer acquisition cost (CAC) is total spend on sales and marketing divided by the number of new customers acquired in the same period — but the devil is in what you include, which time window you use, and how you segment the result. Get any of those wrong and you'll either overspend thinking growth is cheaper than it is, or kill a profitable channel because the blended number looked bad.

The basic formula and what most founders get wrong

The formula itself is simple: CAC = (total sales + marketing spend) / new customers acquired. The mistakes happen in execution. The most common error is omitting non-obvious costs: founder time spent on sales calls, the salary of engineers who build onboarding flows, the cost of content or SEO work, and the commission paid to a payment processor or platform. If you sell through Apple's App Store, for example, you're handing over 15–30% of every transaction — that's an acquisition-side cost that compounds as you grow, and it belongs in your denominator's companion metric (LTV) if not directly in CAC.

The second mistake is mixing time windows. If you run a big paid campaign in October and count all the customers who convert through November and December, your October CAC looks inflated and your November CAC looks free. Match spend to the customers that spend produced. For paid channels with short attribution windows (search, social), a 30-day lag is usually sufficient. For content, SEO, or word-of-mouth-driven businesses, you may need a 90-day or even 6-month window to see true payback.

The third mistake is computing only one number. Blended CAC — all spend divided by all new customers — is useful for board decks but nearly useless for decisions. You need CAC by channel: paid search, organic, referral, outbound, and so on. A blended CAC of $120 might hide a paid CAC of $400 and an organic CAC of $20. Those two channels need completely different decisions.

Segmenting CAC so it actually drives decisions

Once you have channel-level CAC, the next layer is cohort-level CAC. Customers acquired through a discount or promotion often have materially worse retention and lifetime value than customers who came in at full price. If you lumped them together, you may have greenlit a discount campaign that was actually destroying value. Run the LTV calculation separately for each cohort and compare it to the CAC used to acquire them.

For B2B companies, segment by company size or persona. The cost to close an SMB customer through a self-serve funnel and the cost to close an enterprise customer through a six-month sales cycle are completely different numbers, and treating them as one number will cause you to misallocate your sales team. A useful heuristic: if the sales motion is different, the CAC calculation should be separate.

For consumer subscription businesses, pay attention to how platform fees change over subscriber lifetime. As one analysis of Apple and Google's commission structures points out, Apple drops its subscription commission from 30% to 15% after a subscriber's first year — meaning the effective blended platform fee for a retained user base is lower than the fee for a newly acquired one. This kind of dynamic should factor into how you model payback period, not just headline CAC.

CAC payback period: the number that actually matters

Raw CAC in isolation tells you almost nothing. The number that drives funding decisions, hiring plans, and channel bets is CAC payback period: how many months of gross margin does it take to recover what you spent to acquire a customer. The formula is: CAC / (monthly recurring revenue per customer × gross margin %). If your CAC is $300, monthly revenue per customer is $50, and gross margin is 70%, payback is 300 / (50 × 0.70) = 8.6 months.

For SaaS, a payback period under 12 months is generally healthy for SMB customers; enterprise can stretch to 18–24 months because churn is lower and expansion revenue is higher. Consumer businesses with high churn need payback under 6 months or the math rarely works. These aren't universal laws, but they're useful sanity checks before you pour more money into a channel.

What changes the payback calculation dramatically is expansion revenue. If customers upgrade, add seats, or buy adjacent products, their contribution margin grows over time without additional acquisition spend. Factor this in by using net revenue retention when modeling LTV, but keep it out of your CAC payback calculation — payback should reflect how long it takes to recover cost from the initial contract, not from optimistic future upsells.

Using CAC to make better growth decisions

The real purpose of calculating CAC correctly is to answer one question: which channel should get the next dollar? The answer comes from comparing CAC to LTV by channel, with consistent methodology across both. A channel where LTV:CAC is 3:1 or better and payback is under 12 months is worth scaling. A channel where LTV:CAC is under 1:1 should be shut off regardless of volume.

Paul Graham's observation about word-of-mouth growth is instructive here: a product that spreads organically has a superlinear growth curve, and the compounding effect means the effective CAC drops over time as existing users refer new ones. This is why tracking referred customers separately matters — they reveal whether you've built a product with inherent virality or one that's purely dependent on paid acquisition to grow.

Finally, recalculate CAC at least quarterly. Channel efficiency degrades as you exhaust your best audiences, competitors bid up keywords, and platform algorithms change. A channel that had a $60 CAC six months ago may have a $180 CAC today. Founders who set their budget once and don't revisit the math often discover this problem only after burning through a significant part of their runway.

The one thing to do

Calculate CAC by channel, match spend to the cohort it produced, include all real costs, then divide into LTV to get a ratio — and only scale channels where that ratio is 3:1 or better with payback under 12 months.

Frequently asked questions

Should founder time be included in CAC?

Yes, especially early on. If a founder is spending 20 hours a week on sales, that time has an opportunity cost. Exclude it only if you're modeling a scaled state where a sales team has replaced that effort — and even then, include the fully-loaded cost of that team.

What's the difference between blended CAC and channel CAC?

Blended CAC divides all marketing spend by all new customers and is useful for high-level reporting. Channel CAC isolates spend and customers by acquisition source (paid, organic, referral, etc.) and is what you actually use to decide where to invest. Always track both.

How do I calculate CAC if I have a long sales cycle?

Use a lagged attribution model: match the marketing spend from the period when leads were generated to the customers who closed, even if that means looking back 3–6 months. Matching spend to the period when deals close will distort the number during any growth or contraction phase.

At what LTV:CAC ratio should I start scaling a channel aggressively?

A ratio of 3:1 or higher with a payback period under 12 months is the conventional threshold for scaling. Below 3:1, the margin of error is too thin — small increases in churn or channel costs can make the unit economics negative.

Sources

More playbook answers · Growth Prophet home