glossary · metrics
Customer acquisition cost (CAC)
Total sales and marketing spend divided by new customers acquired in the same period.
CAC is the metric that converts growth ambition into an economics question. The formula is simple: CAC = Total Sales and Marketing Spend / New Customers Acquired. What trips up most PM candidates is not the formula but knowing what belongs in the numerator, when blended CAC misleads, and how to connect the number to payback period and LTV:CAC in a single coherent argument. Reciting the formula and citing 3:1 is not a pass.
What goes in “total sales and marketing spend”
The numerator is not just ad spend. It includes: salaries and commissions for every sales and marketing headcount, paid media (SEM, social, display), content production costs, martech and salestech tool subscriptions, referral and affiliate fees, and allocated overhead for sales ops and marketing ops. Miss headcount and your CAC is understated by its largest component.
Blended vs. channel-level CAC
Blended CAC divides all S&M spend by all new customers. It is the board-level efficiency snapshot, useful for trend analysis and investor reporting.
Channel-level CAC divides spend on one channel by customers acquired through that channel. This is the operational lever. A blended CAC of $400 that hides a $2,000 paid social CAC and a $50 organic CAC is not a health signal: organic is subsidizing a channel you should probably kill. For any acquisition or pricing decision, channel-level is almost always the right number.
Attribution is where this gets hard. A customer who touched SEO, a webinar, and a paid retargeting ad before converting must be assigned to something. Last-touch attribution is easy and wrong. Data-driven attribution is better and harder. Know which method your data team uses before citing a channel CAC number in an interview.
CAC, payback period, and LTV:CAC as one narrative
These three numbers only mean something together.
Payback period = CAC / (Monthly Recurring Revenue per customer x Gross Margin). Healthy SaaS targets: 12-18 months for SMB, 18-36 months for mid-market and enterprise. A 3:1 LTV:CAC ratio with a 48-month payback period is a cash flow crisis, not a success signal: you run out of working capital before customers pay back their acquisition cost.
LTV:CAC ratio: 3:1 or better is the standard SaaS benchmark. Below 1:1 means you lose money on every customer. Between 1:1 and 3:1 means growth is capital-inefficient. Above 5:1 often signals under-investment in acquisition. Cite the benchmark, then explain what it implies for reinvestment capacity.
Cohort-level CAC matters. Customers acquired via paid in a competitive quarter cost more and often retain worse than organic cohorts. Blended CAC can look stable while your best-cohort economics deteriorate underneath it.
2026 benchmarks worth knowing
First Page Sage data from roughly 120 client firms (2021-2024 campaigns): B2B organic CAC averages $942 (range $510-$1,786); B2B inorganic averages $1,907 (range $802-$4,664); B2C organic averages $480; B2C inorganic averages $319. Project management software CAC has risen from $891 to $1,020 year-over-year, a 14.5% increase, with enterprise-focused vendors averaging $3,870 per closed account. Separately, 53% of marketing budgets in 2026 now target existing customers rather than new acquisition, which tells you where the ROI has shifted.
The PLG and AI adjustment
PLG compresses CAC by making the product the primary acquisition channel. SaaS companies allowing sign-up without a credit card generate roughly 2x as many paying customers from free trials. Virality and in-product referral loops reduce dependence on inorganic spend.
For AI products, gross margin in the payback formula must account for inference cost per active user, not just SaaS-style hosting. A product with a $50 CAC and $0.10 per query inference cost at high usage volume can have a true margin far below what a naive calculation shows. The PM who ignores inference cost in unit economics will give a confident and wrong answer in any growth or pricing discussion. See LTV:CAC unit economics for the full treatment.
Strong vs. weak answer in an interview
weak
"CAC is total sales and marketing spend divided by new customers acquired. A healthy LTV:CAC ratio is 3:1." This fails because it treats CAC as a single blended number, ignores the time dimension of payback period and its cash flow implications, says nothing about what a PM would actually do with the number, and shows no awareness of PLG or AI cost structure.
strong
"CAC equals all sales and marketing spend, including headcount, ad spend, tools, and referral fees, divided by new customers acquired. I'd distinguish blended from channel-level: blended signals board-level efficiency, but channel-level drives decisions. If blended looks healthy at $400 but paid social is running $2,000 while organic is $50, the right call is probably to cut paid social and invest in the organic motion. I'd connect CAC to payback period using gross margin, not just MRR, and target 12-18 months for SMB. For an AI product, I'd also make sure gross margin accounts for inference cost per active user, because a product can look profitable on CAC alone and still destroy value at scale. The PM's actual levers on CAC are usually product-side: reducing time-to-value to improve trial-to-paid conversion, adding referral loops, or designing for PLG so the product acquires the next user."
The viability question made concrete
In 2026, feasibility is effectively free. You can build almost anything. That shifts the entire weight of product strategy onto viability: is this a problem people will pay to have solved, and can you reach those people at a cost that revenue can sustain? CAC is that question made concrete. A PM who treats it as a finance metric to recite has missed the point. The real question is always: what is the cheapest true way to reach the person who has the problem, and does the product itself reduce that cost over time? That question connects directly to proving viability in an AI context.