incremental CAC

Also called: incremental, incremental customers, counterfactual baseline

Spend divided by the customers it actually added beyond what would have come anyway. Expected = before × the holdout's growth; incremental = actual − expected; incremental CAC = spend ÷ incremental.

Not every customer who arrives during a campaign was caused by it. To find what the spend really added, compare with a holdout: a similar group, region or period without the campaign.

$\text{expected} = \text{before} \times \frac{\text{holdout after}}{\text{holdout before}} \qquad CAC_{incr} = \frac{\text{spend}}{\text{actual} - \text{expected}}$

The expected value is the counterfactual baseline: what would have happened anyway, growing at the holdout's rate. Only the difference above it is incremental.

Incremental CAC is usually higher than the naive one, because part of the growth was coming anyway (season, word of mouth). Budget decisions should compare incremental CAC with a customer's value.

Example

Lviv: 2,000 sign-ups before, 2,300 during the ads; the holdout grew 5%; spend 300,000 UAH.

Expected: 2,000 × 1.05 = 2,100.
Incremental: 2,300 − 2,100 = 200.
Incremental CAC = 300,000 ÷ 200 = 1,500, not 300,000 ÷ 300 = 1,000.

Common mistakes

Crediting all growth to the campaign. Seasonal growth happens without ads.
A holdout that isn't similar. A different region or season makes the baseline wrong.
Dividing by all new customers. CAC on total sign-ups understates the cost of the ones the spend bought.

Learn it in the course

What do we divide by? Denominator traps · ARPU vs ARPPU, LTV per cohort user vs per survivor, win rate with or without "no decision", gross vs net margin, concentration with the item in the total, and incremental results against a control group.
Flows: customers and cash · One rule for every stock, end = start + in − out: MRR, churn, NRR and GRR; profit vs cash, working capital, DSO and runway; CAC payback and incrementality against a holdout.