Decide what the CAC line on a board slide should say when blended CAC fell 20% and paid CAC rose, and whether to approve doubling paid spend.
casemediumExercise · 3 min · 19 of 60
On this page
The board slide says CAC fell by a fifth and asks to double paid spend. Both statements rest on one number, and it is the wrong one for the decision being asked for.
Last quarter the invoicing app spent $50,000 a month on paid media for 96 paid customers a month, and 64 more arrived organically. This quarter paid spend rose to $60,000 a month. Paid customers rose to 104 a month and organic to 136, after the content library passed 200 posts and a referral credit launched. A customer returns $40 a month of margin at 5% monthly churn, so LTV is $800, and cumulative margin by month n is 800 × (1 − 0.95ⁿ). The company's target is payback within 12 months, which caps CAC at about $368.
The founder's board slide reads: "CAC $250, down from $313. LTV/CAC 3.2. Plan: double paid spend to $120,000 a month." The board meets on Thursday, and you have been asked to check the slide before it goes.
$250 is blended CAC, $60,000 ÷ 240, in the terms of blended versus paid CAC. It fell because the organic share rose from 40% (64 of 160) to 57% (136 of 240), not because paid got cheaper. Paid CAC went the other way: $60,000 ÷ 104 = $577, up from $521, and its payback moved from month 21 to month 25 against a 12-month target. The last $10,000 bought 8 customers, $1,250 each. That is more than the $800 a customer can ever return, so the increment never pays back.
The line should read: "Paid CAC $577, up from $521; payback month 25. The last $10,000 bought customers at $1,250. Blended CAC $250, because 57% of new customers now arrive without a paid click: the content program and referral working." LTV/CAC either goes or is shown on paid, 1.4.
Do not approve the doubling. Further spend buys customers at or above the $1,250 the last increment cost, so another $60,000 a month buys customers who never repay. Recommend the opposite: return paid to $50,000, which saves $10,000 a month that was buying $6,400 of lifetime margin, and fund the content program that produced the growth.
What would change this: some organic customers may be caused by paid, people who saw an ad and later typed the name. Pause paid in a fifth of regions for four weeks. If organic sign-ups there fall enough that the last $10,000's true cost comes in under $368 a customer, the case for more paid reopens on that number.
"Report blended; it is the company's true CAC." It is a true average and the wrong figure for a spend decision. It credits the content program's customers to paid, and it approves $60,000 a month of customers who never pay back.
"Report paid, $577: LTV/CAC is 800 ÷ 577 = 1.4, above 1, so doubling still makes money." This is an average too. The decision is about the next dollar, and the last $10,000 already cost $1,250 a customer; even the average pays back only in month 25.
"Organic is free, so cut paid to zero." Untested, and possibly expensive: if some organic customers first saw an ad, cutting paid cuts organic with it. The regional holdout answers that question at a fifth of the risk.
