Your Real CAC Is 3x the Dashboard Number

Thursday afternoon at one software company, and the owner is on her third demo of the day. Say the firm does $4M a year in software and setup services. She will write the quote herself tonight and run the setup call next week, because she is the only person who knows how the product bends. The monthly report her bookkeeper sends says customer acquisition cost is $1,400. That number counts ad spend and the rep's commission, divided by new customers. Her afternoon is not in it.

Say she loads it properly. In her case, a two-week calendar tally shows about 35% of her working hours go to demos, quotes and setup calls. Say her salary is $200,000; that makes $70,000 a year of it acquisition spend. Spread across the 30 or so new customers she signs a year, that's $2,333 per customer. There is no customer-success team, so her account manager does onboarding, roughly 12 hours per customer at $40 an hour: another $480. In her books, loaded CAC comes to about $4,200, three times the dashboard figure. Every input was defensibly somewhere else on the P&L. Stacked, they are the real price of a customer.

The difference matters because of payback. Her customers pay $500 a month at a 70% gross margin, so each one returns $350 a month. On the dashboard number, $1,400 per customer, payback takes 4 months, which says pour money into ads. On the loaded number it takes 12. Same customers, same channel, a very different instruction.

Three decisions turn this from a report into a control. First, the owner makes the allocation call, not the bookkeeper. Which hours count as selling is a judgment, so she writes the rule down once (share of salary from the calendar tally, re-tallied each January) and the bookkeeper applies it the same way every month. If she would rather use an hourly method, the cleaner rate is what she would pay a contract sales engineer per hour, because that is what her time would cost to replace. Either works. Booking it at zero doesn't.

Second, she sets a payback ceiling before she looks at the channels. Say 12 months on loaded CAC. If loaded payback runs above it two months in a row, she freezes new spend in the worst channel. If it stays above for a quarter, she cuts that channel and puts the hours back into referrals or price. The rule has teeth only because the person who signs the ad invoices also wrote the ceiling. This is the step that usually dies in the unit economics nobody reopens.

Third, she prices the bank line into payback. If acquisition is funded by the line of credit, payback in months is the number of months the bank carries each customer. At a 12-month payback, a customer signed in January is still on the line at renewal. At a 9% rate, carrying $4,200 down to zero over a year adds roughly $190 of interest, so loaded CAC is closer to $4,400 and payback slides past 12. A ceiling longer than the line's renewal cycle means growth is borrowing against a review she has not passed yet. Shorter is safer, and it's why reporting the flattering figure costs more when the money is borrowed.

The monthly report keeps both numbers. The dashboard CAC tells the rep which ads to tweak. The loaded CAC, next to the payback ceiling, tells the owner whether to keep writing the checks.