How to read incremental CAC on Facebook and Google ads

The figure reads low because Facebook and Google report one blended cost, mixing in buyers who were coming to you regardless. Your earliest customers were the cheapest to convert, so the average starts there and drifts upward as you scale. Angel investor Avishai Sam Bitton assumes that drift, and treats a strong early number as easy demand a company was always going to win.

Raise your spend on a single platform, keep the rest flat, and read the blended cost as it rises. A number that jumps means you were mostly harvesting demand already yours; one that holds means the spend is truly reaching new buyers. Etgar Shpivak reads that climb as the clearest signal of incremental cost a founder gets without special tooling.

Accurate at counting clicks and sales, and unreliable as a verdict on their own worth, because the same platform selling the ads is also scoring them. Treat the reported cost as the optimistic case. Avishai Sam Bitton backs founders whose attribution, the way credit for a sale gets assigned, runs deliberately pessimistic, so the numbers they log already discount the platform's own math.

Assume the cost climbs, because every new tranche of budget reaches buyers a little further from the sale than the last, so the cost per customer rises even when nothing is wrong. Avishai Sam Bitton also assumes lifetime value, what a customer pays you over the whole relationship, drifts down at the same time, because the first customers were the most eager. Etgar Shpivak builds that discount into the model before a founder scales the spend.

Investors want the acquisition cost read beside what a customer is worth over time and how long the spend takes to pay back, never one figure on its own. Avishai Sam Bitton told Etgar Shpivak on Founders' Marketing Compass that misaligned numbers are the real red flag: change one and the rest should move with it. Hold the whole set together and the ratio worth raising on is a harder number than the tidy 3:1 a deck tends to show.

Very little belongs on brand at the start, and Avishai Sam Bitton ranks it no higher than third in an early company's priorities. He argues the cheap version, building in public and documenting the work, carries a young company until it can see a path to an exit. Put the budget into winning customers now, because revenue is what proves the company works. How much to spend on marketing by stage has real thresholds of its own.

Put more budget behind one platform, hold everything else steady, and watch the blended cost as volume grows. If it climbs, the low figure you saw earlier reflected demand you already had, not the price of the next customer. A rise like that doesn't mean the platform is finished, only that the next customer costs more than the last.

Founders who read every rise in cost as a dead platform talk themselves out of a channel that still works. A climbing cost means the next customer is dearer, not that you've hit the ceiling. Plan for a cost that only rises, but don't treat every upward move as a death sentence, or you'll cut the very thing driving growth.

A cost modeled below reality lets you revise upward when the real number comes in better, which reads as a careful company rather than one that missed. The reverse, revising a confident number down, burns trust. Better to walk in with a conservative figure that has more room to surprise on the upside than on the down.

A cost that looks too cheap can mean you are under-scaled, not that you've found a magic channel. A return that looks too good to be true usually means there was room to spend far more and win more customers at the same price. Your first customers are always the easiest, so a low early cost is mostly demand you already had.

Hire a specialist to run the ads once you know which numbers the spend is judged on, and bring in a consultant while you are still working those numbers out. Etgar Shpivak, a marketing consultant who works with seed and Series A founders, does that with the team before anyone scales a budget. Avishai Sam Bitton would back a founder who can read these numbers over one who hands them to a title. See how Etgar works on his bio.