Home/Insights/How to read a CPA benchmark without fooling yourself
Insight

How to read a CPA benchmark without fooling yourself

A cost-per-FTD number on its own tells you almost nothing. The market, the month on the curve and the method behind it are the whole meaning.

Most published iGaming CPA numbers are quoted without the three things that give them meaning: which market, which month of the campaign, and how the figure was measured.

Read a CPA benchmark as a band tied to a market and a point on the decline curve, never as a single steady-state number. A 250 EUR CPA in month one and a 250 EUR CPA in month nine describe completely different campaigns.

The three questions

First, which market: a Nordic CPA and a Swiss CPA are not comparable, because the LTV behind them differs by a factor of four. Second, which month: month-one CPA runs 1.5 to 2 times the eventual target while the pixel matures, so a number without a month is meaningless. Third, the method: is it blended across GEOs, is it measured to FTD or to registration, and what sample sits behind it.

This is why every benchmark on this site carries a source line, a data-through date and a changelog. A number you cannot date is a number you cannot trust.

Questions operators ask

What makes a CPA benchmark trustworthy?

A market, a point on the decline curve, a measurement basis (to FTD, not registration), a sample size and a date. A single number with none of those is marketing, not a benchmark.

Why compare CPA to LTV, not to other markets?

Because CPA only means something against the value it buys. A high CPA in a high-LTV market can be a better deal than a low CPA in a low-LTV one.

Tell us the market. We tell you what converts there.

A short brief and a 20-minute call. GEO, monthly budget band, vertical. You leave with a channel read, not a sales deck.

Book a call