Blended CAC is lying to you

Blended CAC looks clean on a board slide and hides the one thing you actually need to know: which channel is buying you profitable customers, and which is buying you almost nothing.

What blended CAC hides

Total spend divided by total new customers averages a brilliant channel and a terrible one into a single number that describes neither. Scale the blended figure and you scale the average, not the winner. You are flying on the mean.

New-customer CAC vs blended

If returning customers sit in your denominator, CAC looks wonderful while you are really just paying to re-acquire people who would have bought anyway. New-customer CAC is the honest number, and it is usually a lot less flattering.

The attribution trap

Channel CAC computed straight from platform numbers is fiction, because every platform claims the same customer. Without one deduplicated, governed source, per-channel CAC is just each platform's sales pitch wearing a spreadsheet.

You cannot scale what you cannot see. Blended CAC is a blindfold that feels like a dashboard.

The fix: channel-level, new-customer CAC on governed data

One source of truth for spend and orders, deduplicated attribution, and a clean new-versus-returning split. Then you can finally pour budget into the channel that actually works and starve the one that does not.

How dataeze does it

We build the governed foundation, then put an AI analyst on top that answers "what is my new-customer CAC by channel this week" in plain English, traceable to the SQL. Live in 6 to 8 weeks.