Most ad accounts we open are optimizing for the wrong number.
The platform reports cost per lead. The dashboard looks healthy. Spend goes up, leads go up, and everybody feels good until the sales team quietly mentions that nothing is closing.
The problem with "lead"
To Meta or Google, a lead is a form submission. To your business, a lead is someone who can buy, wants to buy, and is worth the time it takes to sell to them. Those are not the same thing, and the gap between them is where budgets go to die.
When you optimize for form fills, the algorithm does exactly what you asked: it finds the people most likely to fill in a form. That is a different population from the people most likely to become customers. Often a much cheaper one.
What to measure instead
Push a real outcome back into the ad platform. Depending on your sales cycle that is one of:
- Qualified lead — a human looked at it and said "yes, this is a real prospect"
- Booked appointment — they showed up on the calendar
- Closed revenue — the actual number, with a value attached
The further down that list you can get, the better the algorithm optimizes. Even a rough "qualified / not qualified" flag from the CRM, fed back as an offline conversion, changes what the platform goes looking for.
The practical version
You do not need a data team. You need:
- A CRM that records where each lead came from (click IDs, not just "Google")
- One field the sales team actually fills in
- A weekly upload or integration that sends that field back to the platform
That is usually a two-week project. It is the first thing we do on every account, before a dollar of new spend, because everything after it depends on the platform knowing what a good lead looks like.
Cost per lead is a diagnostic. Cost per qualified lead is a result. Report the second one.
- tracking
- paid-media