What ad-level attribution actually changes

Most brokerages can tell you how many leads came from Facebook last month. Very few can tell you which of the eleven ads they were running produced the ones that turned into viewings.

That gap is where the budget leaks.

“Source: Facebook” is not attribution

If every paid lead is tagged the same way, your reporting can answer exactly one question: should we spend on Facebook at all? Since the answer is almost always yes, the report changes nothing.

Useful attribution goes down three levels (campaign, ad set, and the individual ad) because that is the level at which you can act. You cannot switch off “Facebook”. You can switch off one creative that produces enquiries who never answer the phone.

The two numbers that matter, and why cost per lead is not one of them

Cost per lead rewards the ad that collects the most form fills, which is usually the one promising the lowest price on the flashiest render. Those leads are cheap because they are unqualified.

What you want, per ad:

• Leads that reached a real conversation. Not submitted, answered.
• Leads that reached a viewing. The first point where the ad has demonstrably produced something of value.

Rank your ads on those two and the ordering usually differs sharply from the cost-per- lead ordering. Ads that looked expensive move up. Ads that looked efficient turn out to be filling the pipeline with noise your team then pays for in wasted calls.

Sending the result back

There is a second benefit that is easy to miss. When conversions are reported back to Meta (not form submissions, but the leads that actually progressed), the platform’s optimisation starts targeting people who look like your real buyers rather than people who look like form-fillers.

That feedback loop only works if the outcome is recorded against the specific ad in the first place. Which brings it back to the same requirement: attribution at ad level, recorded automatically, not reconstructed from memory at the end of the month.