Where growth breaks: attribution

We can't tell which agency or channel is actually working.

Every agency and platform reports on its own piece, and each one counts differently. None of the reports agree with each other, or with the CRM, so none of them answers the question you're asking.

What's usually behind it.

This is the fog, not anyone's fault. Each partner measures its own piece, usually to the lead and rarely to the signed contract. The same source goes by different names in the CRM. Brand spend gets blended into channel math.

So every report is defensible on its own, and nobody reports on the whole.

Where the good leads come from.

A PE-backed roofing platform. Share of 2026 leads that reached approved work, Jan 1 to Sep 9. Same window for every source; recent leads are still converting.
What a clear read shows4.3x

Search leads reached approved work at 4.3 times the rate of third-party leads.

How I approach it.

I bring every source, partner and number onto one clear read of spend and return, traced to signed contract value by lead source. Not to the lead, and not just to "closed," which lags and misreads any recent month.

I keep brand spend out of channel math, so each channel is judged on what it returns, and you and your board get one number everyone can stand behind. And because I built and sold an agency, I know how to get the best from one. That's often a better outcome than firing it.

Cheaper per lead. Not cheaper per sale.

A PE-backed roofing platform. Working spend and approved jobs. Sources with $50K or more in working spend.
What a clear read shows$2,774

per approved job from search and website leads. Bought leads looked 73% cheaper per lead, then cost $3,994 per approved job. The second number is the one worth managing to.

Proof

The vendor wasn't the problem.

A purchased lead source showed 64% of its leads were never reached. The easy read was a bad vendor. It was how the leads were being worked. We changed when the calls went out, sent an intro text before the first call, and rewrote the scripts. The no-contact rate came down to about 29%.

And when several vendors and tools couldn't give a straight answer on what was working, I built the measurement stack myself.

One source of truth3 brands

brought onto one lead-source list that marketing, the call center and the sales floor all work from.

A cheap lead isn't cheap if it never becomes a sale.

Every month without a clear read, budget drifts toward whichever partner reports best.

Let's find out which of yours is producing contracts.

One read of spend and return, so the next budget decision comes with an answer instead of a debate.

Book a working conversation

Thirty minutes on the challenges ahead, the resources you have today, and a plan.