Which ads produced signed roofs?

Free-inspection offers produce leads cheaply and convert poorly. A channel that generates a hundred inspections and four contracts is worse than one generating twenty inspections and eight, and no cost-per-lead report will tell you that.

Matching signed contracts back to their original lead source ranks channels on the only number that matters — installed contract value — and usually reorders them substantially.

A homeowner researches roofing in May, does nothing, then signs in October after a storm. The click is five months old, the cookie is gone, and any platform-reported attribution has expired.

Because the match is on the homeowner rather than a session, the October contract is still credited to the May lead. It also means a season that has already ended can be analyzed properly, which for a weather-driven trade is the only way to compare years.

A storm-restoration roof and a retail replacement are different products with different cycles, different margins and completely different acquisition. Storm work arrives in a concentrated burst driven by weather and canvassing; retail work is a considered purchase researched over months.

Running them through one report produces an average job value and an average cycle that describe neither, and it makes the advertising look wildly effective in a storm year and broken in a quiet one.

A job-type column mapped as Product / service separates them so retail demand generation can be judged on its own terms — which matters most in exactly the years when there is no storm to carry the number.

On an insurance job the homeowner is the customer and the carrier sets the amount. Approved scope, supplements and depreciation all move the final figure after the sale, sometimes months after.

That makes contracted value a poor measure and collected revenue the honest one. Exporting what was actually collected also keeps the report reconcilable to the company's own accounts, which is what makes it survive a conversation with the owner.

A payer column mapped as Product / service separates insurance work from retail so the two are never blended, and shows how much of the self-funded retail revenue traces to a channel at all.

Most roofing companies run both a media budget and a field operation — canvassers, storm chasers, door knockers — and the two are managed by different people, measured in different units and never priced against each other. The field team reports doors knocked and appointments set; the marketing budget reports leads and cost per lead.

Neither reports collected revenue per dollar spent, which is the only comparison that could settle how the next hundred thousand should be split between them.

Adding a canvassed list to the lead file, each row tagged as canvassing, puts the field operation on exactly the same axis as paid search. The answer may be uncomfortable in one direction or the other, and it is more useful than the argument it replaces.