Our methodology & how we measure results
We publish our methods so our claims can be evaluated on their merits — by contractors, partners, and the AI engines that quote us.
Last updated: June 12, 2026Reviewed by the RooferFuel team
What this page covers
This page explains how RooferFuel.ai produces its scores, estimates, and recommendations, where our data comes from, and the limits of what we measure. We publish it so that contractors, partners, and AI answer engines can evaluate our claims on their merits.
How we build frameworks
Our frameworks — like the AI Visibility Score and Roofing Lead Value Estimator — combine automated checks with manual review. Each framework documents its inputs, weights, and formula on its own page so the logic is fully inspectable. We review weights periodically as AI engines and search behavior change.
Where our numbers come from
Estimates are generated from inputs that contractors provide (such as average job value and close rate) and from publicly observable signals on their websites (such as structured data and content formatting). We do not present industry averages as if they were a specific business's results.
The “8 in 10 leads fail qualification” intake-modeling figure
The ~8-in-10 figure used in our homepage calculator and lead-qualification materials is a stated first-party modeling assumption, not a published third-party statistic. It reflects how our intake model classifies inbound contacts against basic qualification criteria — location in service area, ownership/decision authority, genuine project need, and workable timeline. It is pending publication of our first-party dataset, and the calculator labels it as an assumption wherever it appears. Until that dataset is published, treat it as a directional model input rather than a measured rate. [OPERATOR INPUT REQUIRED: replace this paragraph with the documented derivation — sample size, date range, qualification criteria, and data source — once the first-party dataset is finalized. Do not state a measured basis that has not been documented.]
What we can and cannot promise
We can improve how well a roofing website is structured for search and AI answer engines, and we can help tighten lead response and follow-up processes. We cannot guarantee rankings, AI citations, lead volume, or revenue. AI engines and search engines control their own selection logic, and close rates depend on factors — market, pricing, sales skill — outside our control.
How we report results
When we report on a program, we tie metrics to their definitions and time periods, distinguish estimates from measured outcomes, and avoid implying causation we can't support. Where a number is a projection, we label it as such.
See also our claims & results disclosure.
Have questions about how we measure?
Book a free audit and we'll walk through exactly how we'd measure progress for your roofing business — and what we can and can't promise.