Commercial Audit
For service pages, city pages and listings, against the live top 5: what the buyer checks before ordering and whether your page answers it concretely, where it stops selling, and which of your claims have proof next to them.
Compare against specific pages instead of the live top Optional, 2–10
Two coordinate systems on one page. The buyer tree lists what someone evaluates before ordering — timeline, deliverables, pricing model, who does the work, platform segments. The SERP shows what the top already says. Where they agree it's a must-have; what only the SERP repeats is optional consensus; what neither covers is your whitespace. Concreteness is checked per branch: tools, formats, numbers — present or not. No density, no score.
Every "expert", "leading", "years of experience" is a claim. We look for a passage on the same page that actually entails it — a case with numbers, a named process, a date. Unsupported claims are listed, not scored. Alongside: whether your team has names and roles, whether the page carries time markers, and whether the first screen states the offer, a price anchor, a fact and an action.
Parity with the top
No worse on the table: blocks, schema depth, card attributes — each weighted by how many of the ranking pages have it. This used to be the whole report; now it's the floor, not the goal.
A feature 4 of 5 competitors share is table stakes; a feature only you have is your edge.
Zones are calibrated on the top's own passages — never fixed thresholds.
This tool measures your page. For the local pack, most of the weight sits off the page — and for AI answers the balance flips back toward on-page. Both columns below are survey weights, not our invention.
For being cited in AI answers, on-page rises to the top slot — the work this audit points at weighs roughly twice what it does in Maps.
What it checks
What the buyer checks
Timeline, pricing model, who does the work, deliverables: each branch is marked must-have, optional or nobody-has-it against the top 5, with whether your answer is concrete (a number, a format, a name) or only words.
Where the page stops selling
Each paragraph is compared with the commercial and the informational pages of this same top and marked by the one it sits closer to. Strike paragraphs out and watch the balance move: no credits, no rewrite pushed.
Which claims have proof
“Expert”, “leading”, “15 years”: each claim is checked for a passage on the page that backs it, such as a case with numbers, a named process or a date. Unsupported claims are listed, not scored.
What the report looks like
What buyers in this SERP expect and whether you answer it, where the page stops selling, and which claims stand without proof.
What we don’t do
Clear limits, so you know what the audit judges and what it leaves to you.
Invent scores or weights
No page score or readiness percentage. A feature weighs as much as the share of the top 5 that has it; claims are a list with a count, each traceable to the SERP or your page.
Promise rankings or AI answers
For local queries most of the weight sits in your Business Profile and reviews, off the page. For “which provider” prompts, assistants lean on listicles; whether they cite you is in AI Visibility.
Ask you to rewrite boilerplate
“Call us today” may read like everyone else’s, and that’s fine. Uniqueness is measured only on proof (cases, numbers, process), never as a keyword-density percentage.
Score your design
No screenshot beauty contest. Filters, calculators and maps count as user value the top also offers, and layout is read as where content sits, not how polished it looks.