The method behind the verdict.
Every result is classified by a written rule, and the whole thing runs as a loop rather than a one-time audit. Here's the full mechanism — first the scan, then the Gap Report, then the loop that holds your position.
Real queries. Real responses. No simulation.
You give six inputs. The scanner runs real buying-intent queries against all four engines and classifies what comes back. Every verdict is derived from a live API response — and if an engine can't be reached, that result is clearly labelled as a demo placeholder, never passed off as live.
What you provide
- Company website — grounds the analysis in your domain.
- Full name — personalises the Gap Report.
- Professional email — where results and the report are sent.
- Company name — never included in the queries; a separate judge checks whether the engines' answers surface you unprompted.
- City & region — builds geographically accurate queries.
- Area of operation — the highest-leverage input. Be specific: "rhinoplasty," not "plastic surgery."
What the scan returns
- A classification for each engine — one of four states, by written rule.
- A 0–100 visibility score per engine, plus a cross-engine average.
- Competitors surfaced by name where a rival took your position.
- Your primary gap — the single category most blocking recommendation.
- A one-line reason behind every verdict, in plain language.
OpenAI
Widest consumer familiarity.
Strongest for local signals.
Perplexity
Search-native answers, always live.
Anthropic
Strong reasoning and analysis.
Every score is traceable. Every gap maps to a fix.
Each engine is asked three buying-intent questions; we read its actual answers and grade whether you appear, which sets a 0–100 visibility score per engine and an overall average across all four. Each result is assigned exactly one gap category — no free-text guesswork.
| Classification | What it means |
|---|---|
| Recommended | Named and actively recommended — the highest-value result. |
| Mentioned | Present but passive — visible, not endorsed. |
| Cited | Used as a source — content authority, not recommendation authority. |
| Excluded | No signal detected — invisible for the query. |
Weak Local Relevance
The engine can't confidently tie you to a specific geography.
Thin Service Pages
Your pages lack the structured, extractable content engines can read.
Missing Citations
You're absent from the third-party sources engines draw on.
Low Review Density
Too little review volume or recency for recommendation logic.
Weak Entity Clarity
The engine can't identify you as a distinct, well-defined business.
Competitor Dominance
One rival's signal density crowds you out of the answer.
Each gap maps to a specific, repeatable fix — the detailed plan for yours is reserved for your Gap Report.
A founder-reviewed diagnosis, not an instant readout.
The scanner shows scores and verdicts. The full 11-section Gap Report is prepared separately, checked against public evidence, and personally reviewed by the founder before it reaches you — within 24–48 hours, from a personal address, not a no-reply.
Specific to you
Built on your actual scan results — classifications, competitors, gaps and scores. Never a generic template.
Verified, not asserted
Claims about you and your competitors are checked against what is publicly findable, and anything unverifiable is softened.
A full diagnosis
Competitor comparison, prompt intent matrix, authority and structure gaps, and a prioritised fix plan.
A human quality gate
Every report is reviewed and sent personally by the founder. Nothing is auto-delivered.
AI visibility is measured by sampling live engine answers across multiple buying-intent prompts. Movement should be read directionally — month-to-month change can reflect optimisation work, competitor activity, or platform updates rather than any single cause. This caveat appears in every client report.
We close the highest-impact gaps for you.
The Gap Report names what's holding you back. The Foundation Build is the one-time, fixed-scope engagement that fixes the priority gaps — making your most important pages easy for engines to read, extract from, and trust.
Pages rebuilt for extraction
Up to 5 priority pages restructured for answer extractability, with FAQ blocks engines can lift directly.
The markup engines read
Organisation/LocalBusiness and FAQPage schema deployed across the selected pages, plus internal linking fixes.
A consistent, recognisable business
One primary business profile verified and made consistent, delivered with a before/after baseline report.
A system that holds your position, not a one-off audit.
Answer engines change weekly. The only thing that holds against that is a loop run on a cadence — observe, diagnose, optimize, distribute, verify, repeat.
Observe
Run real buying-intent prompts across four engines and classify each result against the four-state taxonomy.
Diagnose
Assign each result one of six fixed gap categories, so every finding ties to a concrete primary fix.
Optimize
Restructure priority pages for answer extractability and roll out the schema engines read.
Distribute
Place off-site authority signals on trusted channels, governed by a test-and-learn framework.
Verify
Re-sample prompts on a cadence; report mention, citation and recommendation rates directionally.
Repeat
Engines shift constantly. The loop defends the position and keeps adapting as they change.
Run the method on your own business.
One scan, four engines, real queries. See your verdicts now and get the founder-reviewed Gap Report within 48 hours.