Methodology
How VisiScan measures AI visibility
TL;DR: VisiScan asks real buyer questions to up to four AI engines, repeats those questions, detects business and competitor mentions, records citations, and turns the observed answers into a transparent score.
What a scan measures
- Whether the scanned business is named in answers from ChatGPT, Claude, Perplexity, and Gemini.
- Which competitors are named instead, including the position and context observed in the answer.
- Which websites and domains the engines cite as supporting sources.
- Paid reports add Brave grounding, competitor citation gaps, cited-answer verification, and applicable local profile evidence. These are evidence surfaces, not a fifth primary AI engine.
- More than 30 on-site and off-site AI visibility signals used for readiness recommendations.
How a scan works
- VisiScan profiles the business website and identifies its market and buyer context.
- It generates 5 buyer-intent questions for the free scan or 12 for the full report.
- Each question is sent to up to four engines (ChatGPT, Claude, Perplexity, and Gemini). Full reports run 3 samples per question and engine.
- The returned answers are checked for the business name, competitors, and cited sources.
- The report combines those observations with readiness checks and cites the underlying evidence.
How the score is calculated
Visibility is the weighted percentage of tested question-and-engine cells where the business is named. Trust and comparison questions are weighted 1.5x because they are closer to a purchase decision. Best-near-me, emergency, and price questions use 1x weight. Repeated samples are majority-voted per question and engine, and confidence intervals are shown where sample size matters.
The separate paid Citation Score weights query coverage at 50%, reciprocal rank at 20%, competitive share at 20%, and surface coverage at 10%. If a component fails or does not apply, it is excluded and the remaining weights are renormalized. Brave cited-answer verification is additional evidence and does not change the four-engine Visibility Score.
Data provenance and freshness
Every answer has an engine, sample, timestamp, and measurement mode. Live mode uses configured provider APIs; replay or unavailable-provider data is labeled in the report. Results are observations of a changing AI response, not a permanent ranking or a guarantee of future recommendations.
Plans
| Plan | Coverage | Price |
|---|---|---|
| Free scan | 5 questions × up to 4 engines | $0 |
| Full report | 12 questions × up to 4 engines × 3 samples | $49 one-time |
| Monitor | Automatic weekly re-scans and trends | $29/month |
Frequently asked questions
What does VisiScan measure?
VisiScan measures whether ChatGPT, Claude, Perplexity, and Gemini name a business in buyer-intent answers, which competitors they mention instead, and which sources they cite.
How is the Visibility Score calculated?
The score is a weighted named-rate across tested questions and AI engines. Trust and comparison questions receive 1.5x weight; best-near-me, emergency, and price questions receive 1x weight.
Are VisiScan results real AI answers?
Live scans query configured AI providers and record the returned answer, engine, sample, and provenance. A report identifies its measurement mode so replay or unavailable-provider results are not presented as live data.
How is the paid Citation Score calculated?
The paid Citation Score weights query coverage at 50%, reciprocal rank at 20%, competitive share at 20%, and surface coverage at 10%. Failed or inapplicable components are excluded and the remaining weights are renormalized.