Mention rate
The share of buyer questions where each model names you, overall and split by ChatGPT, Gemini, Claude and Perplexity. Includes position when mentioned, so you know whether you are the first suggestion or an afterthought.
Six sections, built from real responses captured for your business. Every figure links back to the question and model that produced it.
The share of buyer questions where each model names you, overall and split by ChatGPT, Gemini, Claude and Perplexity. Includes position when mentioned, so you know whether you are the first suggestion or an afterthought.
The businesses each model recommends most often for your questions, ranked by frequency and position. This is the competitor list the market actually sees, which is often different from the one you track.
Every incorrect claim about your business, quoted verbatim, with the model, the question that triggered it, and how many times it recurred. Grouped by severity: harmful, misleading, outdated.
Which websites, directories and review platforms the models rely on when they talk about your industry and your business. Sorted by influence, with the pages that need your attention first.
A crawl-based score for how well AI systems can read and trust your site: clarity of what you do and where, structured data, crawlability, freshness and consistency with third-party listings.
Prioritised fixes ordered by expected impact and effort: which source to correct, which page to rewrite, which schema to add, which listing to claim. Written for a marketing lead, not a data scientist.
EXCERPT
The summary sits on page one. Behind it are the full question-by-question tables, the quoted responses, and the source map, so nothing has to be taken on trust.
USING THE REPORT
Fix the record. Most errors trace back to one or two stale sources — an old directory entry, a review site with last year's hours, a page nobody has touched since the redesign. Correct those and the answers usually follow.
Claim the gaps. When a model never mentions you for "best for X" questions, it is usually because nothing on the web connects you to X in plain language. The readiness section shows exactly which claims your site is failing to make.
Track the right competitors. Whoever gets named instead of you is your competition in this channel, and it is not always who you would expect. Sometimes the name that keeps appearing is a business you have never tracked; sometimes your fiercest rival turns out to be invisible here too.
Measure change. Run the same 30 questions again after your fixes and compare. Monthly monitoring plans keep the question set fixed so the trend is meaningful.
FORMATS
Every report arrives as a PDF you can hand to a board or a client, plus a private web version with the full response tables behind it. Monitor plans also include a CSV of every captured answer, in case you want to slice it yourself, and a 45-minute walkthrough call with the person who ran the report.
Reports are in English by default. Spanish-language reports and question sets are available for businesses serving Spanish-speaking buyers; ask when you order.
One business, one industry, one location, 30 questions, four models.