Search behaviour moved. Customers now ask ChatGPT and Gemini for recommendations before they ever open Google. Most company websites are invisible in that moment — not because they rank badly, but because AI cannot verify who they are.
Most SEO tools were built to chase rankings on a search results page. That model is being replaced by a single generated answer, and the rules are different: AI cites sources it can verify, and it names businesses whose pages state clear proof — products, credentials, pricing signals and real customer outcomes.
Most company websites publish too little of that proof. So they lose the recommendation without ever knowing it happened.
Before scoring anything, a model reads your site and writes down what you sell, who you sell to and what proof you already publish.
Not fixed templates. The questions are generated from your category and market, the way a real customer would actually phrase them.
We look for the companies that occupy the same answer slot you want — even when you never compete for the same keyword.
No fifty-rule checklist. A short, prioritised list of what to change and why it matters to a customer.
Read the homepage and sample pages, then build a profile of your category and target market.
Generate customer questions, infer competitors, then sample AI answers and technical health in parallel.
Compare your mention and citation rates against the competitors occupying the same answers.
Store the profile against your domain so the next diagnosis skips discovery and tracks change.
Produce a plain-English report with prioritised actions and measured evidence.
For partnerships, data questions or feedback: [email protected]