1. Identity signals
Name, service, location, and contact details that identify your business unambiguously across your site and, where available, directory sources.
Technical audit →VeriSight · AI search
AI search (generative answers in Google, ChatGPT, Perplexity, and other systems) is changing what people see before they contact you. VeriSight measures the signals that decide whether your website is a credible, citable source — and what a diagnostic can and cannot prove.
When someone asks an AI assistant a question about a topic, the assistant may quote, summarize, or recommend sources. That behavior is not a search ranking that a diagnostic can read. It is produced inside a model, at inference time, by a provider that does not publish how it selects sources.
What you can measure is the evidence your website leaves behind:
What you cannot measure: whether a specific AI system was shown your website, whether ChatGPT or Perplexity "recommends" you, or whether a generative answer included your content. Those outcomes depend on model training, provider policies, and real-time selection rules that are not observable by the website owner.
A report that claims to know what ChatGPT or Google AI Overviews will say at 3 p.m. would have to guess at model behavior. VeriSight does not guess. Instead, it shows you what your website does and does not make visible to a machine reader, so you can improve the signals that AI systems can actually use.
Name, service, location, and contact details that identify your business unambiguously across your site and, where available, directory sources.
Technical audit →Schema markup (Organization, LocalBusiness, Product, FAQ, etc.) that describes your pages to machines.
Metadata & schema →Whether AI crawlers can access your pages without blockers, and whether the page renders into a clear answer.
Crawlability →AI answer presence, rankings, recommendations, and model behaviour. Nothing in VeriSight claims to predict those.
Run a comparative audit →Illustrative example (labeled, not a real audit)
Imagine a business with a homepage that states its name, services, and address; a services page with a clear question-and-answer format; and a LocalBusiness schema block. An illustrative scan would report strong entity clarity, valid structured data, and pages that render completely for machines. The same illustrative report would also state clearly that it cannot say whether an AI system chose that business answer, because the answer depends on the model and the user's question — not on the website alone.