An AI Visibility audit: what we assess, what the report looks like and what to do with it

The report shows whether the brand appears against the agreed buying questions, in what context it is described, which sources are cited and which gaps carry the highest priority. The number of models, questions and comparison brands follows from the scope agreed before the work starts.

What the audit assesses

An AI Visibility audit has four main elements. Every result stays linked to a complete answer record.

Element 1: visibility analysis across AI systems

We test an agreed question set matched to the category, the market and the buying intent. For each run we record the question, the complete answer, the model, the mode, the date, brand presence, the context of the mention and the sources cited.

Element 2: competitor benchmark

Comparison brands can be nominated by the company or drawn from the models' answers. We analyse them on the same question set and under the same rules. That gives the result context without pretending to be a universal truth about the whole market.

Element 3: content assessment for citability

We analyse your key pages - the home page, service pages and articles - against four criteria: clarity of specialisation, question-answering structure, citable passages of 130 to 170 words containing specific data, and technical barriers for AI crawlers such as robots.txt and JavaScript rendering.

Element 4: external mention analysis

We check how your company is mentioned beyond your own site: industry publications, LinkedIn, reviews and mentions in articles. We compare that with competitors and identify which mention channels give them a visibility advantage.

What the report looks like

The report combines an executive summary, a measurement table, the context of each mention, the sources cited, the content analysis and the priorities. Any result can be traced back through a record identifier, the complete answer and the version of the method.

What to do with the report

The report identifies priority actions based on the gaps found in the measurement. That might mean tidying up the offer, rebuilding pages, correcting structured data, developing sources or planning the next experiment.

You can implement the recommendations yourself or move on to an AI Recommendation Sprint. The free stage before the audit serves only to check category fit and select the right product.

How long does the audit take?

The timeline depends on the number of models, questions and markets, and on the benchmark required. We confirm the scope, the output format and the delivery date before the paid analysis begins.

Who benefits most from a diagnosis?

• B2B owners who hear from clients that they "found them through AI" - or, more often, never hear it and do not know why
• Marketing managers who want to understand how AI affects their funnel and where leads are being lost
• Marketing agencies assessing the potential of a new service for their clients
• Companies preparing to implement an AI Visibility strategy and looking for a starting point

A diagnosis makes little sense for B2C businesses with a short buying cycle, for e-commerce built on impulse purchases, or for companies whose buyers do not use AI to find suppliers. For everyone else it is the fastest and cheapest way to understand your situation.