Visibility audit
We ask ChatGPT, Gemini, Perplexity and Google's AI summaries what your customers would ask, and record who appears, with which source and in what tone.
More and more customers skip Google: they ask ChatGPT or Gemini, or read the AI summary before they click. We make those answers cite your company: we audit how the models see you today, adapt your website and content so they can read it, and measure every month whether you show up.
It is not about ranking first in a list. It is about being the answer.
AI search engines do not rank links: they write an answer and cite whoever is clear, trustworthy and easy to read. That changes which content you need, how it is written and how it is measured. We work on those three things with a repeatable method, and without promising positions nobody can guarantee.
AI summaries already appear in close to half of all searches and cut clicks to the first result. If you are not in the answer, you do not exist for that customer.
Four blocks that repeat every month: measure, fix the technical base, publish citable content and check the effect.
We ask ChatGPT, Gemini, Perplexity and Google's AI summaries what your customers would ask, and record who appears, with which source and in what tone.
Structured data, an llms.txt file, permissions for AI crawlers, speed and an architecture the models can navigate.
Guides, comparisons, FAQs and product pages with figures and authors, written so a model can summarise them without dropping your brand name.
Tracking of mentions, citations and traffic referred by assistants, alongside classic organic, to decide what to publish next month.
Concrete deliverables, not a slide deck.
It works best where customers research before they buy.
Dermocosmetics, health, equipment, software: categories where the customer asks "which is best for…".
Consultancies, law firms, industry: when the AI recommends suppliers, being on the shortlist is everything.
If organic traffic drops with no clear reason, the reason is usually that the answer is now given without a click.
A two-week start, then a monthly rhythm.
Two weeks to measure how assistants see you, review the website and deliver the prioritised plan.
We fix what stops the models from reading you and publish the first batch of content.
We measure, publish and adjust. Every month you know which questions you have won and which are missing.
What we get asked before starting.
Citable content comes out of Creative Studio, and AI video feeds the same questions.
Tell us your sector and your three competitors. We will send back a first reading of how assistants see you and what it would take to get into the answer.