Enter the domain
We normalise the URL, validate the domain and run a limited scan of public pages only. The crawler rejects private addresses, localhost, cloud metadata and suspicious redirects.
How it works
Eight steps, each with a verifiable output. No step claims a result it cannot show.
Confirm the brand profile
Name, products, value propositions, tone, forbidden claims, colours. Every field indicates whether it was extracted from the site or is a hypothesis. Nothing becomes a fact without your confirmation.
Choose markets, languages and competitors
No country is set automatically: the server location or the browser language does not prove commercial eligibility in a market.
Keywords and AI questions
We propose candidates by intent. You can add, edit, exclude and group them. AI-generated candidates are kept distinct from keywords with measured volume.
Baseline with real data
We connect only what is available: if Search Console or the data provider are not configured, the baseline shows real coverage instead of inventing numbers.
Editorial plan and first article
The plan links intent, relevance to the offer, existing coverage, gaps versus competitors and available evidence. If essential facts about the offer are missing, the product asks for clarification instead of inventing them.
Approval and publication
Editor with versions, quality checks, AI change diffs and approval tied to a precise version. Publication stores channel, external ID, URL and outcome.
Measurement over time
Search Console, AI visibility, citations, verified backlinks. Comparisons only between homogeneous windows, with a warning when prompts, engines or sample change.
Start from your domain
The first scan is limited to public pages and does not require credentials. You will immediately see what was extracted and what remains a hypothesis to confirm.