The blog
Long, technical, checkable guides on AI visibility. Every external link published here was fetched and answered 200 before it shipped — the rule we apply to the files we generate applies to us first.
Why Moroccan businesses vanish from generated answers, and how to fix it: identity across three scripts, +212 phone formatting, Ramadan hours in markup, local corroboration that actually exists.
What llms.txt is, what a good one looks like, who really reads it, what Google's May 2026 documentation says about it, and three commands that tell you whether yours works.
Travellers no longer compare ten tabs — they ask an assistant for an itinerary. What a riad, a tour operator or a transfer company must publish to appear in that answer, and what keeps them out.
How an answer engine picks its sources, the six conditions that decide whether you get cited, and how to measure the outcome honestly instead of buying an invented score.
Which AI robots visit your site, which ones train, which ones index, which ones are triggered by a human — and the robots.txt configurations to copy once you have decided.
The @id graph pattern, the types that actually matter to answer engines, the consistency rule almost every site breaks, and a complete example to adapt.
One URL is enough. The audit returns the score, the twelve checks and the detail of every failure. The first file generation is on us.
Custom project or enterprise
By phone
REPLY WITHIN 48H · ENGLISH & FRENCH
© 2026 Indexonic. All rights reserved.