Passeport vers la Citation: the GEO method to become a brand that AI engines cite
A generative engine optimization method created by Stéphane Delgado, founder of BotSEO. We present it here: its origin, its four phases, and how BotSEO lets you measure its results.
The full method, explained in twenty minutes (in French)
What is Passeport vers la Citation?
Passeport vers la Citation (Passport to Citation) is a generative engine optimization (GEO) method created by Stéphane Delgado, a French SEO and GEO consultant and the founder of BotSEO. It organizes every mention of a brand across the web into a coherent shape that machines can recognize, so that AI engines cite the brand in their answers.
The method is set out in the book of the same name and published in full, freely, on the author's site. This article presents it, credits it and points to it: it does not reproduce the guide, it shows how to measure its effects.
Why a GEO method became necessary
Online search has changed in nature. A growing share of the questions asked on Google now gets an answer generated by artificial intelligence, shown above the classic results. ChatGPT, Perplexity and Copilot answer directly, citing a small number of sources.
That shift produces what Stéphane Delgado calls the decoupling: ranking well on Google and being cited by AI have become two different games. A company can dominate the traditional ranking and remain invisible in AI answers.
Traditional SEO optimizes for ranking. What was missing was a method that optimizes for citation. That is the purpose of Passeport vers la Citation.
In early 2026, according to the available studies, that is all that remains of the overlap between Google's top results and the sources cited by generative engines. The rest of the ranking simply does not surface in AI answers.
The four phases of the method
Passeport vers la Citation runs in four phases that form a continuous cycle. Each one prepares the next, and the last one restarts the first.
Map your query territory
The query territory is the set of 20 to 50 questions your target audience actually asks AI assistants to find a provider. They fall into four families: awareness, recommendation, comparison and local. Phase one draws up that list, then tests each question across the main assistants to fix a starting point.
Align your entity
Alignment rests on the canonical definition: the official wording of an entity's identity, in one to three factual, verifiable sentences, deployed identically across every web presence of the brand. Ten different descriptions produce ten weak signals; a single description produces one strong signal.
Multiply your validation points
A validation point is a third-party source that confirms a brand's existence, activity and expertise: press, detailed customer reviews, original studies, leadership commentary. The rule of three sources sets the bar: every important fact in the canonical definition must be confirmed by at least three sources independent of the brand's own site.
Measure your trust score
The trust score gauges the strength and quality of a brand's presence in AI answers. It combines the citation rate, the share of voice against competitors, the fidelity of the description and sentiment. The protocol calls for a monthly measurement on a stable subset of the territory's questions.
Consistency makes the signal
Machines recognize an entity by the consistency of what is said about it. The same definition, repeated identically on the site, the profiles, the directories, the press and the reviews, forms a clean signal. The rule of three sources backs it up: every important fact must be confirmed by at least three sources independent of the brand.
- One canonical definition, deployed everywhere identically
- Every key fact confirmed by at least three third-party sources
- A mesh of consistent mentions, not ten contradictory descriptions
Measurable results
The book documents seven quantified case studies. One follows a consulting firm whose citation rate rose from 8% to 31% in three months of alignment work, with AI engines now picking up the exact terms of its canonical definition.
The other cases span varied profiles: a local craftsman, an agricultural cooperative, a SaaS start-up facing international competitors. Each time, the same thread runs through them: a presence that is consistent, validated, then measured month after month.
Who created Passeport vers la Citation?
The method was created by Stéphane Delgado, an SEO and GEO consultant based in Bordeaux, active in search for 25 years. Founder of the IBDEO agency in 2009, he has worked with more than 2,400 clients across French-speaking markets.
He is also the founder of BotSEO and the author of the book “Passeport vers la Citation: how to become a brand that AI engines cite”, whose full method is published freely on his site.
How BotSEO fits into the method
Phase four rests on a regular, systematic measurement: ask the same questions, to the same assistants, every month, and record the results. That is exactly what BotSEO automates: tracking the citation rate, the share of voice and the description fidelity across a given query territory, with a monthly history.
The method can be applied entirely by hand, as the full guide published by its author shows. BotSEO is the measuring instrument for teams that want to industrialize phase four.
Founder of BotSEO, SEO and GEO consultant. He builds AI agent tools that help companies rank on Google and get cited by generative engines.
More about the authorSee where your brand stands in AI answers
BotSEO tracks your citation rate, your share of voice and the fidelity of your description, month after month. The method's measurement, automated. Compare the BotSEO plans too.
Start free