What is GDO (Generative Distribution Optimization)?
Published on
GDO, which stands for Generative Distribution Optimization, consists of distributing a company's information across a network of relevant third-party sources, so that it is present where AIs go looking when the question does not name anyone. This is the term we use at FaireDuBruit to name this work: unlike SEO, and GEO which was born from a research paper in 2023, it is not an established industry term, and no external study measures it under this name. What it designates, however, can be measured: we no longer count the pages of your website, we count the independent sources that name you.
For what type of question? It all depends on the question asked to the AI. When a customer names your company, the AI reads your site, and the work is that of SEO and GEO. When the question does not name anyone—'which software for...', 'who to call for...'—the AI looks for sources that compare, test, and answer. This second case is what this page is about.
Why a third term after SEO and GEO
SEO seeks a rank in a list of links. GEO, which appeared in a paper presented at the KDD conference in 2024, seeks to get a page included in the response drafted by a generative engine, by working on the content of that page: clear answers, dated facts, detailed text. Google, for its part, considers that optimization for its AI answers remains organic search and recommends sticking to SEO best practices rather than tricks presented under the acronyms GEO or AEO.
These two approaches have one thing in common: they work on your website. However, in our measurements, for a market question that does not name any company, AIs mainly choose third-party sources—media, publishers, specialized sites they already know—and do not read the websites of the suppliers concerned. Optimizing a page is only useful if that page is read: a word was missing for the work that takes place outside your own website, and that is what we call GDO.
What GDO covers, in practice
Distributing, here, means publishing a company's information—what it does, for whom, with what limits—on distinct thematic sites, in the form of questions that its market actually asks, and in the language in which these questions are asked. Each page answers a specific question, using the words employed by the models in their own searches, and links back to the product listing.
This division into distinct sources is not a presentation choice. In our measurements from September 2026, an AI most often selects only a single page per domain in a response: 95% of Claude's responses, 92% of Gemini's, and 79% of ChatGPT's cite only one page per site, across three corpora and 166 responses. Therefore, adding a fifteenth page on the same site adds little; adding one more independent source adds another door.
What GDO does not promise
It does not promise any citations. An AI decides on its own what it searches for and what it cites, and does not give the same answer twice. What can be guaranteed are real, dated, and linked pages; what cannot be guaranteed is their inclusion, a rank, or a score.
Nor does it transfer a ranking. Moz, in its announcement of a study covering approximately 40,000 queries from February 2026, indicates that 88% of Google's AI Mode citations do not correspond to the organic top 10 for the same query. Ahrefs, in August 2025, noted for its part that only 8% of the pages cited by ChatGPT appeared in Google's top 10. Being well-ranked in a traditional search engine is therefore not enough to be cited, and being cited does not require being well-ranked there.
What we do not know
Our measurements cover the API of GPT-5, Claude Sonnet 4.6, and Gemini 3.5 Flash, not the consumer interfaces, which may behave differently. They describe what models search for and cite at a given moment; they do not measure the effect of a publication on citations, and an AI changes its response from one time to another. And a limitation specific to this page: 'GDO' is our term. It names a work we measure, not a category recognized by model publishers or by research.
Sources
- Aggarwal et al. — “GEO: Generative Engine Optimization”, KDD 2024, .
- Google Search Central — “Optimizing your website for generative AI features on Google Search”, .
- Moz — announcement of its study on Google AI Mode citations, approximately 40,000 queries from February 2026; the study page is not accessible to our reader, .
- Ahrefs, Louise Linehan — overlap between AI search and the top 10, .
- FaireDuBruit measurements, from September 4 to 16, 2026 — see "How we measured".