//

What is GEO: how AI answers questions about you — and whether you can influence it

What is GEO: how AI answers questions about you — and whether you can influence it

More and more people ask ChatGPT rather than Google: “who is …”, “can … be trusted”. The answer is put together by a model — and the usual rules of SEO don’t apply to it. The discipline that deals with presence in AI answers is called GEO. We explain how it works, how it differs from SEO and what to do if AI tells untruths about you — or says nothing at all.

In short. GEO is the optimisation of your presence in AI answers: so that ChatGPT and Perplexity choose you as a source and retell you correctly. The lever you can control is live-search sources: clear facts, consistency, authoritative platforms.

definition

What GEO is, in simple terms

GEO (generative engine optimization) is the optimisation of presence in the answers of generative AI systems: ChatGPT, Gemini, Perplexity and the AI modes of search engines.

If SEO competes for positions in a list of links, GEO competes for something else: for the system to choose you as a source and retell you correctly — after all, an AI user usually doesn’t click through links but reads a ready-made answer.

Not to be confused with geo-targeting in advertising: the overlap of abbreviations is a coincidence.

plan

What to do: practical steps for your own name

01

Run an AI audit of yourself.

Ask the main systems about yourself — in different languages and with different wording (“who is …”, “can … be trusted”, “what is known about …”). Record the answers and the sources they cite.

02

Find the root of the errors.

Fabrications and confusion arise where reliable sources are scarce: the model fills the void on its own. Each error will point to a gap in the field.

03

Put your sources in order.

One version of the facts everywhere: profiles, reference pages, publications. Explicit wording: “founded in 2011”, not “has many years of experience”.

04

Give the models something to cite.

Pages with direct answers to direct questions about you, authoritative mentions, structured data. It is the same work on the information field — for a new reader.

05

Repeat the audit.

Systems are updated; the picture in the answers is not a result but a state that has to be maintained.

mechanics

Where AI gets its answer about a person

The answer is made up of two layers.

Training data is what the model absorbed during training: a snapshot of the internet at a certain point in time, including materials that were removed long ago. This layer cannot be controlled directly. Live search is what the system finds at the moment of the question and cites with links. This layer can be controlled — and this is exactly where GEO works.

What models choose to cite: sources with clear facts (dates, names, figures stated explicitly), that are consistent with each other (one version of the facts across different sources), authoritative (reference resources, prominent media, Wikipedia) and structured (direct answers to direct questions, data markup). A model doesn’t cite vague marketing copy — there is nothing in it for the model to take.

differences

GEO vs SEO: what has changed

Unit of result: in SEO, the position of a link; in GEO, a mention and an accurate retelling in the answer.

Currency: in SEO, the click; in GEO, the model’s trust in the source.

Errors: an SEO error means you rank lower in the search results; a GEO error means the model confidently tells a fabrication about you or confuses you with a namesake, and the user reads it as fact.

Speed: search results can be shifted with weeks of work; AI answers change along with their sources and system updates — it is a longer and less deterministic game.

The tools overlap — publications, reference resources and structured data work for both disciplines — but the criteria for success are different.

question

FAQ

Can information be removed from ChatGPT?

From the training data — practically no. From the answers — indirectly: by changing the live-search sources. Some providers have procedures concerning personal data — they work for certain categories.

AI makes up facts about me. Can this be fixed?

Yes — by saturating the field with reliable sources: hallucinations live in a void. The denser the field of facts, the less the model invents.

Will GEO replace SEO?

No — they coexist: some background checks remain in classic search, some are migrating to AI. The field has to be prepared for both readers.

When should I deal with this — now or “when it goes mainstream”?

Now: the field is still empty, and whoever builds a presence now will become the default answer — and correcting someone else’s “default answer” later is much more expensive.

next step

What AI says about you — right now

The most useful thing you can do after reading this article is to complete step 1 and see your AI picture.

And if it needs work — that’s our territory: Reputation in AI and GEO optimisation →

Read also

›

What online reputation is and what it is made of

›

What Is SERM: How Reputation Management in Search Works

Gregory (Grygorii) Malenko portrait

Author — Hryhorii Malenko

Founder of SERM.UA, the first reputation management agency in Ukraine. Has been managing the reputations of private individuals since 2011.

This material is for informational purposes only and does not constitute legal advice.