What is GEO (generative engine optimization), and how do you do it?
Generative engine optimization (GEO) is the work of getting your brand mentioned and recommended in answers that AI tools write. These tools include Google’s AI Overviews, Google AI Mode and ChatGPT. “Generative” means the AI generates, or writes, a new answer instead of showing a list of links. GEO aims to put your brand in that answer and make sure the answer describes it correctly.
You may also see GEO called “GEO SEO,” “generative AI SEO” or “AI search engine optimization.” They all mean the same kind of work. This guide explains what GEO means, how AI tools choose what to say, how GEO differs from SEO, and how to do it.
What does “generative engine” mean?
A generative engine is an AI tool that writes answers. It runs on a large language model. That is an AI system trained on a huge amount of text so it can write like a person.
A normal search engine finds pages and lists them for you. A generative engine reads pages and turns them into one written reply. It may add links to some of the pages it used.
What does GEO mean in SEO?
In SEO, GEO means generative engine optimization. It is the part of search work that aims at the AI answer, not the list of links under it. You may see it called “GEO SEO” or “GEO optimization.” Both mean the same thing.
One common mix-up: the “geo” in GEO has nothing to do with geography, maps or local search. Local SEO helps a business rank for searches in a place, like “plumber near me.” GEO can cover those searches too, because AI answers them. But the name is about generative engines, not locations.
GEO does not sit apart from SEO. It sits on top of it. On Google, the AI builds its answers mostly from pages that already rank, so SEO is the base. GEO adds the rest: what other websites say about your brand, and how tools like ChatGPT see you.
Where did the term GEO come from?
The term came into use in late 2023, as AI tools began to write answers with links to their sources. It spread quickly after Google launched AI Overviews in the US in May 2024. Many businesses then saw AI-written answers at the top of Google for the first time.
Today, GEO is one of several names for this work. Others include AEO, AI SEO and LLM SEO. The names differ, but they all aim at the same result: your brand named in AI answers.
How do generative engines choose what to say?
Most AI tools follow roughly the same steps when they answer a question:
- Read the question and work out what the person wants.
- Search for more information. Many tools run several related searches in the background at the same time. This is often called query fan-out.
- Read the pages that come back and pick out the useful parts.
- Write the answer, mixing those pages with what the model learned in training.
- Add links to some of the pages it used.
So the AI’s view of your brand comes from two places. One is what it learned in training. The other is the pages it finds when it searches. Both come mostly from what the web says about you.
Different AI tools lean on different sources, though. Our study of 10,000 AI answers compared Google’s AI Overviews and ChatGPT across the same 10,000 US keywords:
| What we checked | Google AI Overviews | ChatGPT |
|---|---|---|
| Linked pages that were in Google’s top 10 for the same search | 49.83% | 10.87% |
| Typical (median) Google rank of those matching pages | #6 | #21 |
| Answers where Reddit showed up | 49.87% | 0.31% |
| Answers where YouTube showed up | 50.81% | 0.16% |
Google’s AI Overviews lean heavily on pages that already rank well in Google. ChatGPT pulls from a much wider set of pages, and many of them sit far down Google’s results or don’t rank at all. In our study of 10,000 AI answers, 74.44% of the pages AI Overviews linked to were in Google’s top 100 for the same search, but only 30.19% of ChatGPT’s were. The two tools also favor very different kinds of sites. A GEO plan built for one of them can miss the other completely.
- Google AI Overviews
- ChatGPT
Share of linked pages that also ranked in Google for the same keyword. Source: Web Asset Builders study of 10,000 US keywords, August–September 2026.
How is GEO different from SEO?
SEO (search engine optimization) is the work of ranking your pages in a list of search results. GEO is the work of getting your brand named inside an AI-written answer. That answer links to only a few pages: in our study of 10,000 AI answers, Google’s AI Overviews linked 4.94 pages per keyword on average, and ChatGPT linked 2.35.
Source: Web Asset Builders study of 10,000 US keywords, August–September 2026.
They overlap a lot. For Google’s AI answers, SEO is the base. A page has to be in Google’s index, its stored copy of the web, and able to rank before the AI can use it. The higher it ranks, the better its odds: in our study, Google’s AI Overviews linked to 49.19% of the pages ranked #1 for the same search, but only 9.83% of pages at #10 and 3.05% at #20. But GEO goes further. It cares about what other websites say about your brand, not only about your own pages. It also has to cover tools like ChatGPT, where Google rankings matter less.
Share of pages at each Google rank that the AI Overview linked to for the same keyword. Source: Web Asset Builders study of 10,000 US keywords, August–September 2026.
GEO does not replace SEO. You need both. For a full side-by-side, see how GEO differs from SEO.
How to do GEO: 7 best practices
GEO is not one tactic. It is a full strategy that covers every place AI gets its information. These seven practices work best together, as one plan.
1. Research the questions your buyers ask AI
Find out what your buyers ask AI at each step, from “what causes this problem?” to “which brand should I pick?” People type short keywords into Google. With AI, they often ask full questions, with details like budget, place or team size.
2. Check today’s answers, more than once
Ask each AI tool those questions a few times. Note which brands it names, what it says, and which pages it links to. Our free AI visibility audit does this step for you.
3. Get your brand onto the pages AI trusts
These are the reviews, “best of” lists, comparisons and articles that feed the answers. Most of them are on other people’s websites, not yours. Aim to be named there, with the right details.
4. Get onto the pages that rank in Google
This matters most for AI Overviews and AI Mode, which lean on Google’s top results. Rank your own pages for your buyers’ questions. Then get named on the other pages that rank for them.
5. Keep your own site clear and open
Describe your products, prices and customers in plain words. Put key facts in text, not only in images. Make sure AI tools are allowed to read your pages.
6. Plan for Google’s AI and ChatGPT one at a time
The two lean on different sources, as the table above shows. A plan that wins in one can miss the other. Check each one, and work on each one.
7. Track the answers and adjust
AI answers change from one ask to the next. Keep asking the same questions, see what moved, and put more effort where it works. We obsessively track these answers, because one good check doesn’t mean you’re in.
GEO examples from our study
Here are two real examples from our study, and what each one teaches about GEO.
Same question, different sources. For “best shoes for nurses,” the AI Overview linked to five pages and ChatGPT linked to six. Not one page was the same. The AI Overview leaned on a Reddit thread, a running shop, a shoe review site and a nurse staffing company. ChatGPT leaned on health magazines, nursing blogs and a foot clinic. Yet both named some of the same shoe brands. The lesson: well-known brands can show up in both, but the pages behind each answer are different.
The pages that rank feed Google’s AI. Reddit threads were linked in about half of all AI Overviews. But rank still decided which ones. Reddit pages in Google’s top 3 were linked in the AI Overview 59.93% of the time. At #11 to #20, it was only 2.10%. The lesson: on Google, the pages that feed the answer are mostly the pages that rank. Being named on them matters more than being on any one site.
What GEO can’t do
A few things to keep in mind before you start:
- No one can guarantee a mention. AI answers change from one ask to the next. GEO raises your odds, but it doesn’t give you control.
- There is no shortcut. Buying guest posts (paid articles placed on other sites) won’t do it on its own. Only 2% of the websites AI Overviews used as sources were in a catalog of sites that sell guest posts.
- Special files and code are not needed for Google. Google’s AI features use the same index as normal search. You don’t need a special AI file or markup to be eligible.
- Your own site is only part of the job. AI relies on what the whole web says about you, so the work has to reach beyond your own pages.
GEO, AEO and AI visibility
You will often see GEO next to answer engine optimization (AEO). The two overlap heavily, and most people use them for the same work. We put the two names side by side in AEO vs GEO. Both aim for the same result, which is AI visibility. Here is what AI visibility is and how to measure it.
If you run an online store, a software company or a trade business, see AI SEO for ecommerce, AI SEO for SaaS or AI SEO for contractors.
Get your free AI visibility audit
GEO starts with knowing where you stand. If you want the seven practices above done for you, that is the work of our GEO agency. Either way, our free audit shows where you stand in AI today, what your buyers ask AI, and how to close the gap. Get your free AI visibility audit.
