---
title: "GEO vs SEO: what changes, what does not, and what TIN measured"
description: "GEO vs SEO, argued from TIN's own measurement: how much of a Google ranking carries into AI answers, what still pays, and which GEO tactics measured as nothing."
url: "https://theinternetninja.com/guides/geo-vs-seo/"
updated: "2026-10-16"
---

# GEO vs SEO: what changes, what does not, and what TIN measured

By the TIN research desk. Updated 2026-10-16. A guide in AI search and visibility (https://theinternetninja.com/guides/ai-search-visibility/), The Internet Ninja.

GEO is not SEO with a new name, and it does not replace SEO. Search authority carries over: across 238 AI and automation agencies, AI citations rose with ranking keywords at a Spearman 0.92 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/). A Google ranking is still not an AI citation. In TIN's US measurement, 30.7% of top-10 organic domains were also cited in that search's AI Overview, and 63.0% of the domains the AI Overview cited were not in the top 10 [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). Keep the SEO that builds authority and access. Stop paying for the GEO tactics that measured as nothing.

Both studies are snapshots. The Google top 10 vs AI Overviews study (the top 10 study below) says of its own findings: "These are associations in one snapshot, not causes" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

## GEO, SEO and AEO in one paragraph each

**SEO** is the work of getting a page to rank in a search engine's list of results, so that a searcher clicks through to it. The prize is a position and the visit that follows.

**GEO**, generative engine optimization, "is the practice of making a page more likely to be retrieved and cited by an AI answer engine such as ChatGPT, Perplexity, Gemini or Google’s AI Overviews" source. A 2026 survey of 45 studies describes the target as "a stochastic, partially observable pipeline spanning search activation, crawling and indexing, retrieval, reranking and context allocation, citation, prominence, factual absorption, fidelity, and user behavior" [source](https://arxiv.org/abs/2607.14035). The prize is a citation inside the answer.

**AEO**, answer engine optimization, is the older label for optimising toward a direct answer rather than a list of links: featured snippets first, AI answers now. This guide treats AEO as part of GEO, so "AEO vs SEO" and "GEO vs SEO" get one answer here.

One more meaning to clear up. Some readers who search "what is geo in seo" mean geo-targeting: location pages, local listings, country targeting. That is local SEO, a different subject. Local pages come up once below, under page types.

## How much of SEO carries over: TIN's measurement

TIN read the Google results and the AI Overview for 438 buyer searches in seven markets [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). The US headline rests on 69 searches with an AI Overview, read in one crawl window, 2026-10-01 and 2026-10-02 [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). The match is at domain level, and the surface is Google AI Overviews only: not ChatGPT, not AI Mode. The study publishes its dataset and its amendments on the same page.

The headline counts 152 of 495 top-10 organic domains that were also cited in that search's AI Overview [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). The average hides a wide spread: "the median search shares 29%, the range runs from 0 to 86%, and 12 searches share no domain at all" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

Read the other way, the AI Overview reaches well outside the top 10. Of the domains US AI Overviews cite, 259 of 411 are not among that search's top-10 organic domains; that count "includes 35 opaque Google redirect links, which cannot be matched to any page and so count as outside" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). Matched by exact URL rather than domain, at least 25.7% of US top-10 pages are the page the AI Overview cites (118 of 460), and the study gives that as a lower bound [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

The domains at the top two organic positions are the ones most often cited [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

| US organic position | Share of searches where that domain is also cited | n |
|---|---|---|
| 1 | 44.9% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 31 of 69 |
| 2 | 46.4% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 32 of 69 |
| 3 | 34.8% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 24 of 69 |
| 4 | 26.1% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 18 of 69 |
| 5 | 24.6% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 17 of 69 |
| 6 | 19.4% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 13 of 67 |
| 7 | 21.1% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 12 of 57 |

The overlap barely moves with what the searcher is trying to do. "The US overlap is between 28.8% and 31.5% for each of the four query types" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

| Query type | US overlap | Searches |
|---|---|---|
| Choosing | 31.0% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 33 |
| Comparing | 30.6% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 14 |
| Pricing | 31.5% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 13 |
| Doing | 28.8% [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/) | 9 |

The top 10 study's advice to anyone who wants to be cited in AI Overviews: "Track both, because one is not a stand-in for the other" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).

TIN's agency study ranked 238 AI and automation agencies by their AI citations in Ahrefs' index and compared that with their search metrics [source](https://theinternetninja.com/recon/agency-ai-citations-2026/).

| Search metric | Spearman correlation with AI citations |
|---|---|
| Ranking keywords | 0.92 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/) |
| Organic traffic (Ahrefs estimate) | 0.86 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/) |
| Domain Rating | 0.81 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/) |
| Referring domains | 0.78 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/) |

The sample is 161 cited agencies and 77 uncited [source](https://theinternetninja.com/recon/agency-ai-citations-2026/). Cited agencies have a median Domain Rating of 41; agencies with no citations, 9 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/). The study's own advice: "Treat AI citation as a search-authority problem first" [source](https://theinternetninja.com/recon/agency-ai-citations-2026/).

Two limits travel with those numbers. The correlations run across all AI platforms in Ahrefs' index, not ChatGPT alone, and the population is AI and automation agencies. And they "say which agencies tend to have both, not that one produces the other"; in the study's words, "Nothing here shows that raising Domain Rating, rankings or case-study quality raises AI citations" [source](https://theinternetninja.com/recon/agency-ai-citations-2026/).

TIN's inference, not a measured finding: SEO decides whether a site is in the pool AI engines draw from, and other factors decide which page in that pool gets cited.

### Ahrefs' count by URL

Ahrefs has re-run its own study of the question across 863K keyword SERPs and 4M AI Overview URLs, and found that "37.9% of URLs cited in AI Overviews also appeared within the first 10 blocks" [source](https://ahrefs.com/blog/ai-overview-citations-top-10/). Counting standard blue links only, 37.10% of cited URLs ranked in the top 10, 26.20% ranked 11 to 100, and 36.70% did not rank in the top 100 [source](https://ahrefs.com/blog/ai-overview-citations-top-10/). Ahrefs compares this with "~76% in July 2025", and the same page says its parsing method improved between the two studies, so the change is not a clean before and after [source](https://ahrefs.com/blog/ai-overview-citations-top-10/).

Ahrefs counts URLs across a corpus; the top 10 study counts domains per search, on a different sample and a different date. They are two separate measurements, and they are not to be averaged. Both point the same way: most of what AI Overviews cite is not a top-10 organic result.

## "GEO" is several channels, not one

The phrase "GEO vs SEO" assumes GEO is one thing. TIN's citation sources study asked seven AI engines the same 60 everyday questions in 10 countries, and the engines barely agreed on whom to cite: "1,091 of 1,395 cited sites (78%) appear in only one engine's answers" [source](https://theinternetninja.com/recon/ai-citation-sources-2026/).

| Engines citing the site | Cited sites |
|---|---|
| 1 | 1,091 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |
| 2 | 194 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |
| 3 | 72 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |
| 4 | 28 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |
| 5 | 9 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |
| 6 | 1 [source](https://theinternetninja.com/recon/ai-citation-sources-2026/) |

Even the two Google surfaces differ. "The closest pair, AI Overviews and AI Mode, share 8 of their 25 most-cited sites; ChatGPT shares at most 1 with any other engine" [source](https://theinternetninja.com/recon/ai-citation-sources-2026/). The engines also lean on different kinds of source: "ChatGPT sends 68% of its citations to government and associations; Gemini sends 58% to company sites" [source](https://theinternetninja.com/recon/ai-citation-sources-2026/). And "YouTube is the most-cited site in Google AI Overviews and AI Mode; ChatGPT cites no social or video platform" [source](https://theinternetninja.com/recon/ai-citation-sources-2026/).

A larger audit points the same way. Across 11,500 queries, URL overlap between Google organic results, AI Overviews and Gemini ran at a Jaccard similarity of 0.11 to 0.18 [source](https://arxiv.org/abs/2607.14035). The survey that relays it draws the conclusion that visibility is indexed by engine and surface [source](https://arxiv.org/abs/2607.14035).

The citation sources study covers everyday questions, not business-to-business ones, and its edition 1 is one weekly run, a snapshot, not a trend [source](https://theinternetninja.com/recon/ai-citation-sources-2026/). Within that scope, its advice: "Do not optimise for “AI” as one channel" [source](https://theinternetninja.com/recon/ai-citation-sources-2026/).

So the honest answer to "SEO vs GEO" depends on the engine. The top 10 study measured the overlap with Google rankings for Google AI Overviews only. TIN has not measured how far Google rankings carry into ChatGPT's citations. What the agency study does show is that ChatGPT is the hardest of them: "173 of 238 agencies were never cited by ChatGPT", and its advice is to "Expect ChatGPT to be the last place you show up" [source](https://theinternetninja.com/recon/agency-ai-citations-2026/).

## What does not change

Three parts of SEO practice still hold for AI answers, each with the evidence that supports it.

| Holds for both | Evidence |
|---|---|
| Be fetchable, with facts in the HTML | Most major AI crawlers did not render JavaScript in a December 2024 log analysis [source](https://vercel.com/blog/the-rise-of-the-ai-crawler) |
| Match the topic in the searcher's words | Topic match was a gatekeeper in all six models of a 252,000-trial experiment [source](https://arxiv.org/abs/2605.25517) |
| Build authority off the page | Ranking keywords track AI citations across 238 agencies [source](https://theinternetninja.com/recon/agency-ai-citations-2026/) |

### Be fetchable

A page that cannot be fetched loses in both channels, and the bar is higher for AI. Log analysis by Vercel and MERJ, published 17 December 2024, found that "None of the major AI crawlers currently render JavaScript", naming OpenAI's OAI-SearchBot, ChatGPT-User and GPTBot, Anthropic's ClaudeBot, Meta-ExternalAgent, ByteDance's Bytespider and PerplexityBot [source](https://vercel.com/blog/the-rise-of-the-ai-crawler). Two do render: Google's Gemini through Googlebot, and AppleBot [source](https://vercel.com/blog/the-rise-of-the-ai-crawler). Google's crawler forgives a page that builds its content in the browser. In that analysis, most AI crawlers did not. Every fact you want cited has to be in the server-rendered HTML, and the crawlers have to be allowed in robots.txt.

### Match the topic in the searcher's words

A SIGIR 2026 controlled experiment ran 252,000 paired trials across six language models and 18 content factors [source](https://arxiv.org/abs/2605.25517). Topic match was a gatekeeper, unanimous across all six models, at odds ratios of 221 to >10k [source](https://arxiv.org/abs/2605.25517). Having the query terms on the page was a differentiator, at 5.99 to 40 [source](https://arxiv.org/abs/2605.25517).

That is intent matching, which SEO already teaches. So is the warning against stuffing. In the KDD 2024 paper that named GEO, keyword stuffing brought "little to no improvement"; on position-adjusted word count it scored 17.7 against a 19.3 baseline [source](https://arxiv.org/abs/2311.09735).

### Build authority off the page

Authority is the strongest carry-over TIN has measured, per the agency correlations above. What shifts is the kind of authority. In Ahrefs' study of 75,000 brands, against AI Overview visibility, branded web mentions correlated at a Spearman 0.664 and backlinks at 0.218 [source](https://ahrefs.com/blog/ai-overview-brand-correlation/). Across ChatGPT, AI Mode and AI Overviews, YouTube brand mentions were the strongest signal at approximately 0.737, branded web mentions ran 0.656 to 0.709 by platform, and backlinks 0.191 to 0.244 [source](https://ahrefs.com/blog/ai-brand-visibility-correlations).

These are correlations, not tests. Read with the agency data, they say the discipline is the same and the weight inside it moves: being named across the web counts for more than being linked.

## What changes

| Changes for AI answers | Evidence |
|---|---|
| You win a citation, not a click | Products asked for by name were recognised far more often than they surfaced unprompted [source](https://arxiv.org/abs/2601.00912) |
| A current date counts for more | A recent timestamp against a 2019 one was a gatekeeper in all six models [source](https://arxiv.org/abs/2605.25517) |
| One reading proves little | Identical daily runs disagree on their sources [source](https://arxiv.org/abs/2607.14035) |
| Several sold tactics measured as nothing | Schema, llms.txt, length and formatting (table below) |

### The unit you win is a citation, not a click

A Google rank is public and stable enough to track. An AI citation is a mention inside an answer to one prompt on one engine, and the usual way of checking it misleads. A January 2026 preprint tested 112 startups across 2,240 queries. Asked about a product by name, ChatGPT recognised it 99.4% of the time and Perplexity 94.3%; asked the buyer's question without the name, the products appeared 3.32% and 8.29% of the time [source](https://arxiv.org/abs/2601.00912). It is one author and one preprint. On that sample, asking an engine about a product by name measured recognition, not discovery.

### A current date counts for more

In the same 252,000-trial experiment, a recent timestamp against a 2019 one was a gatekeeper, at 14.4 to >10k [source](https://arxiv.org/abs/2605.25517). A date against no date at all was weak, 1.15 to 13.0, and a missing date and a stale date were near-equivalent at 1.31 to 2.32 [source](https://arxiv.org/abs/2605.25517). The effect was unanimous across models: a page that reads as old loses to one that reads as this year.

### Measurement is noisy

Across four engines and 45 days, identical daily runs agreed on their sources at a Jaccard of approximately 0.34 to 0.42, and the authors propose seven to eight repetitions per prompt as a starting point [source](https://arxiv.org/abs/2607.14035). The survey that relays them warns the number comes from a small set of Swiss queries [source](https://arxiv.org/abs/2607.14035). A rank check is one reading. A citation share needs repeated runs, and every figure has to name its engine, its surface, its date and its prompt.

### Some GEO tactics measured as nothing

| Tactic sold as GEO | What was measured | Source |
|---|---|---|
| Schema markup | 1,885 pages: +2.4% on AI Mode and +2.2% on ChatGPT, both "statistically indistinguishable from zero"; a 4.6% decline on AI Overviews | [source](https://ahrefs.com/blog/schema-ai-citations/) |
| llms.txt | 137,210 domains: 28% publish one, and 97% of those received zero requests in May 2026 | [source](https://ahrefs.com/blog/llmstxt-study/) |
| Longer content | Spearman 0.04 against citation position across 174,048 pages cited in 560,346 AI Overviews; 53.4% of citations went to pages under 1,000 words, 16.0% to pages over 2,000 | [source](https://ahrefs.com/blog/short-vs-long-content-in-ai-overviews/) |
| Organised sections over dense prose | Odds ratio 0.78 to 1.68 across six models: no consistent effect | [source](https://arxiv.org/abs/2605.25517) |

Two limits. The schema pages were already cited before schema was added, so that study does not test an uncited page [source](https://ahrefs.com/blog/schema-ai-citations/). And the llms.txt study measures whether the files get fetched, not whether they lead to citations [source](https://ahrefs.com/blog/llmstxt-study/).

## Can GEO hurt your SEO?

It can. SAGEO Arena (Kim et al., 2026) ran a full search pipeline over 171,003 documents and 2,700 queries, and found that existing GEO approaches "often degrade performance in retrieval and reranking" [source](https://arxiv.org/abs/2602.12187). Optimising only the body of a page for citation "reduces average top-20 presence by approximately 9%, top-10 presence after reranking by 16%, and final citation by 6%", in a 2026 survey's reading of that study [source](https://arxiv.org/abs/2607.14035).

The safer question is which pages to write for which channel. TIN's GEO post gives a ruling per page type, and these are TIN's editorial rulings, not measured findings source:

| Page type | Ruling | Why |
|---|---|---|
| Comparison and alternatives | GEO first | The format answer engines lean on, and the structure serves the organic result too |
| Glossary, FAQ and how-to | GEO first | Passage retrieval fits them and SEO gives up nothing |
| Product and pricing | Split | The facts and the price belong to the machine, the space above the fold to conversion |
| Location and sector | SEO first | The local pack still decides the local page |

The local ruling leans on one study. In Whitespark's 540-query local study, AI Overviews appeared on 68% of local business queries, and in its Houston plumber sample 60% of the citations went to third-party publishers and 40% to individual businesses [source](https://whitespark.ca/blog/case-study-the-prevalence-of-ai-overviews-in-local-search/). The GEO post calls that "one city and one trade", and the only measured split it could find source.

## What this guide cannot tell you

- **Associations, not causes.** Every TIN figure here comes from one snapshot. None of the studies shows that raising rankings, Domain Rating or mentions raises AI citations.
- **Google AI Overviews only, at domain level.** The top 10 study covers one Google surface. It says nothing about ChatGPT, Perplexity or AI Mode, and its exact-URL figure is a lower bound.
- **A sample of buyer searches, in one market that counts.** The top 10 study's intents "are TIN's buyer queries in its four channels. They are not a random sample of all searches" [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). The European markets have 4 to 10 searches each and are not compared [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/). In the US, no search returned more than 8 organic results, so "top 10" means the organic results the page showed [source](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/).
- **Everyday questions for the engine comparison.** The citation sources study does not cover business-to-business questions.
- **One population and one index for authority.** The agency study counts AI and automation agencies, from one index, Ahrefs, on 2026-09-26 [source](https://theinternetninja.com/recon/agency-ai-citations-2026/). Other industries can differ, and TIN has not measured them.
- **Outside studies, read as published.** The controlled experiment, the arena study and the noise figures are other people's work. TIN read them; it did not re-run them.

**Author and method.** Written by TIN. Every figure links to the page it was read from on 2026-10-16. Last updated 2026-10-16.

## Read next

- [Google top 10 against AI Overviews](https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/): the full overlap data by position, query type and market, with the dataset.
- [Which agencies AI engines cite](https://theinternetninja.com/recon/agency-ai-citations-2026/): how AI citations track search authority, and why ChatGPT is the last place most show up.
- [Which websites AI engines cite](https://theinternetninja.com/recon/ai-citation-sources-2026/): the source mix for all seven engines, and how little they share.
- Generative engine optimization: the controlled evidence: the full evidence table behind "what does not change" and "what changes".
- How ChatGPT ranks passages: how passages get fetched, scored and dropped.

**Get cited.** If you want this overlap measured for your own searches, [tell us what you sell](https://theinternetninja.com/about/#work-with-us).

## Sources

1. TIN, "Google top 10 vs AI Overviews" study, edition 2026-10-02. https://theinternetninja.com/recon/google-top10-vs-ai-overviews-2026/
2. TIN, "Agency AI citations" study, edition 2026-09-27, Ahrefs AI index 2026-09-26. https://theinternetninja.com/recon/agency-ai-citations-2026/
3. TIN, "AI citation sources" study, edition 2026-10-02. https://theinternetninja.com/recon/ai-citation-sources-2026/
4. TIN, "Generative engine optimization: what the controlled evidence shows". https://theinternetninja.com/blog/generative-engine-optimization-what-the-controlled-evidence-shows-2026/
5. TIN, "ChatGPT ranks passages, not pages". https://theinternetninja.com/blog/ai-search-optimization-what-the-retrieval-record-shows-2026/
6. Louise Linehan, Ahrefs, AI Overview citations and the top 10, 2026-03-02. https://ahrefs.com/blog/ai-overview-citations-top-10/
7. Olivier Martinez, "Optimizing Visibility in Generative Engines: A Critical Survey of Generative Engine Optimization (2023-2026)", 2026-07-15, relaying Schulte et al. (2026) and Grossman et al. (2026). https://arxiv.org/abs/2607.14035
8. Vishwakarma, Kumar and Jamidar, SIGIR 2026, controlled experiment on citation factors. https://arxiv.org/abs/2605.25517
9. Aggarwal et al., "GEO: Generative Engine Optimization", KDD 2024. https://arxiv.org/abs/2311.09735
10. Vercel with MERJ, "The rise of the AI crawler", 17 December 2024. https://vercel.com/blog/the-rise-of-the-ai-crawler
11. Ahrefs, branded mentions and AI Overview visibility, 75,000 brands. https://ahrefs.com/blog/ai-overview-brand-correlation/
12. Ahrefs, AI brand visibility correlations. https://ahrefs.com/blog/ai-brand-visibility-correlations
13. Ahrefs, short against long content in AI Overviews. https://ahrefs.com/blog/short-vs-long-content-in-ai-overviews/
14. Ahrefs, schema and AI citations. https://ahrefs.com/blog/schema-ai-citations/
15. Ahrefs, llms.txt study. https://ahrefs.com/blog/llmstxt-study/
16. Amit Prakash Sharma, "The Discovery Gap: How Product Hunt Startups Vanish in LLM Organic Discovery Queries", preprint, January 2026. https://arxiv.org/abs/2601.00912
17. Kim, Jeong, Kim, Lee and Lee, "SAGEO Arena: A Realistic Environment for Evaluating Search-Augmented Generative Engine Optimization", 2026. https://arxiv.org/abs/2602.12187
18. Miriam Ellis, "New Research: AI Overviews for Local Business Searches", Whitespark, 12 May 2025. https://whitespark.ca/blog/case-study-the-prevalence-of-ai-overviews-in-local-search/

## Questions

**What is the difference between GEO and SEO?**

SEO works to rank a page in a search engine's list of results so a searcher clicks it. GEO works to get a page retrieved and cited inside an AI engine's answer. Search authority carries over, but a ranking is not a citation: in TIN's US snapshot, most of the domains Google's AI Overviews cited were not in that search's top 10. These are associations, not causes.

**Is SEO being replaced by GEO?**

Not on TIN's data. Across AI and automation agencies, the closest correlate of AI citation TIN measured was search authority, an association rather than a cause. What changes is what you win (a citation, not a click), how you measure it (repeated runs per engine), and which tactics are worth paying for.

**What does GEO mean in SEO?**

Usually generative engine optimization: making a page more likely to be retrieved and cited by an AI answer engine such as ChatGPT, Perplexity, Gemini or Google's AI Overviews. Some searchers mean geo-targeting, which is location SEO for local pages and listings, a different subject.

**Is SEO dead now with AI?**

No. In the agency study, AI citations tracked ranking keywords more closely than anything else TIN measured, and in TIN's Google top 10 vs AI Overviews study the domains at organic positions 1 and 2 were the ones Google's AI Overviews cited most often. Ranking is not enough on its own: most of the domains those AI Overviews cited were outside that search's top 10.

**What is the difference between AEO and SEO?**

AEO, answer engine optimization, is the older label for optimising toward direct answers rather than a list of links. This page treats AEO as part of GEO, so the differences between GEO and SEO here apply to AEO as well.

**Can GEO hurt traditional SEO performance?**

It can. In the SAGEO Arena study, optimising only the body of a page for citation lowered its presence at retrieval, after reranking and in the final citations. In the KDD 2024 paper that named GEO, keyword stuffing brought little to no improvement.

**Does ranking on Google get you into AI answers?**

Partly, on what TIN measured. In its US snapshot of Google AI Overviews, some top-10 organic domains were also cited, but most cited domains were outside the top 10. TIN has not measured how far Google rankings carry into ChatGPT's citations.

## Method and author

Written by the TIN research desk. Every figure carries its source link, and nothing is rounded or invented. How TIN works: https://theinternetninja.com/about/