Gareth Puckey Say hello

SEO, GEO and AEO

SEO (search engine optimisation) is the work of helping search engines understand your pages so the right people find them. GEO (generative engine optimisation) and AEO (answer engine optimisation) are newer names for getting your content used and cited inside AI-written answers, such as Google’s AI Overviews and AI Mode or ChatGPT search. On SEO vs GEO, Google’s own position is that optimising for its AI features is “optimizing for the search experience, and thus still SEO” (Google Search Central), so the practical way to treat GEO and AEO is as an extension of SEO with new ways to measure it.

By Gareth Puckey Hands-on in SEO and GEO Updated 19 sources 4 guides

What it is

SEO: search engine optimisation

Google describes SEO as helping search engines understand your content, and helping people find your site and decide whether to visit it (Google SEO Starter Guide). The groundwork is familiar: pages that can be crawled and indexed, internal links that help them get found, important content in text form, and a good experience once people arrive (Google Search Central). Success has usually been judged by where a page ranks and how many clicks it earns.

GEO: generative engine optimisation

The term comes from a research paper, “GEO: Generative Engine Optimization”, first posted in November 2023 by researchers at IIT Delhi and Princeton and accepted to the KDD 2024 conference (arXiv). It defined generative engines as search tools that use generative models to gather sources and summarise them into an answer, and GEO as a framework to help content creators become more visible in those answers.

The authors built a benchmark of 10,000 queries and tested nine content changes on two generative engines: one they built for the research and Perplexity, a public engine (paper, full text). Citing sources, adding quotations and adding statistics did best, with relative gains of 30 to 40% on their main visibility measure. Keyword stuffing, a long-standing SEO habit, gave little or no improvement.

Read those results with care. Visibility was scored by how much of an answer drew on a source and how prominently it appeared, so the paper says nothing about traffic or sales. The authors also say their methods may need to adapt as engines change, and that they did not test the effect on search rankings (paper, full text).

AEO: answer engine optimisation

Google’s guide groups AEO with GEO as labels for work aimed at visibility in AI search experiences (Google Search Central). Google doesn’t draw a line between the two. The way I’d separate them (my view, not an industry standard): AEO is about being the direct answer to a specific question, and GEO is about being one of the sources an AI system reads, summarises and cites in a longer answer. Both depend on the same pages being found and trusted first.

How SEO, GEO and AEO overlap and differ

They overlap heavily. Google says its generative AI features are rooted in its core ranking and quality systems. They use retrieval-augmented generation (also called grounding) to pull relevant pages from the Search index, and query fan-out, where the model runs several related searches at once to gather more sources (Google Search Central). To appear as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to show in Search with a snippet (Google Search Central). If the SEO groundwork fails, Google’s AI features have nothing of yours to work with.

The differences are real, though:

  • What the searcher sees. Classic results are a ranked list of links. AI features give a written answer with links to supporting pages, and Google says fan-out lets them show a wider, more varied set of links than a classic search (Google Search Central).
  • What counts as success. In SEO it’s ranking and clicks. In GEO and AEO it’s being cited or mentioned in the answer, which may or may not lead to a click.
  • Which systems you deal with. Google’s AI features draw on Google’s own index. ChatGPT search relies on OpenAI’s crawler, OAI-SearchBot, and sites that block it are not shown in ChatGPT search answers (OpenAI).

SEO vs GEO vs AEO side by side

SEOGEOAEO
Stands forSearch engine optimisationGenerative engine optimisationAnswer engine optimisation
The aimHelp search engines understand your pages so the right people find themBe one of the sources an AI system reads, summarises and citesBe the direct answer to a specific question
Where it showsA ranked list of linksLonger AI-written answers, such as AI Overviews, AI Mode and ChatGPT searchShort, direct answers inside the same AI features
Success looks likeRankings and clicksBeing cited or mentioned in the answerBeing the answer given
How to measure itThe Search Console Performance reportThe Search Console generative AI report and Bing’s AI Performance reportThe same reports as GEO
Google’s viewThe foundation for all threeStill SEOStill SEO
The split between GEO and AEO is my working view, not an industry standard. Google treats both as SEO.

Why businesses need it

AI answers now sit in front of UK searchers

Ofcom’s Online Nation 2025 report found about 30% of searches show an AI overview, and 53% of UK adults say they often see these summaries, mostly because their usual search service now includes them by default (Ofcom, 10 December 2025). Google is used by 82% of adults, with 3 billion UK searches a month, while ChatGPT had 1.8 billion UK visits in the first eight months of 2025, up from 368 million in the same period of 2024. Ofcom’s summary page doesn’t explain how each figure was collected, so treat them as a guide to direction.

Google’s own figures show the scale. It says AI Overviews has more than 2.5 billion monthly active users and AI Mode passed 1 billion within a year (Sundar Pichai, Google, 19 May 2026). AI Mode started rolling out in the UK on 28 July 2025, and Google said early users were asking questions two or three times the length of traditional searches (Google, 28 July 2025).

Fewer clicks on many informational searches

Independent studies point the same way, each with limits worth knowing:

  • Pew Research Center tracked the browsing of 900 US adults in March 2025. They clicked a traditional result on 8% of visits to pages with an AI summary, against 15% on pages without one, and clicked a link inside the summary on 1% of visits (Pew Research Center). It covers US users only and one month.
  • Ahrefs studied 300,000 keywords using aggregated Search Console data and found an AI Overview correlated with a 58% lower average click-through rate for the top-ranking page in December 2025 (Ahrefs, 4 February 2026). It used desktop data only, and the 58% compares actual results with a forecast built from similar keywords without AI Overviews.
  • Seer Interactive tracked 53 brands from January 2025 to February 2026. On informational queries, brands cited in an AI Overview earned about 120% more organic clicks per impression than brands that weren’t cited, but still about 38% fewer than when no AI Overview appeared (Seer Interactive, April 2026). The authors say they cannot claim causation, as higher-authority brands are more likely to be cited, and they report the decline reversing in early 2026.

Google disputes the bleaker readings. In August 2025 its head of Search said total organic clicks from Google to websites were relatively stable year on year, that click quality had risen, and that some third-party reports used flawed methods (Google, 6 August 2025). The post doesn’t publish the figures behind those statements.

What this means for performance

The commercial question is whether search still brings in enquiries and revenue at a cost that works. Fewer clicks on early research questions may not matter if the visits that remain convert well. Google says clicks from results pages with AI Overviews are higher quality, meaning people spend longer on the site (Google Search Central). That’s a claim you can only test against your own revenue data.

My work covers SEO and GEO as well as Google Ads, Meta ads and GA4, with CRM data connected so revenue can be traced back to marketing. I apply the same discipline to search: judge organic and AI-referred visits by the enquiries and revenue they lead to. Seer’s data also has a paid search angle: it found paid ads had a higher click-through rate on informational queries where the brand was cited in the AI Overview (15.74% against 11.19%), though again without proof of cause (Seer Interactive).

Where it’s heading (2026 onwards)

  1. Google has published its own GEO and AEO guidance

    On 15 May 2026 Google published a guide to optimising for its generative AI features, including a section on common “AEO/GEO” misconceptions (Google Search Central blog). The guide, last updated on 10 July 2026, says that for Google Search you can ignore llms.txt and other AI-specific files, breaking content into small chunks, rewriting content just for AI, chasing inauthentic mentions and over-focusing on structured data (Google Search Central). It also says a site must be included in the Search generative AI features setting in Search Console to be eligible for those features.

  2. Google is adding agents to Search

    At Google I/O on 19 May 2026, Google announced information agents in Search, which people set up to watch for information in the background and help them act on it. They were due to roll out over the summer, starting with Google AI Pro and Ultra subscribers (Sundar Pichai, Google). Google’s guide now includes early advice on AI agents that carry out tasks such as comparing product specifications, including browser agents that read a site through screenshots, its page structure and the accessibility tree. It also mentions emerging standards such as the Universal Commerce Protocol (Google Search Central).

  3. AI visibility can now be measured, partly

    On 3 June 2026 Google launched generative AI performance reports in Search Console, covering AI Overviews, AI Mode and generative features in Discover, and by 31 August 2026 they had rolled out to all sites worldwide (Google Search Central blog). The launch post lists impressions, pages, countries, devices and dates; it doesn’t list clicks. AI feature traffic also stays in the main Performance report under the “Web” search type (Google Search Central).

    Microsoft launched an AI Performance report in Bing Webmaster Tools as a public preview on 10 February 2026. It shows how often your pages are cited in Copilot, Bing’s AI summaries and some partner services, which pages were cited and the grounding queries behind them (Bing Webmaster Blog).

  4. More links inside AI answers

    On 15 October 2026 Google said it was adding more inline links to AI Mode and AI Overviews, showing them higher in responses more often, and testing a section at the top of some responses that highlights official websites (Google, 15 October 2026).

My view is that citations will sit alongside rankings as a standard search metric, and that the sites that keep earning them will publish things an AI system can’t assemble from existing pages, such as first-hand data and clear, evidenced positions. The measure that matters to a business won’t change: whether search, in whatever form, brings in revenue at a cost that works.

Common mistakes

Treating GEO as a replacement for SEO

Google’s AI features retrieve pages from the same index, so a page that isn’t indexed or eligible for a snippet can’t be used (Google Search Central). Crawl and index problems come before any AI-specific work.

Paying for tactics Google says it ignores

Google says llms.txt files neither help nor harm visibility in Google Search, because Google Search doesn’t use them, and that there’s no need to break content into chunks or rewrite it just for AI (Google Search Central). Other services may read llms.txt, so it’s your call, but don’t buy it as a Google tactic.

Publishing a page for every possible question

Creating separate pages mainly to match fan-out queries breaches Google’s scaled content abuse policy (Google Search Central), and so can generating many AI pages that add no value. Google also asks you to fact-check AI output before publishing, including titles, meta descriptions, structured data and alt text (Google Search Central). If AI drafts your content, keep a named person responsible for what goes live, as covered in using AI in a marketing workflow without losing control.

Blocking the wrong crawler

Google-Extended controls whether your content is used to train Gemini models and to ground answers in Gemini apps, and it doesn’t affect inclusion in Google Search (Google crawler documentation). To limit what appears in AI Overviews and AI Mode, Google points to nosnippet, max-snippet and noindex (Google Search Central). For ChatGPT, OAI-SearchBot (search answers) and GPTBot (model training) are separate settings (OpenAI).

Trusting tools that claim inside knowledge

Google says third-party tools don’t have access to its internal ranking data, and recommends checking AEO and GEO advice against its official guidance (Google Search Central). Lab studies such as the original GEO paper are useful pointers, but they don’t show what any live engine does today.

How to get started

  1. Check you’re eligible. Verify your site in Search Console, confirm your key pages are indexed and not blocked by robots.txt or your hosting, and check that the Search generative AI features setting includes your site (Google Search Central). Add the site to Bing Webmaster Tools too, so you can see Copilot citations (Bing Webmaster Blog).
  2. Decide on AI crawlers on purpose. Write down which you allow: Googlebot for Search and its AI features (Google Search Central), Google-Extended for Gemini training and grounding outside Search (Google crawler documentation), and OAI-SearchBot for ChatGPT search or GPTBot for OpenAI model training (OpenAI). Then make robots.txt match that decision.
  3. Take a baseline before you change anything. Record impressions from the Search Console generative AI report, citations from Bing’s AI Performance report, and organic clicks, enquiries and revenue by landing page.
  4. Improve the pages that make money first. Start with the pages that lead to enquiries. Add what a summary of other sites can’t provide, such as your own data, real examples and a clear point of view, and cite your sources. Google’s guide favours this kind of non-commodity content, and the GEO paper found cited sources and statistics helped in its tests (Google Search Central, arXiv).
  5. Keep business details current. Google’s guide says AI responses can include product listings and local business details, and that Merchant Center and Google Business Profile can help keep yours visible (Google Search Central).
  6. Review every quarter. Compare results with your baseline, check Google’s guide for changes, and record anything you can’t attribute as a judgement call rather than a result.

Questions people ask

Is GEO replacing SEO?

No. Google says optimising for its AI features is still SEO, and AI Overviews and AI Mode pull pages from the same Search index. A page that isn’t indexed, or can’t show with a snippet, can’t be used in those answers (Google Search Central).

What is the difference between GEO and AEO?

Google doesn’t separate them: its guide treats both as labels for visibility in AI search (Google Search Central). My working split is that AEO is being the direct answer to one question, and GEO is being one of the sources an AI system cites in a longer answer.

Do I need an llms.txt file?

Not for Google. Google says llms.txt files neither help nor harm visibility in Google Search, because Search doesn’t use them (Google Search Central). Other services may read one, so it’s a choice rather than a requirement.

How do I know if AI answers cite my site?

Use two free reports. The generative AI performance report in Search Console shows impressions in AI Overviews, AI Mode and Discover (Google Search Central blog). Bing Webmaster Tools’ AI Performance report shows Copilot citations and the grounding queries behind them (Bing Webmaster Blog).

Should I block AI crawlers?

Decide one crawler at a time. Blocking Google-Extended affects Gemini training and grounding in Gemini apps, not Google Search (Google crawler documentation). Sites that block OAI-SearchBot are not shown in ChatGPT search answers, while GPTBot only covers model training (OpenAI).

Does AI search still send traffic?

Less on many informational searches. Pew found people clicked a traditional result on 8% of visits to pages with an AI summary, against 15% without one (Pew Research Center). Google says total organic clicks are relatively stable. Judge it on your own enquiries and revenue, not clicks alone.

Guides on SEO, GEO and AEO

Sources

19 sources, all opened and checked
  1. Optimizing your website for generative AI features on Google Search, Google Search Central
  2. A new resource for optimizing for generative AI in Google Search, Google Search Central blog
  3. AI features and your website, Google Search Central
  4. Search Engine Optimization (SEO) Starter Guide, Google Search Central
  5. Using generative AI content, Google Search Central
  6. Google Search’s guidance on using third-party SEO tools, services, and advice, Google Search Central
  7. Google’s common crawlers (Google-Extended), Google for Developers
  8. Introducing Search Generative AI performance reports in Search Console, Google Search Central blog
  9. I/O 2026: Welcome to the agentic Gemini era, Google
  10. Google Search: Introducing AI Mode in the UK, Google
  11. More ways to connect with the web from Search, Google
  12. AI in Search is driving more queries and higher quality clicks, Google
  13. GEO: Generative Engine Optimization (abstract) and full text, Aggarwal et al., arXiv (KDD 2024)
  14. Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center
  15. AI Overviews reduce clicks: 2026 update, Ahrefs
  16. AIO impact on Google CTR: 2026 update, Seer Interactive
  17. From apps to AI search: how the UK goes online in 2025, Ofcom
  18. Introducing AI Performance in Bing Webmaster Tools public preview, Microsoft Bing Webmaster Blog
  19. Overview of OpenAI crawlers, OpenAI