Is GEO a New Name for SEO? What 9 Years in Search Taught Me

Is GEO a new name for SEO? Mostly, yes. After nine years in search, I recognise most of the work: make your content easy to reach, answer real questions better than anyone else, and earn trust beyond your own website. What changed is how those sources are retrieved, combined and shown. A ranking is not a citation, a citation is not a recommendation, and a recommendation is not revenue. The new work starts in those gaps.

I first asked this question in a short section of how I built an SEO and GEO practice from zero. This is the full answer, with the evidence from Google, Bing, OpenAI, Ahrefs and Semrush, and the mistakes I made along the way.

What is the difference between SEO and GEO?

SEO (search engine optimisation) is the practice of making a website rank and earn clicks in search engines such as Google and Bing. GEO (generative engine optimisation) is the practice of getting a brand named, cited or recommended inside AI-generated answers from ChatGPT, Google AI Overviews and AI Mode, Gemini, Perplexity, Copilot and Claude. The foundation is shared. The win, the surface and the measurement are not.

AEO (answer engine optimisation) sits in the middle: winning the direct answer, such as a featured snippet or a voice result. I covered how the three differ for beginners in the SEO vs AEO vs GEO section of my career guide.

Who says GEO is just SEO, and who disagrees?

The strongest "it is just SEO" voice is Google itself. Its guide to optimising for generative AI features defines AEO and GEO, then says optimising for generative AI search "is still SEO". It also says you do not need llms.txt files, special schema or content chopped into small chunks to appear in its AI features.

Ahrefs made the same case early. Ryan Law's April 2025 piece, GEO, LLMO, AEO… It's All Just SEO, called GEO "a byproduct of SEO". On the other side, Michael King of iPullRank told Digiday in March 2026 that GEO has its own technical demands: "Each channel needs its own love." Lily Ray of Amsive, in the same piece, saw a familiar hype cycle: "We've all lived through this a million times."

My read is that both camps are right about different things. Google is describing Google. Its AI features are built on its own ranking systems, so of course the advice is SEO. ChatGPT, Perplexity and Copilot pick their sources differently, and Google's guide does not settle the question for them.

How has the surface of search changed since 2017?

The surface moved from a list you click to an answer you read. When my career started in 2017, Google was ten blue links with featured snippets creeping in. By 2026, Google, Bing and ChatGPT all write the answer first and show sources second.

The milestones that changed my day-to-day work:

1. 2019: Google's BERT update made search better at understanding sentences, not just matching words.
2. 2022 onwards: Google's helpful content and core updates. A site I worked on lost traffic while doing what we thought was "right". That was the first time I felt the ground move.
3. November 2022: ChatGPT launched, and the idea of an engine that answers instead of listing went mainstream.
4. May 2024: Google launched AI Overviews in the US, and expanded them to India and the UK in August 2024. On my accounts, impressions climbed while clicks did not keep pace.
5. 2025: Google launched AI Mode. Around this period, leads started telling us an AI assistant had recommended a brand to them.
6. February 2026: Bing launched its AI Performance report, showing how often Copilot and Bing's AI answers cite your pages.
7. 31 August 2026: Search Console's generative AI report went live worldwide, with impressions but no clicks. I unpacked its limits in my zero click search guide.

None of these happened overnight. I felt the change in steps from 2022 onwards, and each step took some clicks away and added some new way to be seen.

What carried over from SEO into GEO?

Almost everything you learnt about authority carried over, but the currency changed. Backlinks still matter. Brand mentions, reviews, press coverage and community discussion now matter alongside them, because AI systems judge you by what the wider web says about you.

Link building became earning mentions

Early in my career, I did what everyone did: guest posts and link exchanges. They moved rankings for a while, then stopped. The work that lasted was getting a brand talked about by sources people already trust.

Ahrefs studied 75,000 brands and found branded web mentions correlated at 0.664 with AI Overview visibility, against 0.218 for the number of backlinks. That is correlation, not cause. Big brands collect both mentions and AI visibility, so you cannot buy mentions and expect the same result. Google's guide also warns that chasing "inauthentic mentions" is less helpful than it seems.

Reputation management and digital PR became GEO

One of my early SEO problems was a VR company that did not show up on Google for its own name. I fixed it with owned pages and trusted third-party profiles. Years later I realised that was a GEO problem too: if AI cannot find a clear, consistent story about you, it either leaves you out or tells someone else's version.

I have since seen digital PR do something it never used to: a press mention I helped earn for a brand later appeared as a source in AI answers. Wikipedia, credible directories and industry listings work the same way. They are where AI looks to confirm who you are.

Reddit and communities became a source of evidence

Reddit and forum threads now get cited for the topics my clients care about. That does not mean you should post there for links. It means you should know what is being said about your brand there, because AI engines read it.

Schema and E-E-A-T moved in, with less magic than I expected

Structured data helps search engines understand a page and qualify for rich results. E-E-A-T (experience, expertise, authoritativeness and trust) is how Google's quality raters judge content. Neither is a switch that makes an AI cite you. Google says plainly that no special schema is needed for its AI features. I learnt this the hard way, which I come back to below.

How did AI start "seeing" websites?

AI engines answer in two ways: from what the model learnt in training, and from pages it retrieves live when you ask. GEO mostly works on the second. Google's AI Overviews and AI Mode use "query fan-out", running several related searches across subtopics and citing pages from those results, according to Google's AI features documentation.

That is why ranking in the top 10 is no longer enough. Ahrefs found that in March 2026 only about 38% of AI Overview citations also ranked in the top 10 for the query, down from about 76% in July 2025. Ahrefs notes its method changed between the two studies, so treat the drop as a direction, not an exact figure.

ChatGPT works differently again. OpenAI's publisher FAQ says any public site can appear in ChatGPT search if it does not block OAI-SearchBot, and that ChatGPT may also get results from a third-party search provider. Copilot is grounded in Bing. So your Bing index status matters more now than it did when I ignored it.

Retrieval is only the first step. A retrieved page is not always cited, a cited page is not always recommended, and a recommendation does not always lead to a sale.

How did keywords turn into prompts?

Keywords did not disappear. They became the starting point. A prompt like "which accounting firm should a UK startup use" can trigger several searches behind the scenes, and those searches still return ranked pages.

My own process changed in three stages. First, I moved from volume-sorted keyword lists to question and intent mapping, using People Also Ask. Then I added prompt sets: 25 to 50 real buyer questions, run across engines several times, tracking who gets named and who gets cited. Today I use both. Keyword volume still drives the content plan, and prompts check whether that plan shows up in AI answers.

The data for this article shows why you still need keyword research. In Ahrefs, "geo vs seo" gets about 11,000 searches a month worldwide and "seo vs geo" about 8,000. "Is geo the new seo" gets almost none. People search the debate as a comparison, so that is how this page is built.

Which sources shaped AI answers in the last two years?

The sources AI engines trust change fast, and they differ by engine. Semrush tracked over 230,000 prompts for 13 weeks in 2025 and found that Reddit's share of ChatGPT responses fell from about 60% to about 10% between early August and mid September. Wikipedia fell from about 55% to under 20% over the same weeks. Google's AI Mode leaned on LinkedIn, YouTube and Reddit instead.

Two cautions. Semrush sells tools for this problem, and it says its data does not explain why the shift happened. The lesson I take is not "post on Reddit" or "avoid Reddit". It is that no single source is safe to depend on, which is exactly what good link building taught us ten years ago.

Where does SEO win but GEO fail, and the other way round?

This is the part most "SEO vs GEO" articles skip, and it is where I have earned my opinions. The same work can succeed in one place and fail in the other:

• Ranks on Google, missing from AI answers. A page can rank well and still be left out when AI compares options, because nothing else on the web backs it up.
• Mentioned, but someone else is cited. In audits I have seen AI name a brand while linking a third-party page as the source. I wrote about this in what happens when AI mentions you but cites someone else.
• Visible in English, weaker in the local language. On one consumer app audit, the share of AI answers naming the brand fell from 30% in English to 22% in the local language.
• Technically perfect, no change in AI visibility. Clean code and schema did not move citations on their own.
• More impressions, flat enquiries. Across three client accounts in my look at whether SEO is dead, impressions grew on all three while enquiries grew on only two.

What mistakes did I make in nine years?

My biggest mistake was focusing only on the page. I spent years perfecting titles, headings and internal links while ignoring what the rest of the web said about the brands I worked on. GEO punished that habit first. Four more followed from it:

1. Chasing rankings, not revenue. I reported positions and traffic for years before I tied SEO to enquiries. AI answers made that gap obvious, because you can be cited without getting a single visit.
2. Link volume over reputation. Guest posts and swapped links looked good in a report. Mentions in sources people trust did more, and still do.
3. Ignoring Bing. I treated Bing as an afterthought. Then Copilot and Bing's own AI answers started citing pages from the Bing index, and I had to catch up on Bing Webmaster Tools.
4. Treating schema as a magic fix. I expected structured data and technical work to drive AI citations far more than they did. Original information and third-party mentions mattered more.

My earlier mistakes, the ones from starting out, are in how I started in SEO and moved to GEO.

Which AI engine should you optimise for first?

Start with ChatGPT and Google's AI Overviews and AI Mode, because that is where the users are. OpenAI said ChatGPT passed 1 billion weekly active users in July 2026, and Google said in August 2026 that the Gemini app passed 1 billion monthly users. After that, the right engine depends on your region and on how your buyers search.

Region matters more than most guides admit. For an enterprise B2B client, Copilot sits inside Microsoft 365, where the buyers already work. For a consumer brand in India, the local-language gap I mentioned above can matter more than which engine you pick.

What should a sceptical CMO ask about GEO?

The fairest question is: if the work is mostly SEO, why pay separately for GEO? My answer: if a GEO audit gives you the same recommendations as your SEO audit, with a new dashboard and a higher invoice, it has not earned a separate budget. That is why my team had to prove GEO worked before we sold it.

The traffic numbers also call for honesty. Marie Haynes analysed 100 websites and, as MarTech reported in October 2026, ChatGPT sent 0.62% as many sessions as Google organic. Yet 13.8% of ChatGPT visitors converted against 3.02% from Google, on the sites that tracked conversions. AI is small for traffic, strong for intent, and important for how buyers see you before they ever visit.

How do I approach SEO and GEO together? My six-step method

I run SEO and GEO as one programme, in this order:

1. Set the SEO baseline. Crawlability, rankings, demand, organic conversions and technical issues. If this is broken, nothing else matters.
2. Map real questions. Keywords, conversational prompts, the sub-searches behind them and the objections your sales team hears.
3. Audit AI visibility. Mentions, citations, which competitors get recommended, and whether what AI says about you is accurate. Run every prompt several times, because answers vary.
4. Fix owned content and access. Make original, useful information easy to reach. Allow the crawlers you want. This is where my three-pillar GEO framework comes in.
5. Work on earned reputation. Find which independent sources discuss the brand, correct what is wrong, and fill the gaps with real PR, reviews and community presence.
6. Measure outcomes, honestly. Compare search performance, AI visibility, referrals, branded demand and qualified leads, and say plainly what you cannot attribute.

The Princeton GEO study found that adding evidence, such as statistics and quotations, raised visibility in generative answers by up to 40% in its tests. That supports step 4. It is a lab result, not a promise of traffic.

What should you expect next from each engine?

Expect measurement to keep improving and sources to keep moving. In 2026 alone, Bing added citation reporting, Google added an AI report to Search Console, and OpenAI documented how ChatGPT search finds pages. Each made GEO more measurable and more like SEO. What I expect, and this is my opinion rather than anyone's roadmap:

• Google will keep folding AI answers into normal search, so ranking for the sub-questions will matter more than ranking for the headline query.
• ChatGPT will keep changing which sources it favours, as the Reddit and Wikipedia swing showed. Spread your evidence across sources.
• Copilot will keep tying web visibility to work tools, which makes Bing more important for B2B.
• Perplexity will keep rewarding fresh, specific answers.

So, is GEO a new name for SEO?

Mostly, yes, and calling it a whole new discipline undersells nine years of work that still applies. But "it is just SEO" is lazy if it means nothing changed. The surface moved from lists to answers, the unit moved from rankings to citations, and the people who vouch for you now matter as much as the links they give. If you understood SEO as earning trust rather than gaming rankings, you are already most of the way to GEO.

If you are new to all of this, start with how I would start SEO and GEO today. If you want help working out where your brand stands in AI answers, get in touch.