SEO isn't dead: how to get found by Google and ChatGPT

As the only marketer in a B2B company, I have one pair of hands, ten channels and an algorithm that changes its mind every month. I can't afford to chase every hype.

That helps with filtering. Every year someone on LinkedIn declares SEO dead. Usually it's the same person who previously buried the newsletter, email and presumably the fax.

SEO is alive and well. It has just gained two little brothers, AEO and GEO, with abbreviations that sound like a Swedish furniture range. What are you supposed to do with them as a business owner?

The short answer: SEO makes sure you get found in Google, AEO that your text becomes the answer, and GEO that AI tools like ChatGPT mention you. All three reward the same thing: a text that answers directly, is technically sound, tells the same story everywhere and backs up every claim.

One question in three outfits

SEO, search engine optimisation, is about how you get found in Google.

AEO, answer engine optimisation, is about how you become the answer: in the featured snippet, in the AI summary at the top of the search results, in the voice of a smart speaker.

GEO, generative engine optimisation, is about whether ChatGPT, Perplexity, Gemini and Claude mention you when someone asks them about your field.

It sounds like three disciplines. In practice it's one question in three outfits: does a machine understand who you are, what you know and why it should believe you?

How is AI changing the way we search?

AI is shifting search from a list of links to one ready-made answer. It used to be that someone typed in two keywords and clicked on the third link.

Now that same person types a whole question into a chat window, gets four neat paragraphs back and may not click on anything at all. The battle for the click has partly become a battle for the mention. You want to be the source that answer relies on.

That doesn't mean keywords disappear. What does disappear is the text written for an algorithm, with the same keyword crammed in seventeen times. Google already hated that. Language models mostly find it boring.

How do you write a text that AI models quote?

The answer first. A language model and a reader in a hurry want exactly the same thing: the core in the first two sentences. Anyone who warms up for three paragraphs before getting to the point gets skipped by both.

Good markup. I'm fanatical about it. Structured data is the package leaflet of your website. You tell a search engine in black and white: this is a person, this is her profession, this is a frequently asked question with this answer. A visitor never sees it, a machine reads it first.

The same story everywhere. Language models build a picture of you from everything they come across: your site, your LinkedIn, your business profile, mentions on other sites. If you're called a copywriter in one place, a content strategist in another and a creative all-rounder in a third, a model doesn't know what to make of you. You can read more about that in Why ChatGPT doesn't know your business.

Only what you can back up. A model would rather quote a sentence with a number, a name and a date than a sentence full of adjectives. The same goes for people, by the way.

How to pour that approach into questions and answers, you can read in Are FAQs still worth it?

Can you use AI without your text sounding like AI?

Yes, as long as AI does the groundwork and you do the thinking. I work with AI every day and, to be honest, I'm a little bit in love with it.

I use it to map search intent, to test the structure of a page, to come up with forty headlines of which thirty-nine are no good, to write FAQ schema and to read figures from ad campaigns. What I don't let it do is think for me.

You can spot a text that AI wrote entirely by itself from a hundred metres away. It's neat, it's fluent and it says nothing. It contains the word seamless.

Ironically, language models themselves aren't keen on such texts as a source either: what everyone has already written, they don't need to come to you for.

The gain lies in the combination. AI makes the research faster and the manual work lighter. The human supplies what a model doesn't have: experience, an opinion, an example from last week and a joke that goes just over the edge.

Why is GEO good news for small marketing teams?

Because AI models don't choose the biggest source, but the clearest one. Big brands have the budget to appear everywhere at once. Small teams don't, but they do have something else: specific knowledge.

A small organisation that explains exactly what it does and for whom, and delivers that in a technically tidy way, can simply stand next to a multinational in an AI answer.

That's why I believe in this approach, especially with my own ten channels. It's not a trick that expires next month. It's good writing, good structuring and being honest about what you know.

What you don't need to do

You don't need to buy secret AI tricks. Google itself says that no special optimisation or special markup is needed to appear in Google's AI features. The ordinary SEO basics still apply.

Nobody can promise you that ChatGPT will mention you. You can measure whether it happens, though, and you can read how in Does ChatGPT mention your business?

All grown up

So SEO isn't dead. It has grown up and nowadays it listens to robots too.

Frequently asked questions

Has AI killed SEO?

No. AI tools base their answers largely on web pages, and Google says the ordinary SEO basics also apply to its AI features. SEO has gained two little brothers, though: AEO and GEO.

What is the difference between SEO, AEO and GEO?

SEO is about being found in Google, AEO about becoming the answer in featured snippets and AI summaries, and GEO about being mentioned by AI tools such as ChatGPT, Gemini and Perplexity.

How do you write a text that AI quotes?

Answer directly in the first two sentences, use structured data, tell the same story about who you are everywhere and back up every claim with facts, names or figures.