AI search isn’t a shortcut around SEO.
It’s a reason to do it properly.

What Google actually says

Google published guidance on optimising for its generative AI features, and placed it inside the SEO Fundamentals section of Search Central rather than in a separate AI area. That placement is the argument. Google states directly that from Search’s perspective, optimising for generative AI search is optimising for the search experience — and is therefore still SEO.

It goes further and names things you can ignore: no llms.txt file, no content chunking, no AI-specific rewriting, and no special schema.org markup for AI Overviews or AI Mode. Pages still need to be crawlable, indexed, snippet-eligible and genuinely useful.

There is no hidden schema that makes an AI recommend you.

The one distinction worth internalising

Google draws a line between commodity and non-commodity content. Commodity content is assembled from what is already available everywhere — "7 tips for first-time homebuyers". Non-commodity content comes from direct experience: what you actually did, what it cost, what went wrong.

That distinction matters more now, because a language model can produce the commodity version instantly. If 500 sites publish "10 social media tips for businesses", there is no reason for yours to be the one cited.

What to do instead

1. Answer real questions. Write the pages that answer what customers already ask you — what a rebrand costs, how long a website takes, what a brand guidelines document should contain.
2. Say what you do plainly. "We create impactful digital experiences" tells nobody anything. "We design conversion-focused websites, landing pages and SEO foundations" is a sentence a search system can use.
3. Build topic depth, not single posts. A cluster of connected articles linked to the relevant service page does more than one long piece.
4. Get the boring information right — services, locations served, contact details, expertise, FAQs — and keep it consistent everywhere you appear.
5. Earn mentions rather than manufacturing them. The same spam policies apply to AI features as to search.

Measurement is changing too

Google began rolling out Search Generative AI performance reporting in Search Console in June 2026, to a subset of sites. Traffic from AI features is already counted inside overall Search performance under the Web search type. If you do not have the new report yet, the honest answer is to watch overall Search Console performance, engagement and branded search demand.

The actual strategy

Do not build a website for an AI. Build a website an AI would have a good reason to recommend: useful content, clear information, sound technical SEO, and something original that could only have come from you.