Google’s new generative AI Search guidance (May 2026): what it means for SEO

Okapi & Co • May 2026

Google just published new advice for AI search. Here’s what it really means for SEO.

Search is changing fast, but Google has just made one thing much clearer: optimizing for AI-driven search experiences is not a separate discipline from SEO—it’s SEO evolving.

In May 2026, Google published new resources to help site owners understand how to improve visibility in generative AI features in Search (like AI-generated answers and summaries). If you’ve been seeing “GEO” (Generative Engine Optimization) and “AEO” (Answer Engine Optimization) everywhere, Google’s message is refreshingly practical:

• There’s no shortcut or “hack” to win AI answers.
• The fundamentals still matter.
• The bar for quality, structure, and trust is rising.

Below is our take at Okapi & Co: what Google’s guidance actually means, what we’re telling clients to do next, and how to make sure your brand is the one AI systems cite and recommend.

1) Google’s key point: GEO/AEO is still SEO

Google’s new “AI optimization” guidance explicitly calls out terms like AEO and GEO and reframes them: from Google’s perspective, optimizing for generative AI search experiences is simply optimizing for the modern search experience.

That matters because it cuts through the noise. You don’t need a new playbook full of speculative tactics. You need to execute the fundamentals at a higher standard—because AI answers expose weaknesses faster:

• Thin, repetitive content becomes invisible.
• Vague claims don’t get trusted.
• Poor structure means your best insights can’t be extracted into answers.

Okapi & Co view: the winners won’t be the brands producing more content—they’ll be the brands producing more usable evidence.

2) AI content isn’t “bad”—scaled unhelpful content is

One of the most important parts of Google’s updated documentation is its clarity on generative AI content:

Using AI can be perfectly fine for research, drafting, and structuring. The risk is publishing lots of pages “without adding value,” which can cross into spam policy territory (scaled content abuse).

This is where many teams get it wrong. They hear “Google is okay with AI” and interpret that as “we can mass-produce content again.”

You can’t. Not sustainably.

Okapi & Co rule of thumb: if a page could be published by any competitor with the same prompt and the same sources, it’s unlikely to be the page AI systems choose.

3) What actually increases AI visibility: extractability + trust

Generative AI features don’t just “rank” pages. They assemble answers. That changes what you need your pages to do.

Make your pages extractable

AI systems favour content that’s easy to lift into a reliable answer:

• Clear headings that mirror real sub-questions
• Direct answers early on (not buried 700 words down)
• Definitions, constraints, and “when it depends” nuance
• Structured lists, steps, tables (where appropriate)

Make your pages trustworthy

AI answers tend to reward what can be grounded:

• First-hand experience (screenshots, walkthroughs, real examples)
• Specific data (even small datasets are powerful)
• Transparent authorship and credentials
• Updated content that reflects current reality

This is the new baseline. Not “SEO copywriting.” Not “content volume.” Useful pages that can be confidently reused.

4) The KPI shift: from rankings to recognition

Google hasn’t killed rankings—but AI features are changing what “visibility” means.

For many queries, the first impression of a brand now happens in an AI-generated summary, not a list of blue links. That means your measurement has to widen:

• Are you cited in AI answers?
• Are you recommended, compared, or merely mentioned?
• Do your pages appear as sources for the key questions in your category?
• Are you building recognition before the click ever happens?

Okapi & Co recommendation: track a prompt set per client (category + “best X for Y” + comparisons + “how to” questions) and monitor whether your brand appears, and in what role (citation vs recommendation).

5) What we’re advising clients to do this month

If you’re trying to turn Google’s new guidance into action, here are the moves we’re implementing with clients right now.

A) Build your “entity spine”

AI systems need clarity on who you are:

• Strong About page (who you help, your POV, your credentials)
• Clear author pages with experience and real work examples
• Consistent contact details (especially for local/regional businesses)
• Organisational trust signals (policies, case studies, proof)

B) Upgrade your money pages for AI extraction

For your most commercially important pages:

• Add a short “direct answer” section near the top
• Add FAQ-style subheadings that match real search questions
• Add examples and evidence blocks (screenshots, mini case studies)

C) Stop publishing “interchangeable” content

Replace “10 posts a month” with “2 posts that are hard to copy.”

If your content doesn’t contain proprietary experience, it’s increasingly at risk in a world where AI can summarise everyone’s generic advice.

Final thought: the SEO advantage is becoming editorial discipline

Google’s new advice isn’t a “GEO trick list.” It’s a signal: the future belongs to sites that publish content worth reusing.

That’s what we do at Okapi & Co: we help brands earn visibility not only in rankings, but in the answers that shape customer decisions before a click ever happens.

If you want help translating this into a practical roadmap—technical, editorial, and measurement—get in touch.

Call to action: Want an AI search readiness audit? We’ll review how your brand shows up in AI answers and give you a prioritized plan to improve citations and recommendations.