GEO is becoming citation-first. Here's what that means for your business.

Search is changing, and it's not about following a new trend. It's about how people actually get answers now.

Google AI Overviews have become a default assumption in SEO planning. That shift is real. And it's forcing everyone in our industry to rethink what success actually looks like.

We've spent the last month studying what practitioners are seeing, what the data shows, and what matters for your business right now. This post summarises our thinking and the actions we're taking with clients.

The shift from rankings to citations

For years, SEO has been about rankings. Your keyword ranks position 3, you get X% of clicks, revenue follows.

That model is breaking down.

When an AI Overview appears for a query, it does cite sources. But those citations are selective. The AI picks which sources are worth mentioning, and which ones stay silent. A page can rank in the top 10 and still never be cited.

This is what the industry calls citation-first SEO or GEO (Generative Engine Optimisation). It's not hype. It's what's happening in practice.

Search Engine Land recently refreshed their explainer on how AI Overviews work, signalling that this is now a standard planning assumption for SEO programmes.

Ranking is necessary. But ranking alone is not enough anymore.

Why citations matter more than rankings

Citations do two things rankings do not:

They drive traffic differently. An AI Overview citation is a direct reference, often with your brand name, your specific claim, or a fact attributed to you. The person reading the overview has already found an answer. Some will click the citation link. Others will not. But of those who do click, the intent is usually clear (they want to verify, learn more, or engage with you directly).

They signal authority to the AI system. When an AI repeatedly cites you for certain topics, it's learning that you are a reliable source on those topics. Over time, this affects whether you get cited again, and in what contexts.

What needs to change

This is not about tricks or optimisation tactics. It's about being easier to cite and harder to ignore.

There are three practical areas:

1. Your content needs to be extractable

If an AI system cannot reliably read your page, it cannot cite it.

This sounds obvious, but many sites fail this test:

• Heavy JavaScript that delays content rendering
• Complex layouts that bury key information
• Missing or unclear structure (no headings, no FAQ sections, no definitions)
• Images with no alt text or captions
• Thin or repetitive content that adds no new value

When we audit a site for AI visibility, we render the page as Google and AI agents would see it, then check if key claims, facts, and definitions are immediately visible and quotable. Many sites fail. Most can be fixed.

If you're worried about your site's structure, our technical SEO guidance covers how to optimise pages for both people and machines.

2. Your entity (your business) needs to be clear

AI systems think in entities. An entity is a named thing with consistent attributes. Your business, your team, your services, your location.

If your entity signals are weak or scattered, the AI cannot reliably associate your content with your business. That means even good content may not be cited under your name.

Practical steps:

• Create a clear About / Team / Credentials page that explains who you are, what you do, and where you operate
• Use structured data (Schema.org Organization and Person tags) to tell Google what your entity is
• Keep your Name, Address, Phone consistent across your site, Google Business Profile, and third-party listings
• Link your team members (authors) to the organisation clearly, with their credentials where relevant

3. Your content needs to be quotable

AI systems tend to cite specific, well-formed statements.

This means:

• Clear definitions at the start of sections
• Numbered steps or processes
• Data or evidence you have collected yourself
• Before and after examples
• Small datasets or findings from your own work
• FAQ sections that directly answer common questions

Rambling paragraphs are harder to cite. A paragraph that says "We did X, here's what we found" is much easier.

This is not about writing for AI. It's about writing clearly.

What to do this week

Don't wait for perfect. Start small and build from there.

Week 1 actions

1. Pick 20 to 30 priority queries (the topics you most want to be visible for). For each one, search the query in Google and note whether an AI Overview appears. If it does, note which sources are cited.
2. Create or strengthen your entity page. You need a clear About section that states what you do, who you are, key credentials, and where you operate. Keep it under 300 words. Make it factual. Add structured data (Organisation schema at minimum).
3. Audit one high-traffic page. Render it as Google would. Check if key claims are visible and clearly stated. Rewrite any rambling sections into definitions or numbered steps.
4. Add a FAQ section to one key page. Answer the 5 to 10 most common questions about that topic. Short answers, clear questions. FAQ schema is optional, but helps.

Month 1 actions

1. Build a citation tracker. A simple spreadsheet where you record, weekly, which queries show AI Overviews and whether you are cited. Do this for your 20 to 30 priority queries. Track it for four weeks to see patterns.
2. Refresh your 5 most important pages. For each page, add a clear definition, a process section (with numbered steps), and a FAQ. Keep changes factual and useful. Don't add text just to be longer.
3. Add author information to key posts. Show who wrote it, their credentials, and when it was last updated. Use Person + Author schema if you can.
4. Create one data or research post. Test something small in your industry (customer feedback, case study metrics, process observations). Document the findings. Share the specific numbers. AI systems prefer citing content that includes original research or specific claims backed by evidence.

What we do not know (and won't pretend we do)

There is a lot of uncertainty in this space right now.

We don't know the full weighting of different signals in Google's AI citation decisions. We can see what practitioners report seeing. We can test and measure our own results. But the exact algorithm is not public, and it will change as Google learns.

We don't know how long the current citation landscape will stay stable. AI search is very new. Google is learning what works. Your citation performance now may look different in six months.

Some practitioners have reported that the March 2026 core update increased dominance of UGC and big brands in AI citations. We don't know if smaller, independent sites can compete in all niches. Some queries appear to be "closed loops" where the AI mostly cites the same large sites repeatedly. We are testing strategies to break through those loops (original data, unique insights, community distribution), but results are mixed.

We don't know the full impact on long-term traffic. AI Overviews are not yet showing on all queries, not in all regions, and not for all users. Click-through rates from AI citations vary wildly by topic and context.

Concerns about accuracy in AI-generated answers have prompted debate across the industry. Even "~90% accurate" responses still imply huge absolute error volumes at Google scale.

What we do know is this: ignoring AI visibility is no longer an option. Whether you get citations or not matters more now than it did a year ago. And the sites that are investing in clarity, extractability, and authority are seeing better results than those that are not.

A different way to think about this

GEO is not SEO 2.0. It's not a replacement. It's an addition.

You still need to rank in traditional search results. Keywords still matter. Backlinks still matter. Page speed still matters.

But you now also need to think about whether you will be cited when your topic appears in an AI answer. And that requires different thinking.

It requires asking: Can this content be easily read and quoted by a machine? Is my business identity clear and consistent? Am I saying things that are specific enough and evidence-based enough to be worth citing?

These are good questions for any content. They make your site clearer, more useful, and more trustworthy. Even if AI search disappeared tomorrow, you would benefit from acting on them.

That's why we are focused on this. It's not about optimising for a feature. It's about building a clearer, more evidence-based online presence that works for people and machines alike.

Next steps

We are publishing three more guides this month on:

1. How to audit your content for AI extractability (a practical walkthrough)
2. Entity clarity and schema (what to set up, why, and how)
3. How to build a citation tracker and measure progress

Get in touch if you want to discuss how this applies to your specific business. We can audit your priority queries and content in about two hours, and give you a clear action plan.

In the meantime, start with the Week 1 actions above. They are low-cost, high-value, and they matter regardless of what happens next in AI search.