Skip to main content
SEO InsightsAI Search

AI Overviews Own the Research, Not the Sale: The Query Divide Reshaping SEO

Summarize with ChatGPT
JK
John Kyprianou
July 01, 2026
6 min read
A search results page splitting into two paths, one labelled research absorbed by an AI answer and one labelled buy still sending a clean click

Most of the panic about AI Overviews treats search as one undifferentiated thing that is quietly being eaten. It isn't. The interesting story in the data right now is not "AI Overviews are taking your clicks." It's which clicks they take, and which ones they leave completely alone.

Here is the divide, and it should change how you spend your effort. AI Overviews have made themselves at home on top of research queries. They have almost entirely stayed out of the way of buying queries. If you treat those two as the same problem, you will pour work into the wrong pages.

A search results page splitting into two paths, one labelled research absorbed by an AI answer and one labelled buy still sending a clean click

What the intent split actually looks like

Seer Interactive tracked 5.47 million queries and 2.43 billion organic impressions from January 2025 through February 2026. Their AI Overview CTR analysis found AI Overviews appearing on about 36% of informational queries, roughly 8% of commercial queries, and just under 5% of transactional ones. That is a 7x gap between the top and bottom of the funnel.

Shopping tells the same story from a different angle. Analyses of millions of retail keywords put AI Overview presence on "best [product]" style research queries in the 80s as a percentage, while pure "buy [product]" and "[product] price" queries sit in the low teens. The AI answer wants to help you decide. It rarely wants to stand between you and the checkout.

So the honest summary is this: AI Overviews own the top of the funnel and leave the bottom mostly intact. The research phase, where someone is comparing, learning, and shortlisting, is where the answer box now lives. The transactional query, where someone already knows what they want, still sends a clean click to a real page.

Why this happens, and why it will hold

This is not an accident of rollout timing. It is what the model is good at.

Research queries are exactly the kind of thing a generative summary handles well. "What is the best CRM for a small law firm" has a defensible synthesized answer. Pulling three options together and explaining the tradeoffs is genuinely useful, and Google can do it without much risk of being embarrassingly wrong.

A transactional query is different. When someone searches to buy a specific product or book a specific service, they do not want a summary. They want the page. Inserting an AI answer there adds friction and, for commercial queries, real liability if it gets a price or a spec wrong. Google has a strong commercial reason to keep those SERPs clean too, because that is where the ads and the Shopping units earn their keep.

Our take: this divide is structural, not temporary. Expect the research layer to keep getting more AI-heavy and the transactional layer to stay comparatively click-rich. Plan around that shape rather than around a single headline CTR number.

What this means for your pages

Stop trying to force your product and service pages into AI Overviews. For most transactional queries there is no Overview to be cited in, and chasing one is effort spent on a surface that mostly does not exist for those searches.

Instead, split your thinking by intent, because Google already has.

On research queries, the game is citation, not the click. When an Overview answers "how do I choose an SEO agency in Cyprus," the win is being one of the named, linked sources inside that answer, and being the brand a reader remembers when they later search for you directly. This is where genuine expertise, comparison content, and clear structure pay off. It is the layer we walk clients through in our AI search optimization work, and the mechanics are the same ones covered in our guide on showing up in AI Overviews.

On transactional queries, the game is still the click, so protect it. Your money pages need to load fast, answer the buying question directly, carry clean product or service markup, and remove friction. The traffic that reaches them is more intact than the headlines suggest, which makes conversion the thing worth obsessing over there, not AI visibility.

Build the bridge between the two. The research query is increasingly answered without a visit. That makes brand memory the real prize. If you are the source cited during the research phase, you are the name typed into the search box during the buying phase. That handoff is the whole ballgame now, and it is closely related to the pattern we described in the great decoupling, where impressions climb while clicks flatten.

Google's own guidance backs the boring answer

It is tempting to treat all of this as a reason to buy some exotic new tactic. Google is telling you not to.

In its AI optimization guide, last updated on June 29, 2026, Google is blunt: "optimizing for generative AI search is optimizing for the search experience, and thus still SEO." It explicitly says you do not need special markup files, artificially chunked content, or manufactured mentions. The generative features run on the same core ranking and quality systems as everything else.

That lines up with what we see in client work. The sites that get cited in research-query Overviews are the ones that would rank well anyway: real expertise, a clear technical structure, content that says something a summary cannot fully replace. There is no separate AI funnel to optimize. There is your existing content, sorted by intent, with the research layer treated as a visibility problem and the transactional layer treated as a conversion one. If you want the longer version of that argument, we made it in GEO vs SEO.

The practical move for Cyprus and US businesses

For a Cyprus SME or a US small business, the resource question is what matters. You cannot do everything, so do the thing that matches how search now behaves.

Audit your key pages by intent. For the research and comparison queries in your space, invest in content that deserves to be cited and structure that makes it easy to lift. For the transactional queries, stop worrying about AI Overviews that mostly are not there and make the page convert. That is the split, and it is a far better use of a limited budget than reacting to whichever CTR statistic went viral this week.

If you want a second pair of eyes on which of your pages sit on which side of that line, our free SEO review is a straightforward place to start, and we work with businesses across Cyprus and the US on exactly this.

The one-line version: AI Overviews took the research, not the sale. Build for both, but stop treating them as the same fight.

John Kyprianou

John Kyprianou

Founder & SEO Strategist

John brings over a decade of experience in SEO and digital marketing. With expertise in technical SEO, content strategy, and data analytics, he helps businesses achieve sustainable growth through search.

Related Articles

Google Universal Cart explained, a single cross-merchant shopping cart spanning Search, Gemini, YouTube and Gmail, guide by SEO Turtle
AI Search

Google's Universal Cart: When the Checkout Moves Off Your Site

At I/O 2026 Google launched Universal Cart, a single basket that works across Search, Gemini, YouTube and Gmail. For years SEO was about winning the click to your site. Universal Cart removes the click even for buyers. Here is our practitioner read on what actually changes and the unglamorous feed work that decides who wins.

June 26, 2026
A Search Console toggle switch labelled AI Overviews and AI Mode being switched off, illustrating Google's new opt-out control, by SEO Turtle
AI Search

Google Now Lets You Disappear From AI Overviews. Almost Nobody Should

Google quietly shipped a switch that removes your site from AI Overviews and AI Mode while leaving your normal rankings untouched. It sounds like a gift to anyone watching AI eat their clicks. We explain why it is a regulatory escape hatch built for a handful of publishers, not an SEO lever, and why flipping it makes most businesses invisible exactly where attention is heading.

June 25, 2026
An AI search answer with a sponsored result woven into the response text instead of sitting above it
AI Search

Ads Are Moving Inside the AI Answer. Here Is What That Means for Your Visibility

For 25 years the deal was simple: ads sat at the top, organic links sat below, and everyone knew which was which. In June 2026 both Google and OpenAI started weaving ads directly into the AI answer. We break down what actually changed, why it makes organic citations more valuable rather than less, and what a business in Cyprus or the US should do about it right now.

June 24, 2026
Google's reimagined AI search box explained, an intelligent box suggesting and expanding the user's query, guide by SEO Turtle
AI Search

Google Just Reimagined the Search Box. Here Is What It Changes for SEO

For 25 years the Google search box recorded the question you already had. The new AI search box helps you write a better one, suggesting and reshaping your query before you finish typing. That moves the decisive moment of search upstream, off the results page and into the box. Here is our practitioner read on what it changes and what to do now.

June 23, 2026
How to structure content so LLMs cite your website as a source, guide by SEO Turtle
AI Search

How to Structure Content So LLMs Cite You as a Source

LLMs do not rank your page, they lift a passage from it and quote it. That single change rewrites how you should write. Here is our tested, practitioner read on structuring content so AI answer engines extract and cite you, what the controlled research actually proves, and the schema and llms.txt myths to stop wasting time on.

June 18, 2026

Continue Your SEO Journey

Explore more expert insights and take action on your SEO strategy