Why AI Overviews Grew 58% in 2026 (And What It Means)

Two completely separate research teams looked at Google in early 2026, measured two completely different things, and landed on the exact same number. 58%.

One was counting how much more often AI Overviews show up. The other was counting how many clicks they swallow. We sat with both data sets for a while because the coincidence felt too neat, almost suspicious. It isn't a coincidence so much as a pincer movement, and it explains why so many sites are watching rankings hold steady while traffic quietly leaks out the bottom.

So let's pull both numbers apart, show you where they come from, and tell you what we're actually doing about it in client campaigns. No vibes, no recycled LinkedIn takes. The real figures, the real sources, and our honest read on what matters.

What does the 58% AI Overviews increase actually measure?

The headline number comes from BrightEdge, who have been tracking AI Overviews since the day they launched. In their one-year analysis covering February 2025 to February 2026, AI Overview presence climbed from roughly 30% of tracked queries to about 48%. That jump, 30 to 48, works out to a 58% increase in coverage.

Worth saying clearly, because the number gets mangled everywhere. The 58% is not "58% of searches now have an AI Overview." It's the growth rate. Just under half of searches trigger one, and BrightEdge notes roughly 52% of queries still show no AI Overview at all. Plenty of search is still old-fashioned blue links.

Where it stops feeling abstract is the industry breakdown. Search Engine Journal's coverage of the same BrightEdge data lists how nine sectors moved over the year:

• Education went from 18% of queries triggering an AI Overview to 83%. That is not a typo.
• B2B technology climbed from 36% to 82%.
• Restaurants jumped from 10% to 78%.
• Healthcare sat highest of all, rising to 88% by late 2025.

The 58% average hides those wild swings. If you're in education or B2B SaaS, your reality is closer to "almost every search now has an AI answer sitting on top of it." If you're in a niche Google still treats cautiously, you might barely notice. Averages lie, basically, and this one lies more than most.

The bit nobody screenshots: AI Overviews are getting taller

Presence is only half the story. The same report found the average AI Overview now exceeds 1,200 pixels in height, up about 15% year on year. On a standard desktop viewport of roughly 900 pixels, that means the AI answer fills the entire visible screen before a single organic result appears.

You can rank first and still be below the fold. We've watched this happen on client dashboards and it is a genuinely strange feeling. The ranking report says "position 1." The traffic report says "where did everyone go." Both are correct.

The second 58%: the one that hits your traffic

Here's the number we actually lose sleep over. Ahrefs ran a study on 300,000 keywords using aggregated Google Search Console data, comparing December 2023 (before AI Overviews) with December 2025. Their finding: the presence of an AI Overview correlates with a 58% lower click-through rate for top-ranking pages.

Their own line on it is blunt. For every 100 clicks you could historically earn for a top-ranking page, Google now keeps 58. The position-1 CTR on AI Overview keywords dropped from 0.073 in December 2023 to 0.016 in December 2025. That is most of your traffic, gone, while your ranking didn't budge.

What makes this one sting is that Ahrefs measured a 34.5% drop back in April 2025. By December it had nearly doubled. The trend isn't flattening out, it's compounding, which is the opposite of what a lot of "AI Overviews are overblown" commentary predicted last year.

So the two 58%s aren't twins. One is how much more often AI Overviews appear. The other is how much harder they hit your clicks when they do. Stacked together, that's the squeeze.

And a third data point that confirms the pattern

If you suspect Ahrefs and BrightEdge are talking their own book, fair enough. So look at neutral data. The Pew Research Center tracked the real browsing behaviour of 900 US adults across nearly 69,000 Google searches in March 2025.

When an AI summary appeared, users clicked a traditional search result in just 8% of visits. Without a summary, they clicked 15% of the time, nearly twice as often. And clicking a link inside the AI summary itself? That happened in 1% of visits. One percent.

Three independent measurements, three different methods, same direction of travel. When we see that kind of agreement across sources that don't share a method or an agenda, we stop debating whether it's real and start asking what to do about it.

Why your rankings can stay flat while traffic falls

This is the part that confuses business owners the most, and honestly we don't blame them. The SEO dashboard is green. The Analytics graph is red. How can both be true?

Because the click economics changed underneath the ranking. You're still number one. There's just a 1,200-pixel answer sitting above you that resolves the user's question before they ever scroll. The BrightEdge data adds a twist here too: only about 17% of sources cited inside AI Overviews also rank in the organic top 10. Roughly five out of six citations come from pages that aren't even on page one of normal results.

Read that again, because it reframes the whole game. Being the top organic result no longer guarantees you'll be the cited source. They've become two separate competitions. We dig into the mechanics of one of them in our breakdown of how AI Overview citation rates work and what drives them, and the traffic side in our look at the click-through rate collapse and how to read it without panicking.

If you only track rankings, you're flying blind to half of what now decides your visibility. That's not a small adjustment to make. It's a different scoreboard.

What Google actually says you should do (and what we'd ignore)

Time for some myth-busting, because the AI Overview optimisation advice floating around is about 70% nonsense. A lot of it tells you to bolt on more schema, create special AI files, and add secret markup. We kept seeing it in the original version of this very article, frankly.

Google's own AI features documentation on Search Central says the opposite, in plain English: "There are no additional requirements to appear in AI Overviews or AI Mode." And on the schema question specifically: "You don't need to create new machine readable files, AI text files, or markup to appear in these features. There's also no special schema.org structured data that you need to add."

To be eligible as a cited source, a page only needs to be indexed and eligible to show in normal search with a snippet. That's it. The same fundamentals that have always mattered.

So here's our honest take. Schema is still useful for rich results, for entity clarity, and for machines parsing your content cleanly. We use it on nearly every project. But anyone selling "AI Overview schema" as a special unlock is selling you something Google has explicitly said doesn't exist. If you want the technical groundwork done properly rather than as a magic trick, that's the territory our SEO service and our notes on technical SEO strategies that still move the needle actually cover.

What does move the needle for AI Overview citations

Right, the useful part. Here's the process we run when a client wants to be the cited source rather than the buried link. It's not glamorous and it works.

1. Answer the question in the first two sentences of the relevant section. AI models lift clean, self-contained answers. Bury the answer under three paragraphs of preamble and you've handed the citation to whoever was more direct.
2. Publish something only you can publish. Original data, first-hand testing, a number nobody else has. Models reward sources that add information rather than rephrase it, which is exactly why we put so much weight on running original research to earn AI search visibility.
3. Build entity clarity, not just keyword coverage. The engines reason about brands and topics as entities. Tighten your internal linking, your About page, your authorship. Our findings on entity optimisation and knowledge graphs go deeper here.
4. Keep it current. Stale content gets quietly dropped from citations. We review cornerstone pages on a quarterly cycle, minimum.
5. Track citations as a metric in their own right. Not impressions, not vanity rankings. Are you actually being named in the answer? If you're not measuring it, you're guessing.

None of that is exotic. It's solid content and structure aimed at a reader who happens to be a language model. For the broader playbook, our guide to getting your brand into AI answers ties the whole approach together, and our dedicated AI search visibility service is where we run it for clients end to end.

Is brand visibility worth anything if nobody clicks?

This is the genuinely open question, and we'll resist the urge to pretend we've solved it.

If a user reads your answer in the AI Overview, sees your brand named, and never visits, did anything good happen? Sometimes yes. They've now associated your name with the authoritative answer to their question. That's a brand impression with intent attached, which traditional display advertising would charge a fortune to fake.

But let's not gaslight ourselves. A brand impression doesn't pay invoices the way a click that converts does. We've stopped telling clients "citations are the new traffic" because that's a comforting half-truth. Citations build authority and compound over time. They are not a like-for-like replacement for a visit that turns into a lead.

What's actually changed is that you need to stop relying on a single channel to do everything. Search used to be discovery, consideration, and conversion in one funnel. Now Google often eats the discovery layer and hands you the user only further down. So you build owned channels that don't depend on Google's mood: email, a real audience, repeat direct traffic. We unpack that survival approach in our zero-click search visibility strategy, and the brutal version of what happens when you don't diversify is laid out in how small publishers lost 60% of their Google traffic.

What we're telling clients to do right now

If you read nothing else, read this. Here's the short version of our 2026 playbook, drawn from what's working across the campaigns we run.

Split your scoreboard in two. Track organic rankings and AI citations as separate metrics, because they're now separate competitions with different winners. Optimising for one does not automatically win the other.

Defend your highest-intent queries first. Not every keyword deserves the same energy. The searches where someone is close to buying or booking are where a lost click costs you real money. Protect those, and obsess less about informational queries the AI was always going to answer itself.

Be the source, not the summary. If your content can be fully replaced by a three-sentence AI answer, it was always going to be. Publish the thing that can't be summarised away: the data, the tested process, the strong opinion, the real example.

Diversify before you're forced to. The sites panicking right now are the ones that treated Google as their entire business. Build the email list, the brand, the direct relationships while you still have traffic to convert into them.

The 58% growth in AI Overviews isn't slowing down, and neither is the 58% hit to clicks. That sounds grim, and for "rank and pray" SEO it genuinely is. For anyone willing to compete on being the most useful, most cited, most genuinely-worth-visiting source, the bar just got higher in a way that quietly favours people who do the work properly.

If you'd rather not reverse-engineer all of this from a dashboard at 11pm, that's literally our job. Tell us what your traffic graph is doing and we'll tell you whether it's AI Overviews, a core update, or something else entirely. We do this every day, and we'd genuinely enjoy the puzzle.