On this page11
Search Engine Roundtable reported on October 26, 2026 that AI Overviews now appear on over 80% of Google searches. That headline blends three studies that measure different things, and the article calls none of them a count of all searches. AEO vs SEO is the right frame. SEO earns the rank. AEO earns the mention inside the answer above it.
What Was Reported
Barry Schwartz of Search Engine Roundtable published the report on October 26, 2026. His headline reads “AI Overviews now show up for over 80% of Google Searches”. He adds that the share was about 50% in June of last year and about 15% the year before that. Those earlier figures link to older articles that this post has not re-checked.
The article points to three data sources. Here is what it says about each.
| Source | What the article says | What it covers |
|---|---|---|
| Semrush | A live chart tracks the trend. A separate study splits results by industry and query type, with no headline percentage given | Chart numbers can change |
| SE Ranking | AI Overviews appear on 82.5% of US branded searches | Branded queries in the US only |
| seoClarity | AI Overviews appear on over 62% of queries, from 55% to 84% by industry. Schwartz says it was shared with him about a month before | A different query set and an earlier date |
Read the Number Critically
Start with the headline. “Over 80%” sits in the title, but the article never assigns it to one study. The two percentages it does give, 82.5% and 62%, are not the same measure. One covers branded searches in the US. The other covers a wider set of queries, about a month earlier. Picking the larger one, or averaging the two, would be a choice, not a finding.
Three more gaps matter. The article gives no sample sizes, no query sets and no collection dates. It names a country only for the SE Ranking figure. And its single claim about non-branded searches is an assumption. Schwartz writes: “So this was specific to branded queries, and I assume it is higher for many non-branded queries.” The article offers no data for that sentence.
Branded searches also deserve suspicion as a yardstick. This point is an inference, not a finding. Someone who types a client’s name already knows the client. Whether an answer appears for that search says little about the harder job, which is being found by a buyer who does not know the name yet.
One more caution. All three sources are SEO software providers, and a rising number suits anyone selling tools for this space. That is context, not an accusation. It is a reason to quote each figure with its limits instead of repeating the headline.
None of this means the report is wrong. Google does appear to be showing more AI answers. Schwartz writes that “Google is really going all in on AI Overviews.” The careful sentence for a client is “reports put it between 62% and 82.5% in different studies”, not “80% of searches”.
AEO vs SEO: What Each One Measures
The debate often sounds like a fight. It is not. The two disciplines answer different questions about the same results page, so one cannot replace the other.
| Feature | SEO | AEO |
|---|---|---|
| Main question | Where does the page rank? | Does the answer name or cite the client? |
| Unit of success | A position in the results | A mention or citation inside an answer |
| What a buyer reads | A list of links | One written answer, often above the links |
| Evidence | Rank, clicks and impressions | The stored answer text, with the date |
| Main risk | Losing position | Being left out, or described wrongly |
The last row matters most. A page can hold position one and still not appear in the answer above it. Nothing in the report says that happens often. It says only that an answer is present on most searches in these studies, so the question is fair to ask. For the full comparison, read SKYA’s guide to AEO vs SEO.
Where Rank Reports Go Blind
Everything in this section is an inference, not a finding in the report. It follows from where the answer sits on the page.
A rank report can show position one for a buyer question. It cannot show whether the answer above that result names the client. A traffic drop can tell you that clicks fell. It cannot tell you whether an answer absorbed them. A steady rank can hide a change in the answer’s wording. A rival at position three can be named in the answer while the client is not. Each gap is the same gap. The rank sits below the answer, and rank tools do not read the answer.
Take a buyer question such as “best accounting software for a small agency”. Through the SEO lens, you find the client’s link and note its position. Through the AEO lens, you read the AI Overview first and note who is named. The two lenses can disagree. A client at position six might be named in the answer. A client at position one might be missing from it. Only the second lens finds that out.
Presence Is Not the Same as Impact
A critical reader should separate two claims. The first is that an AI Overview appears. The second is that it changes what buyers do. The report supports the first and is silent on the second. This is an inference, but a sound one: an answer can sit on the page and still be skipped, clicked through or trusted, and the report cannot tell us which.
That is why agencies should avoid forecasting traffic from this story. Say what the studies measured. Then measure the client’s own pages in Search Console, and read the answers for the client’s own questions. Your report should be built from those two records, not from a market-wide percentage. A client can act on its own data. It cannot act on an industry average.
What the Report Does Not Say
A client will ask five questions that the article does not answer. It does not say which study the 80% comes from. It does not give a share for non-branded queries, only an assumption. It says nothing about the effect on clicks. It gives no figures outside the US. And it does not say whether any particular client is named in those answers, which is the question that matters most. Write these gaps into the client report, so nobody mistakes a headline for a finding.
A Test You Can Run This Week
-
Pick ten buyer questions per client. Mix branded and non-branded ones.
-
Run each in Google from the client’s main market. Note whether an AI Overview appears.
-
Copy the full answer text. Save the date and the device.
-
Mark each answer: client named, client cited, client absent, or a rival named.
-
Note the client’s rank for the same query.
-
Compare rank and answer side by side.
Keep the wording of each question fixed. If you change the words, you change the test. Run the same set again after seven days, then after fourteen, and add a second market at the two-week mark. After thirty days, compare all readings. Look at how often the client is named, and whether that moved with rank or against it. A single reading shows a moment. A series shows a pattern.
Mistakes Agencies Will Make
The first mistake is repeating the headline. Telling a client that 80% of searches show an AI Overview claims more than the article supports. The second is treating branded data as proof for every query, which leads to a plan that ignores non-branded terms. The third is dropping SEO because AEO is rising. Rankings and links still decide which pages exist for an answer to draw on. The fourth is reporting rank alone, which leaves the answer layer unseen.
What to Ask Any Vendor
Good AI visibility tracking puts rank and answer on one page. By hand, that works for ten questions. It breaks at fifty across several clients and assistants. Ask any vendor the five questions below.
| Requirement | Why it matters | How SKYA handles it |
|---|---|---|
| Keeps the answer text and date | You need to compare answers over time | A stored reading behind every number, with an evidence panel |
| Reads Google AI Overview and AI Mode | Both show answers above the links | Nine assistants every day: ChatGPT, Gemini, Perplexity, Claude, Google AI Overview, Google AI Mode, DeepSeek, Grok and Copilot |
| Admits what it cannot read | A guessed score hides gaps | No answers read means no figure. A run that measured nothing gives its credit back |
| Checks claims | A mention can be wrong | A hallucination audit on every paid plan |
| Ties answers to traffic | Answer text and visits need one story | GA4 and Search Console integration, multi-client workspaces and white-label PDF reports |
SKYA publishes how every number is calculated. Its page on daily AI answer readings shows what a daily reading looks like. A client report built on AI visibility tracking should keep rank and answer side by side, so the client sees the whole page and not one half of it.
What to Tell Clients
Say that studies now put AI Overviews on most of the searches they measured, and that you are checking the client’s own questions. Say that rank still matters, and that you now also check whether the answer above it names the client. Say that the numbers come from different studies and that you quote each with its limits. Avoid three lines: “80% of all searches show an AI answer”, “rank no longer matters” and “the data is settled”. The report supports none of them.
Frequently Asked Questions
What is the difference between AEO vs SEO?
SEO aims to rank a page in the results. AEO aims to get a client named or cited inside the AI answer. They measure different things on the same page, so most clients need both.
Do AI Overviews really show on over 80% of Google searches?
Search Engine Roundtable used that headline on October 26, 2026. The article cites SE Ranking at 82.5% of US branded searches and seoClarity at over 62% of queries. It does not tie the headline to one study.
Does the report cover non-branded searches?
Not with data. The author says he assumes the share is higher for many non-branded queries, but the article gives no figure for them.
Does this mean SEO no longer matters?
No. The report says nothing about SEO ending. Rank still shows where pages sit. It does not show what the answer above them says.
What is AI visibility tracking?
It is the daily reading of what AI assistants say about a client, stored with dates and checked for wrong claims. SKYA reads nine assistants every day and keeps the answer behind every number.
Written by
SKYA Editorial Team
The SKYA Editorial Team publishes field-tested research on Answer Engine Optimization, Generative Engine Optimization and AI Visibility. Every article is reviewed by senior practitioners from the SKYA product and research groups before publication.
Related posts
Keep exploring the SKYA Journal.
