SEO Galaxy is now GetMentioned. The platform stays the same. Want a demo? Book your slot here →
Five claims circulate as SEO trends 2026, and only two survive a look at the study behind them. This post names the sample size, market and year of collection for each, and says which of these trends falls outside our trade.
David Hahn · August 26, 2026 · 16 min read
The short version
Trend lists in search follow one pattern. They name eight to twelve developments. A percentage stands before each.
The numbers travel from list to list. Year of collection and method drop out on the way.
Artificial intelligence sits in almost every heading for 2026. Between them stand SEO experts with an assessment and no data.
We took the five claims circulating most as SEO trends 2026 and opened the primary source for each, checking year, sample and method.
Five checks are enough to place a number:
With one missing, the number cannot be verified.
The second check fails most often. A study from 2022 then sits in a list titled with the current year.
GetMentioned is a link building agency with its own placement marketplace. Full service in link building, not full-service SEO.
That sets the cut of this article. On backlinks, mentions and digital PR we write from our own work.
Technical SEO happens on our clients' websites. We name those trends and say whose job they are.
| Trend claim | Study behind it | Finding |
|---|---|---|
| 60 percent of searches end without a click | SparkToro with Similarweb, June 2026 | out of date, measured is 68 percent |
| AI answers cost clicks | Pew 2025, Ahrefs 2026, Seer 2026 | holds as correlation, direction turning |
| 73 percent of search happens outside Google | none findable | does not hold, value inverted |
| Backlinks are losing weight | no longitudinal measurement | does not hold, nobody measures it |
| Mentions without a link replace backlinks | Ahrefs 2025, Semrush 2026 | holds in part, mainly at Google |
"60 percent of all searches end without a click" appears in almost every trend list for 2026. It has a clean primary source and is two years old.
The 60 percent come from SparkToro's analyses. Rand Fishkin has published them for years with changing data suppliers.
The current level is higher. In the analysis from 9 June 2026, 68.01 percent of Google searches ended without a click.
The basis is Similarweb's desktop and mobile panel, US market, January to April 2026. Searches in the Google mobile app are excluded.
For 2024 the same analysis reports 60.45 percent. That value has travelled the trend lists ever since.
At SparkToro a zero-click search is any search ending without a click. Satisfied users count like users who give up.
Rewording a query also counts as a search without a click. It does not measure whether an AI answer sufficed.
A second caveat concerns the origin. Both values come from one analyst with a changed data supplier. No independent second measurement exists.
We found no comparable study for the UK or for Germany.
A result in the listing brings a visit less often than in 2024. That part holds.
Users take information from the results page more often. Visibility shifts from click count to mention.
Not this: "60 percent of all searches end without a click, so SEO no longer pays."
Better: "From January to April 2026, around 68 percent of US Google browser searches ended without a click, measured by SparkToro on Similarweb panel data. We found no comparable UK measurement."
The data here is at its densest. Three independent studies measure the same field with different methods.
Two features get mixed up regularly. AI Overviews sit atop the classic results page. AI Mode replaces that page and the user chooses it.
All three studies measure AI Overviews only. For AI Mode no comparable click figures exist, because that traffic cannot be broken out in reports.
The Pew Research Center is the only source here without a commercial interest. Its analysis from 22 July 2025 covers 68,879 Google searches by 900 US adults in March 2025.
The central value is 8 against 15 percent. With an AI summary users clicked a classic result in 8 percent of visits, without one in 15 percent.
Users clicked a source inside the summary in 1 percent of visits. Trend lists almost never name that.
Pew measures a results page's click rate. A single website's traffic goes unmeasured, and Pew claims no causality.
Ahrefs compares the same keywords before and after the rollout. The analysis from 4 February 2026 covers 300,000 keywords from Search Console data, desktop, December 2023 against December 2025.
For the top result Ahrefs measures an average click rate 58 percent lower with an AI Overview present. The earlier April 2025 value was 34.5 percent.
Ahrefs writes "correlates" in its own finding. The market goes unnamed, the clearest gap in this study.
Seer Interactive measures the same quantity through client accounts since early 2025. The analysis from 24 April 2026 covers 53 brands, 5.47 million queries and 2.43 billion organic impressions.
There the organic click rate on queries with an AI Overview rises again. From 1.3 percent in December 2025 to 2.4 percent in February 2026.
The gap remains. On informational queries the click rate sits around 38 percent below queries without one.
On causality Seer writes itself that it cannot be shown. Brands with higher authority would also be cited more often.
The claim holds in a narrower version. With an AI answer present, click rates come out lower, measured in three independent studies.
None of the three measures causality. Any phrasing like "AI Overviews cost X percent of traffic" exceeds the data.
The most recent figure in the field shows a recovery. Quoting only the Ahrefs value reproduces the state of February 2026.
How AI Overviews come about and where their sources come from is in our article on AI Overviews.
The claim runs "73 percent of search activity happens outside Google". It is the most interesting of the five, because no primary source leads to it.
The number has circulated since 2025 through social media and trade articles, under the label "Search Everywhere Optimization".
We found no study behind it. No panel research, no collection period, no sample, no method.
The arithmetic beside it names around 50 billion daily searches outside Google. No published measurement of that quantity exists.
Blanks like this are a result in themselves. They show where an industry copies itself.
A clickstream measurement across search-carrying platforms does exist. SparkToro published it on 3 March 2026 with data from Datos.
The basis is 41 editorially selected domains, desktop panel, the US plus the EU and the UK, calendar year 2025. In the fourth quarter Google took 73.7 percent of US desktop searches.
So the number 73 does exist. It describes Google's share and not the remainder.
By the same study the remainder splits across three groups:
The direction is worth noting. Fishkin calls the 73.7 percent a downward correction, because Google's share is usually put above 90 percent.
AI search systems came to 3.2 percent, well below what the trend lists suggest.
Not this: "73 percent of search happens outside Google, so you need a presence on every platform."
Better: "Google took around 74 percent of US desktop searches in the fourth quarter of 2025, measured by SparkToro on Datos panel data. AI tools came to 3.2 percent. Both sit far from the circulating numbers."
"Links are losing weight" has belonged in every trend list for years. For 2026 its stated reason is the rise of AI-assisted search.
Statements by Google staff get cited as evidence. We found no primary source with person, date and place of record.
Beside that sits a documented change of wording in Google's own documentation. We traced the archived versions in our link building statistics.
A longitudinal measurement of a falling link weight is one we did not find. We searched at Ahrefs, Semrush and Moz.
What does exist are correlations between link data and AI search visibility. They sit in the middle range and point upwards.
The analysis of 75,000 brands from 26 May 2025 gives 0.295 for referring domains, 0.326 for Domain Rating and 0.218 for backlink count, each against visibility in AI Overviews. A measured decline is absent.
The honest reading is plain. For link weight there is no time series in 2026, upwards or downwards.
That blank is the actual news. A trend nobody has measured for years stays a guess.
For our own work the finding changes little. Our link building strategy goes by relevance and quality.
Ten fitting references from one topic field beat a thousand arbitrary ones. That rule does not hang on an unknown weighting.
For SEO teams it means prioritising target pages first. For marketing leads the budget logic is unchanged.
"Mentions replace backlinks" is the youngest of the five trends. It comes with AI citability as its reason.
The mechanics speak for mentions. A system assembling an answer from several sources finds more reference points for a brand the more independent websites report it.
Ahrefs measured the relationship platform by platform. The analysis from 7 July 2025 covers around 76.7 million AI Overviews, 957,000 ChatGPT prompts and 953,500 Perplexity prompts from June 2025.
The Spearman correlation between brand mentions and visibility comes to 0.65 at Google AI Overviews. At Perplexity it is 0.30, at ChatGPT 0.15.
Author Patrick Stox adds the usual caveat himself. Correlation is no proof of causation.
The effect is therefore largely a Google phenomenon.
Not this: "Mentions work in every AI search."
Better: "Brand mentions on the web correlate strongly with a brand's visibility in Google AI Overviews, weakly at Perplexity and barely at ChatGPT."
A second study delivers the counter-finding. On 9 June 2026 Semrush analysed 3,981 domain appearances from 115 prompts in 14 countries across four platforms.
61.7 percent of cases were citations without the brand named in the answer. Around 25 percent were brand mentions without a linked source.
Mention and link are therefore decoupled. These figures do not show one replacing the other.
A controlled experiment where a placed mention produces an extra naming in an AI answer does not exist. Anyone selling an AI citability guarantee lacks the evidence.
Mentions and backlinks come out of one step of work with us. One editorial placement delivers both in a single article.
For classic search the link counts. For naming in AI answers the brand name in the text counts.
How a brand's share in AI answers gets measured, and where it fails, is in our article on Share of Model.
In planning meetings I now ask for the sample size before I look at the percentage. Most slides in a trend deck do not survive that question.
Three claims turn up almost everywhere in 2026 and are quicker to place.
"AI systems prefer content with demonstrable authority" is a plausible guess. We found no study measuring authority against namings in AI answers.
Better: "Ahrefs measures a correlation of 0.326 between Domain Rating and visibility in AI Overviews, across 75,000 brands. A preference for authoritative content is not measured by that."
"E-E-A-T becomes a ranking factor for AI systems in 2026" shortens two things. E-E-A-T stands for experience, expertise, authoritativeness and trust. It is a grid for human raters.
Better: "E-E-A-T is an assessment grid from the guidelines Google publishes for its human quality raters. No single measured value corresponds to it inside an AI system."
"Generative Engine Optimization replaces SEO" is the best selling of the three. The review of 45 GEO studies by Olivier Martinez, 15 July 2026, finds no stable cross-platform causal effect for any technique.
Better: "A review of 45 GEO studies from 2023 to 2026 finds no stable cross-platform causal effect for any technique. GEO adds to the work on visibility and does not replace it."
Part of these trends concerns work on your own website. We do not take it on, and we say where it belongs.
| Trend claim for 2026 | Where the work happens | Ours? |
|---|---|---|
| Core Web Vitals and mobile first | technical SEO on your website | no |
| Structured data and table of contents | in your own CMS | no |
| Keyword work and content planning | in the company's editorial team | copy yes, publishing no |
| E-E-A-T, author profiles and AI content rules | on your own website | no |
| Backlinks, mentions and digital PR | on other people's websites | yes, as a full service |
Core Web Vitals, mobile first and structured data are measurable and cleared in a few days. That work belongs to your team, an SEO agency or a technical provider.
The primary documentation on it is free and kept current. A summary from us would be shorter and older than the original.
In our experience the technical side is the bottleneck on few websites. It shows up in the first audit and is done after that.
For agentic systems that buy or book for a user, no reliable adoption figures exist. SparkToro's 3.2 percent for AI tools is the closest.
We watch the field and write no guide, because optimising for AI agents cannot be tied to any measured value today.
We have worked in this field for years. The build runs through placements at publishers matching the target page topic.
In the placement marketplace publishers stand side by side, with topic, organic traffic, SEO data and GetMentioned Score. The price shown per placement is what you pay.
Which of them fit your topic field is what MentionIQ proposes. The implementation can be handed to us as a full service.
The trend lists call for new metrics. Two are usable once you know their limits.
A visibility index maps rankings across a fixed keyword set. It measures positions. Clicks are no part of it.
What it stopped reflecting in 2025 is in our article on the visibility index.
For AI visibility several providers offer their own formulas. No provider publishes its prompt set.
That makes no number comparable between two providers. Two systems can report different values for one brand on one day, neither of them wrongly.
Such a value becomes usable as a time series inside your own account.
Four decisions carry through the year, whichever trend prevails.
Typical situations in which the effort pays off:
Of five widespread trend claims for 2026, two hold in a tighter version. Two get passed on with outdated or inverted numbers. One has no findable study.
The findings that hold point one way. A position in the search results brings a visit less often, and a brand's naming gains weight.
Mentions and references from matching topic fields are what a company can actively build. The rest stays observation.
Two developments hold. Searches end without a click more often, 68 percent in the US in spring 2026. And with an AI answer present, click rates come out lower, measured by Pew, Ahrefs and Seer.
In SparkToro's analysis on Similarweb panel data it was 68.01 percent of US Google browser searches from January to April 2026. We found no comparable study for the UK or Germany.
No measurement shows a falling weight over time. What is documented is a changed wording in Google's own documentation. The statements in circulation have no primary source. Correlations between link data and AI search visibility still point upwards.
Not on the available data. Brand mentions correlate at 0.65 with Google AI Overviews and at 0.15 with ChatGPT. In Semrush's ghost citations analysis, 61.7 percent of cases were citations without the brand named, around 25 percent namings without a citation.
Generative Engine Optimization, GEO for short, means optimising content towards being named in AI answers. A review of 45 studies from 2023 to 2026 finds no stable cross-platform causal effect for any technique.
As a technical basis they stay measurable and checkable. They belong to your own website and therefore to your team or a technical provider. We do not take that on.
Through a provider with its own prompt set, read as a time series inside your account. A comparison with other providers does not hold, because no provider publishes its prompt set.
About the author
Managing Director, GetMentioned
David has been building link acquisition and digital PR processes since 2016, first as an agency under SEO Galaxy, today as a platform with GetMentioned. He has scaled his own projects from zero to seven-figure monthly traffic and delivered thousands of campaigns for clients. Here he writes about what works in practice, and about what only costs budget.
Compare publishers, SEO data and prices in one place and book the placements that fit.
Verified link building statistics with source and year of collection. How strongly backlinks work, how many top-ranking pages hold, what placements cost and how common buying links is. Three widely quoted numbers failed the source check.
August 16, 2026 · 10 min read
An AI Overview condenses several websites into one answer above the organic results. This post shows where its sources come from, what Google actually says about optimising for it, and why the circulating click-loss figures measure three different things.
August 23, 2026 · 10 min read
Share of Model tracks a brand's share of all brand mentions in AI answers. Why the formula is uncontested but the denominator is not, and why two vendors can report different values for the same brand on the same day.
August 23, 2026 · 10 min read
