The Scale of the Shift: AI Overviews Now Own 25.8% of Search
The headline number from StackMatix's January 2026 analysis: AI Overviews now appear in 25.8% of all US searches. A year ago that figure was in the single digits for most verticals. For informational queries — the long tail that used to make SEO work — 39.4% now trigger an AIO.
This is no longer an edge case. One in four searches your audience runs today puts an AI-generated answer above the blue links. That answer is synthesised from sources Google selected, cited at Google's discretion, and designed to keep the user on Google's surface. Your rank is still real. Your click is no longer implied.
What Dropped 46.7%: The Real Click Data Behind the Panic
The most rigorous public study so far looked at 68,000 queries and found a 46.7% relative decline in organic CTR when an AIO appeared. Less rigorous studies cluster in the 15–46% range depending on query type, with DataSlayer reporting a 61% CTR impact on some informational segments.
The single most consequential stat for anyone who has ever celebrated a #1 ranking: position 1 CTR falls from 28.5% to 11.2% when an AI Overview appears above it. You can do everything 'right' — rank first, own the featured snippet, have the schema dialled — and still lose 60% of your clicks to a block of grey text that Google wrote using your page as a source.
This is the number that forces the strategy change. Ranking is now necessary but insufficient.
Why B2B Tech Got Hit Hardest (82% AIO Exposure)
AIO exposure is brutally uneven by vertical. The 2026 breakdown:
- B2B Technology: 82% of searches trigger an AIO — up from 36% in 2025. The highest exposure in the index.
- Health and finance: consistently above 55%.
- Informational queries generally: 39.4%.
- E-commerce / transactional queries: just 4%, for now.
If you sell SaaS, cloud, cybersecurity, dev tools or MarTech, Google has decided your category is the one it is most willing to answer for you. The 'ultimate guide to X' template that built a decade of B2B content is now the template Google harvests to build the overview that stops the click from ever reaching you.
E-commerce teams reading this with a sigh of relief: the 4% number is a grace period, not a permanent reprieve. Google's Shopping AI mode is already expanding the surface. Read the B2B pattern as your 12-month forecast.
The March 2026 Core Update Verdict
The March 2026 Core Update was the first one tuned specifically to distinguish 'content written with AI' from 'content that deserves to be cited by AI.' The post-rollout data is unambiguous:
- Sites publishing generic AI output — the indistinguishable, competent, keyword-complete articles that have flooded the index since 2023 — saw traffic drops of 60–80%.
- Sites publishing original data, proprietary research and first-party case studies saw visibility gains averaging 22%, with some outliers reporting 71%+ lifts.
Google's message, finally made explicit in ranking terms: authority and originality are the 2026 signals. Brand strength, first-party data, demonstrable expertise and unique perspective are what the model is now trained to reward. Generic is not a neutral position any more. It is a penalty.
Original Data Is the New Moat
Here is the uncomfortable pivot every content team needs to make. The old model was: identify a keyword, write a good article about it, earn links, rank, collect traffic. That model is dead for informational queries. It worked because Google had nothing better to surface. Now it does — and the thing it surfaces is a synthesis of your content without your URL.
The new moat is content the AI cannot generate without you:
- Proprietary benchmarks from your own customer data.
- Surveys you ran. Not meta-analyses of surveys someone else ran.
- Case studies with numbers only you have.
- Point-of-view arguments grounded in something the model has no training data for.
- Frameworks named and owned by your brand.
If a competent AI could write your article from public knowledge, a competent AI already has. The March update confirmed that Google can now tell the difference — it is what the model is now trained to reward.
Getting Cited Beats Getting Ranked
The new SEO goal, clean and specific: get cited in the AI Overview while maintaining organic position. Citation beats ranking, because the citation is the click path that survives.
Practically, this means optimising for two audiences — the ranking algorithm and the summarisation algorithm — at once:
- Structure for extraction. Clear H2s, direct answers in the first 50 words of each section, definitions that can stand alone.
- Earn entity recognition. Brand, author and organisation schema. Consistent bylines. Knowledge-panel-ready biographies. The AIO cites entities Google already trusts.
- Ship original numbers. Google's AI preferentially cites sources with unique data points, because synthesising without attribution on a stat is a hallucination risk.
- Be the primary source, not the aggregator. Aggregator pages are exactly what AIOs replace.
How to Anchor Your Old Content in the New Reality
This is the blog post where we say the quiet part out loud: we are in the middle of refreshing more than 20 of Anjin's own posts for exactly these reasons. Every one of them was written in the pre-AIO model. Every one of them is being anchored to 2026 data, original framing and the new citation-first structure.
The refresh pattern, if you want to steal it:
- Audit for queries now served by an AIO. Prioritise the ones where you hold a page-one position — you have the most to protect.
- Add a dated 2026 update block at the top with 1–2 hard stats and a primary source link.
- Replace generic paragraphs with proprietary numbers or explicit point-of-view.
- Restructure for extractability — direct answers up front, H2s that are themselves queries.
- Re-publish with an updated-date, not just a silent edit. Google reads the signal.
Do that for every post that still has impressions, and you get most of the way back.
What This Means for Marketing Teams
Here is the operational problem you cannot ignore. The new playbook requires:
- Continuous refreshing, not set-and-forget publishing.
- Original data collection across every post.
- Schema, author authority and entity work on every page.
- Monitoring of AIO presence per query, weekly, not quarterly.
- A velocity of publishing and updating that a normal content team, agency retainer or freelancer stack cannot match.
You cannot hand-write and hand-refresh 50 posts a quarter with original data embedded in each. Not with a team of three. Not with a £10k/month agency. The arithmetic does not work. This is why the category of 'Marketing Operating System' now exists — because the job changed shape and the tooling had to follow, and Anjin is what that operating system looks like in production.
Anjin: The Marketing Operating System for the Post-AIO Web
Anjin is the Marketing Operating System built for exactly this reality. One platform that runs your content engine end-to-end: generation, refresh, schema, distribution, citation tracking, AIO monitoring, internal linking and performance — all inside a single OS, powered by agents that know your brand, your data and your positioning.
What Anjin replaces for AIO-era SEO:
- Your content agency — drafts, refreshes and republishes at the cadence the new game requires.
- Your SEO consultant — continuous technical, entity and schema work.
- Your analytics stack for search — AIO presence, citation tracking, CTR decay per query.
- The manual refresh project you are not going to finish — Anjin runs it as a standing operation, not a one-off sprint.
What it does that no retainer can:
- Ships refreshes the day a core update lands, not the month after.
- Embeds first-party data into every draft because it is plugged into your data, not working from scratch.
- Operates 24/7 against the entire archive, not the three posts this quarter's content lead has time for.
This is how a single operator, or a team of two, runs marketing at a scale that used to require a team of twelve.
The £888 Lifetime License — Offer Closing Soon
Lifetime access to Anjin for a one-time payment of £888. Not a subscription. Not a seat. Not a trial. One payment, unlimited use, for as long as Anjin exists.
The average marketing team spends £888 in about three working days on tooling, freelancers and coordination software. You're buying the platform that replaces most of it — once.
This price will not be offered again once we close our early-access cohort.
Claim your £888 Anjin lifetime license →Founders, agency owners and in-house marketers — this is how you run marketing at AI speed without the team, the burn, or another year of waiting.
Sources: StackMatix, Position Digital, EnFuse Solutions, DataSlayer, Searches Everywhere, Slayly, Wyoming News, SEO.com
