[The Engines]
How Can Brands Adapt Their SEO Strategy for AI Overviews?
Google's May 6, 2026 update rewired how AI Overviews link to sources, and the click data explains why: being cited now matters more than being ranked.
The short answer
On May 6, 2026, Google added inline citations, hover previews, subscription labels, and forum attribution to AI Overviews and AI Mode, a direct response to organic click-through rates that fell 61% between September and November 2025. Pages cited inside an AI Overview now average 2.1% CTR against 0.9% for uncited pages, so the strategy question is no longer 'how do I rank,' it is 'how do I get cited.'
What Changed in Google's AI Overviews Update on May 6, 2026?
On May 6, 2026, Google shipped five changes to AI Overviews and AI Mode, and the headline change is where links live. Citations now sit inline, next to the specific sentence they support, instead of stacked in a list at the bottom of the answer.
The second change compounds the first: hovering over an inline link on desktop now surfaces a preview card showing the site name and page title before the user clicks. Google's stated reasoning is that searchers hesitate to click a link when they cannot tell where it leads, so the preview is meant to close that hesitation gap.
The remaining three changes are smaller but not trivial. Links from a searcher's news subscriptions get a 'Subscribed' label. AI Overviews now sometimes append suggested follow-up articles after the answer. And a new 'Community Perspectives' feature quotes Reddit and forum threads by handle when citing discussion-based content, not just websites.
- Inline citations placed next to the text they support, not grouped at the bottom
- Hover previews on desktop showing site name and page title before a click
- A 'Subscribed' label on links from the searcher's news subscriptions
- Suggested follow-up articles appended after the AI Overview answer
- Community Perspectives, which cites Reddit and forum threads by handle
Why Did Google Make These Changes Now?
The timing lines up with a click-through problem Google could not ignore. An analysis by Seer Interactive covering 5.47 million queries across 53 brands found organic CTR on pages appearing in AI Overviews fell from 2.52% in September 2025 to 0.76% in November 2025, a 61% decline.
Part of that early drop was arithmetic: impressions grew faster than clicks in October, which flattened the CTR without a real collapse in click volume. November was different. Clicks themselves fell, which is the pattern that tends to get a product team's attention.
CTR did climb afterward, from 1.3% in December 2025 to 2.4% by February 2026, an 85% jump in two months. But Seer's own researchers describe that move as a leveling off, not a recovery. The May 6 update reads as Google's next lever: if the answer format is going to keep suppressing raw click volume, making the links inside it more clickable is the available fix.
| Element | Before May 6, 2026 | After May 6, 2026 | What to do about it |
|---|---|---|---|
| Link placement | Grouped at the bottom of the AI Overview | Inline, next to the sentence it supports | Pair each claim with its evidence in the same sentence, so there is something specific for an inline citation to point to |
| Link context | Plain text link, no preview | Hover preview shows site name and page title on desktop | Keep titles and meta descriptions accurate and specific, they now double as the pitch inside the preview card |
| News content | No subscription signal | 'Subscribed' label on links from the searcher's news subscriptions | Mostly relevant to publishers, but a sign Google is testing trust signals beyond the page itself |
| Forum and social content | Rarely surfaced by name | Community Perspectives quotes Reddit and forum threads by handle | Monitor and participate in threads already ranking for your terms, they are now a citation surface |
| Post-answer engagement | The answer ends the interaction | Suggested follow-up articles appended after the answer | Publish the deeper follow-up piece a suggestion widget can actually point to |
Who Does the AI Overviews Update Affect?
It affects almost anyone ranking for a query where an AI Overview appears, but exposure is uneven by intent. Comparison queries triggered an AI Overview 95% of the time in Seer's dataset. Transactional queries triggered one only 5% of the time.
That split matters for who should act first. B2B SaaS and tech growth teams live on comparison and 'best [category] tool' queries, the exact intent most likely to render an AI Overview, which makes citation inside that answer a direct proxy for whether ChatGPT or AI Overviews recommend them over a named competitor. Local and service businesses see less AI Overview pressure on bottom-funnel transactional searches, but plenty on informational 'best [service] in [city]' and how-to queries, where the same citation dynamics apply.
The more useful split, across both groups, is cited versus uncited, not AI Overview versus no AI Overview. Pages cited inside an AI Overview averaged 2.1% CTR against 0.9% for pages that were not cited. Meanwhile, CTR on queries with no AI Overview at all rose from 2.8% to 3.8% over the same window. Both trend lines point the same direction: presence in the answer, not just presence on the results page, is what converts.
How Should You Respond to the AI Overviews Update?
The response is not to write less for AI systems and more for humans. It is to write content built to be extracted and cited in the first place, since that is now the difference between a 2.1% click rate and a 0.9% one.
A handful of tactics follow directly from the mechanics Google just shipped. Pair every claim with the specific evidence behind it in the same sentence or the one right after, since an inline citation needs something precise to point to. Keep title tags and meta descriptions accurate and specific, because they now double as the pitch inside the hover preview card. Monitor and participate in the forum and community threads already ranking for your terms, since Community Perspectives makes those threads a citation surface in their own right. And publish the deeper follow-up pieces that a 'suggested articles' widget can actually point to.
None of this is about tricking the model into citing you. Omnicite's own view, backed by client work, is that the mechanism is coverage, freshness, and quality at a scale most in-house teams cannot sustain, publishing 10 to 12 pieces a day rather than one a week, across ChatGPT, Perplexity, Gemini, Copilot, and AI Overviews rather than Google alone. LeadHaste, run under the same group, went from zero to 1 million monthly impressions and more than 200 AI citations, taking its Domain Rating from 1 to 24, in four months of that approach. A second client, DeepInspect.ai, went from zero to 150,000 monthly impressions in six weeks.
What Should You Track After the Update?
Track Citation Share, the percentage of relevant AI answers in your category that cite you, rather than average rank position alone. Rank without citation now converts worse than it used to, and the gap between cited and uncited pages is the whole story of this update.
Average position was built for a results page made of ten blue links. It was never built to measure whether an AI answer chose you over a competitor by name. Citation Count per day, Answer Presence across the full universe of questions in your category, and Share of Voice against named competitors are the metrics that actually move when an update like this one lands.
- Citation Share: the percentage of relevant AI answers that cite you
- Citation Count per day: raw volume of citations across engines
- Answer Presence: how much of your category's question universe you show up in
- Share of Voice: your citation footprint relative to named competitors
Source: Search Engine Journal, citing Seer Interactive, 2026-04-26
Key takeaways
- Google's May 6, 2026 update moved AI Overview citations inline, added hover previews, subscription labels, suggested follow-up articles, and Community Perspectives quoting forum threads.
- The update follows a 61% CTR decline on AI Overview pages between September and November 2025, per a Seer Interactive analysis of 5.47 million queries across 53 brands.
- CTR climbed from 1.3% to 2.4% between December 2025 and February 2026, an 85% jump that Seer itself describes as a leveling off rather than a full recovery.
- Comparison queries trigger an AI Overview 95% of the time versus 5% for transactional queries, so exposure varies sharply by search intent and by ICP.
- Pages cited inside an AI Overview average 2.1% CTR against 0.9% for uncited pages, making citation, not just ranking, the deciding factor.
- The response is content built to be extracted and cited: claim-evidence pairs, accurate titles for hover previews, forum participation, and follow-up depth, tracked against Citation Share rather than average position.
Omnicite Editorial. "AI Overviews SEO: What Changed and How to Respond" The Citation Report, Omnicite. https://omnicite.co/blog/how-can-brands-adapt-their-seo-strategy-for-ai-o/
Sources
Google's May 6, 2026 update added inline citations, hover previews, subscription labels, suggested articles, and Community Perspectives to AI Overviews and AI Mode. Search Engine Land, 2026-05-06
Organic CTR on AI Overview pages fell 61% from September to November 2025, based on 5.47 million queries across 53 brands. Search Engine Journal, 2026-04-26
CTR rose from 1.3% in December 2025 to 2.4% in February 2026; comparison queries trigger AI Overviews 95% of the time versus 5% for transactional queries; cited pages average 2.1% CTR versus 0.9% for uncited pages. Search Engine Land, 2026-04-24
Frequently asked questions
What did Google change about AI Overviews on May 6, 2026?
Google moved citations inline next to the text they support, added hover previews on desktop, a 'Subscribed' label for news subscription links, suggested follow-up articles, and Community Perspectives, which quotes Reddit and forum threads by handle.
Did click-through rates on AI Overviews actually recover?
Partially. CTR rose from 1.3% in December 2025 to 2.4% in February 2026, an 85% increase, but Seer Interactive, the firm behind the underlying research, describes that move as a leveling off rather than a full recovery from the 61% decline seen earlier in the year.
Which query types trigger AI Overviews most often?
Comparison queries triggered an AI Overview 95% of the time in Seer Interactive's dataset, while transactional queries triggered one only 5% of the time, which means comparison-heavy categories like B2B software carry the most exposure.
Does being cited in an AI Overview actually help click-through rate?
Yes. Pages cited inside an AI Overview averaged 2.1% CTR compared with 0.9% for pages that ranked but were not cited, roughly double the click performance for the same query.
How is responding to this update different from traditional SEO?
Traditional SEO optimizes for rank position on a results page. This update rewards content built to be extracted and cited within an answer, meaning claim-evidence pairing, accurate titles for hover previews, and presence in the forum threads AI Overviews now quote directly.
What metric should replace average rank position after this update?
Citation Share, the percentage of relevant AI answers in a category that cite you, tracked alongside Citation Count per day, Answer Presence, and Share of Voice against named competitors.