How Can Brands Adapt to the AI Overview's Impact on SERP?

What changed in the SERP when AI Overviews expanded?

AI Overviews can now occupy more of the results page before a searcher reaches classic organic listings. The change covered by On-Page's September 2026 analysis describes a design in which selected AI Overviews expand into a fuller answer on load, with an AI Mode-style follow-up prompt available immediately. That shifts the practical competition from winning a blue-link position to earning a supporting citation inside the answer.

Before the reported August 28 change, the overview could appear as a shorter snippet with a Show more control. On-Page reports that the updated treatment removes that extra interaction for selected queries and can place the first organic result below the fold on many screens. Google has not published a query list or the share of searches receiving the treatment, so brands should treat rollout scope as unknown rather than assume every keyword has changed.

Less visible organic real estate means fewer chances to win the first click. It also means a page can contribute to the answer without receiving a visit. This is why traffic reporting alone is no longer a sufficient read on search visibility. Brands need a separate record of where they are cited, where competitors are cited, and which queries still send visits.

Google's documentation says AI Overviews and AI Mode may use different models and techniques, so the links and answers they show can vary. It also says these has may use query fan-out, which searches related subtopics and data sources while forming a response. A single page built around one narrow keyword is less likely to cover the related evidence a complex answer requires.

Classic SERPs rewarded being found. The answer surface increases the premium on being selected as support. That is the distinction behind Citation Engineering: create useful coverage, maintain it, and make each page easy for an engine to understand and cite.

1. Audit priority queries manually on desktop and mobile, then save evidence of the visible SERP layout.
2. Separate queries with an AI Overview from queries that still show a conventional link-first layout.
3. Track whether your brand, competitors, or no external source is cited in the answer.

What does the before-and-after change mean for clicks?

The before-and-after matters because the new interface can reduce the amount of page space available to organic results before the fold. On-Page's analysis reports that a randomized field experiment found showing an AI Overview cut outbound organic clicks by 39.8 percent on queries where an overview appeared. The study was a working paper, not a peer-reviewed result, so the figure should be used as directional evidence rather than a universal traffic forecast.

That analysis also reports that hiding an overview increased outbound organic clicks from 0.37 to 0.62 per affected search and lowered the probability of a zero-click search from 0.73 to 0.54. Those values describe the study population and its observed queries, not every market or device. Still, they explain why a more expanded answer design is consequential for publishers and brands that depend on informational search traffic.

Its control interface used the earlier collapsed overview experience. The newer expanded treatment described by On-Page resembles an AI Mode-style answer inside the main results page. The analysis therefore makes a reasonable operational case for monitoring click changes closely after an overview appears, while avoiding a claim that every site will lose the same percentage of traffic.

Do not respond by trying to block AI Overviews reflexively. Google says sites eligible to appear in Search with a snippet are eligible to be shown as supporting links in AI Overviews and AI Mode, with no additional technical requirements. Removing snippets or indexing can reduce exposure in both classic and AI search. The better first move is to understand whether a page is being displaced, cited, or both.

A brand that loses clicks but gains citations may still be building consideration. A brand that loses both clicks and citations has a clear content and measurement problem. Those are different situations, and they require different fixes.

1. Compare a four-week baseline against a four-week post-change period for affected query groups.
2. Review organic click, impression, conversion, and branded-search movement together.
3. Record citation presence separately from Search Console performance data.

Who is most affected by AI Overview SERP impact?

Brands that rely on top-of-funnel informational pages are most exposed because these searches are easiest for an answer box to summarize. Publishers, software companies with extensive educational content, affiliate sites, local service businesses, and ecommerce teams all face a version of the same problem: a searcher may get a usable answer before seeing the traditional results.

B2B SaaS teams are affected when prospects ask category, comparison, integration, implementation, or evaluation questions. A query such as best customer support platform may invite an answer that names a small number of providers, summarizes selection criteria, and links to supporting evidence. A ranking below that answer can still exist, but it may no longer receive the attention its rank once implied.

Local and multi-location businesses are exposed when users ask for the best service in a city, compare providers, or need a quick answer about availability. In those cases, the answer may surface local sources, Business Profiles, or supporting pages. The goal is not merely to rank for a city term. It is to provide accurate, current evidence that can support the recommendation a person receives.

Teams with thin pages are also vulnerable. Google says existing SEO fundamentals remain relevant, including helpful and reliable content, internal links, page experience, images, videos, and structured data where appropriate. A page that has only a generic definition or repeats vendor copy gives an answer engine little reason to cite it over a clearer primary source.

Impact is uneven. Google states that AI Overviews are shown only when its systems determine they are additive to classic Search and that they often do not trigger. That makes a query-level inventory more useful than a sitewide assumption. Start with pages that already produce qualified traffic and identify which of their query themes now receive an overview.

1. Publishers should prioritize cited evidence, original reporting, and clear source attribution.
2. SaaS teams should prioritize category and comparison questions that shape vendor shortlists.
3. Local businesses should prioritize service-and-location pages with current operational details.
4. Ecommerce teams should prioritize product facts, comparison help, and accurate structured product information.

How should brands respond without chasing a shortcut?

Brands should respond by making their pages better evidence for an answer, not by searching for a special AI markup trick. Google explicitly says there is no special schema.org structured data required for AI has and no need to create new machine-readable AI text files. The durable work is the same work that helps a page earn trust in Search: clear information, technical eligibility, useful internal links, and content that directly resolves a real question.

Start with technical eligibility. Google says a page must be indexed and eligible to show a snippet in Google Search to be eligible as a supporting link in AI Overviews or AI Mode. Check indexability, canonical handling, crawl access, mobile rendering, page performance, and whether an accidental noindex or restrictive snippet directive is limiting the page. Fixing eligibility does not guarantee a citation, but failing it removes the opportunity.

Then rebuild priority pages around answerable questions. Put the direct answer near the top, define terms plainly, distinguish established facts from opinion, and link every material claim to the original source where possible. Add a dated statistic, a transparent comparison table, a product specification, a methodology, or original data when the topic supports it. These assets give an engine something concrete to use rather than a vague marketing statement.

Coverage matters as much as polish. Google says AI has can use query fan-out across related subtopics. A strong pillar should therefore connect to focused pages that answer the next questions a user is likely to ask. Each spoke should link back to its pillar and to relevant companion pages. This helps readers navigate and gives crawlers a clearer map of your expertise.

Keep the information fresh. An answer engine has little use for a page whose product details, prices, regulations, methods, or supporting citations have become stale. Assign owners to high-impact pages, preserve source dates, and use a recurring review schedule tied to changes in your market.

This approach does not promise citation counts or rankings. It creates a defensible basis for visibility: accurate content that is eligible to appear, covers the question well, and can be checked by a reader or an engine.

1. Keep strategic pages indexed and eligible for snippets.
2. Put a concise answer before the longer explanation.
3. Add primary sources, dated evidence, and transparent comparisons where relevant.
4. Connect related pages through purposeful internal links.
5. Review pages when source facts or customer needs change.

How should you measure visibility after AI Overviews?

Measure the result the user sees, not only the rank your software records. Search Console remains necessary because Google says AI-has appearances are included in the Performance report under the Web search type. But Search Console does not provide a separate, complete report that tells a brand every AI Overview citation it received, so it should be combined with direct SERP observation and conversion analysis.

Create a fixed prompt set for each category, product line, or local market. Include non-branded category questions, comparison questions, implementation questions, and high-intent local questions where relevant. Run the same prompt set on a recurring cadence, record the answer, save the cited domains, and identify whether your brand appears by name or as a linked source.

Use Citation Share as the headline metric: the percentage of relevant AI answers in a category that cite your brand. Pair it with Citation Count per day for volume, Answer Presence for breadth across the prompt set, and Share of Voice for the relative result against named competitors. These metrics describe the answer layer that a conventional rank tracker cannot fully describe.

Do not treat every citation as equal. A citation in a direct recommendation, a comparison, or a decision-critical answer may matter more than a citation in a broad definition. Tag prompts by intent and review downstream behavior. Organic sessions, assisted conversions, branded searches, demo requests, calls, and bookings can show whether visibility is becoming commercial demand.

Set a baseline before changing content. Then annotate the timeline with page releases, major source updates, technical fixes, and observed SERP design changes. Without this record, teams can mistake normal query volatility for a content win or loss.

Rankings have not disappeared. The more useful measure is whether your brand is among the sources an AI answer chooses when there is no page two to rescue a weak result.

1. Build a stable, intent-tagged prompt set.
2. Capture citations and cited competitors on a defined cadence.
3. Use Search Console for clicks, impressions, and query trends.
4. Connect visibility changes to conversions and qualified demand.
5. Keep evidence screenshots and source URLs for every major finding.

What should a 90-day response plan look like?

A practical 90-day plan starts with diagnosis, moves into content and technical remediation, then measures whether the work changes answer-level visibility. The first month is for establishing facts. The second month is for improving the pages that can change the result. The third month is for comparing a clean baseline with the first meaningful observation window.

In days 1 to 30, identify the pages and query themes that matter commercially. Pull Search Console data for the relevant pages, classify query intent, inspect representative SERPs, and document where AI Overviews appear. Capture which domains are cited. Review technical eligibility before rewriting content, because a strong page cannot be selected if Google cannot serve it with a snippet.

In days 31 to 60, improve the highest-priority page set. Add answer-first openings, source-backed facts, useful comparison structures, and internal links to related coverage. Remove unsupported claims. Where competitors are repeatedly cited because they publish a primary document, benchmark, study, or clear explainer, decide whether your brand can create an equally defensible asset rather than copying their format.

In days 61 to 90, re-run the prompt set and compare Citation Share, cited-domain mix, organic clicks, and conversion indicators. Segment the results by intent. Informational traffic may move differently from high-intent comparison traffic, and local queries may behave differently from national category prompts. Keep the conclusion narrow enough to match the evidence.

The outcome is a repeatable operating loop. You observe the answer surface, publish material worth citing, monitor the result, and update the work when the market or SERP changes. That is more reliable than treating AI Overviews as a one-time search launch.

1. Days 1 to 30: establish query, citation, traffic, and technical baselines.
2. Days 31 to 60: improve eligible pages with direct answers and source-backed assets.
3. Days 61 to 90: compare Citation Share, organic performance, and conversion outcomes.
4. Repeat the cycle for priority categories, products, and locations.

Sources

Source: Google Search Central

Google says AI Overviews and AI Mode may use different models and techniques, pages need no additional technical requirements beyond Search eligibility, and AI-has performance is included in Search Console Web data. Google Search Central, 2025-12-10

Source: On-Page

On-Page reports an August 28, 2026 expansion of AI Overviews for selected queries and cites a randomized field experiment reporting a 39.8 percent reduction in outbound organic clicks when an AI Overview was shown on affected queries. On-Page, 2026-09-01