How to Adapt to Google's AI-Driven People Also Ask Answers

What changed in Google's People Also Ask answers?

Google's People Also Ask answers have shifted from website-sourced snippets toward AI Overviews. Auspia reported on September 9, 2026 that an analysis from AlsoAsked found AI Overviews in 97% of a 19.2 million-query English-language People Also Ask sample during the first week of September, compared with 12% in July 2025. The report also cited Allintitle data showing a 100% AI-generated result from August 2026. These are third-party measurements, not a public Google metric, so they should be treated as observed platform tracking rather than an official product announcement.

The old People Also Ask interaction was straightforward for publishers. A searcher opened a related question, saw a compact answer pulled from a page, and could follow the associated result. The new interaction puts a generated response between the question and the source page. A website can still matter, but it must now be selected as a supporting link or citation within the answer rather than merely supplying an extractable snippet.

Google describes AI Overviews as AI-generated snapshots with key information and links for deeper research. Its Search Central documentation says AI Overviews may use query fan-out, which runs related searches across subtopics and data sources while a response is generated. That expands the set of pages Google may consider, but it also means a page that matched one narrow question is competing with broader coverage around the answer.

This is a change in distribution, not proof that conventional search has disappeared. Google says AI Overviews show only when its systems determine they add something beyond classic Search. Organic results, snippets, internal linking, crawlability, and page experience remain relevant. The practical change is that People Also Ask presence alone no longer describes how a publisher appears when an answer is generated.

The sensible before-and-after view is therefore not website result versus no website result. It is a move from a visible source answer to an answer surface where attribution, links, and click behavior can vary by query. That calls for a measurement model built around Citation Share, Answer Presence, and traffic quality, rather than one report for a single SERP feature.

1. Before July 2025 tracking: 12% of sampled People Also Ask answers were AI-generated, according to Auspia's cited third-party data.
2. August 2026 tracking: 86% of sampled People Also Ask answers were AI Overviews, according to the same report.
3. Early September 2026 tracking: 97% of the 19.2 million-query sample was AI Overviews.
4. What to do now: audit question pages, retain Search eligibility, and monitor cited links on priority prompts.

Who does the People Also Ask change affect most?

The change most affects publishers that treated People Also Ask as a repeatable way to earn discovery for informational questions. That includes documentation libraries, how-to publishers, comparison content, education sites, and B2B teams that built pages around specific pre-purchase questions. A page can still be useful to Google, but its previous appearance in an expanded People Also Ask result is not the same opportunity when the answer is generated first.

Smaller sites may feel the shift more sharply when a narrow question was their entry point to a wider topic. A focused page could previously surface beside a related question even when the domain was not a prominent organic result for the original query. With AI Overviews, Google can assemble an answer from a wider set of searches and supporting pages. That does not establish a rule that larger brands always win, and Google has not published such a rule. It does make single-query visibility a weaker proxy for discoverability.

B2B SaaS teams should examine category, alternative, integration, and implementation questions. These questions often sit close to evaluation and can reveal whether a buyer encounters the company in an AI-generated explanation. A company that is absent from the cited set for those prompts has an answer-visibility problem even if several pages retain conventional rankings.

Local and multi-location businesses should examine service and location questions. Searchers asking for help with a service, a local comparison, or a practical constraint may now see AI-generated summaries in more answer surfaces. The work is to confirm what appears in the live results for the specific service area, then improve the pages and supporting evidence that can satisfy the question.

Content teams should also avoid treating this as a reason to delete question-answer structure. Clear answers remain useful for readers and align with Google's guidance to create helpful, reliable, people-first content. The lost assumption is that formatting a paragraph as a People Also Ask target guarantees a familiar traffic path. It never guaranteed that outcome, and it is less suited to the current surface.

1. Publishers dependent on informational question traffic should compare affected query groups before and after the change.
2. B2B teams should track citations for category, comparison, integration, and implementation prompts.
3. Local businesses should check priority service-area questions in the markets they serve.
4. Editorial teams should retain clear question-answer pages while replacing feature-only reporting with answer-surface measurement.

How should publishers respond to AI-generated People Also Ask answers?

Publishers should respond by making their pages eligible, useful, and easy to verify, then by measuring whether those pages appear as supporting links in AI-generated answers. Google says there are no additional technical requirements to be eligible as a supporting link in AI Overviews or AI Mode: a page must be indexed and eligible to show a snippet in Google Search. Google also says that meeting its requirements does not guarantee crawling, indexing, or serving.

Start with the pages that answer business-critical questions. Confirm that each page can be crawled and indexed, that its key answer appears early, and that the page gives a reader enough context to verify the claim. For factual content, cite the primary source near the claim, retain publication dates where they matter, and update pages when the underlying evidence changes. A generated answer has little reason to rely on a thin page that merely restates a question.

Build coverage around the question, not a template. Google's documentation says AI has may use related searches across subtopics and sources. A page about a product category should therefore connect naturally to implementation questions, definitions, alternatives, and evidence pages. Internal links make those relationships legible to readers and search systems. They also prevent a useful source from becoming an isolated page with no topical context.

Do not add speculative AI files or special schema solely for AI Overviews. Google says there is no special schema.org structured data needed and that publishers do not need new machine-readable files or AI text files to appear. Structured data can still help when it accurately is page content and serves an existing Search use case. It is not a shortcut to citation selection.

Measure the change with a prompt set, not a one-time search. Record the question, market, date, whether an AI Overview appears, the cited domains, your domain's presence, and the linked page. Repeat the sample on a schedule. This turns a platform change into a defensible baseline for AI search visibility rather than a collection of screenshots.

1. Confirm priority pages are indexed and eligible to appear with a Search snippet.
2. Put the direct answer first, then support it with dated primary evidence and useful context.
3. Connect priority pages to related definitions, comparisons, and implementation resources.
4. Track AI Overview presence, cited domains, cited URLs, and referral or conversion outcomes for the same prompt set.

What should teams stop measuring on its own?

Teams should stop treating People Also Ask appearances as a standalone success metric. A feature-level count can show that a page appeared in an interface, but it cannot show whether the page was cited in an AI-generated answer, whether a visitor followed the link, or whether that visit led to a meaningful outcome. The question now is whether the brand is present in the answer journey for the queries that matter.

Search Console remains important, but Google reports AI has performance within the overall Web search type. Google advises site owners to use Search Console and to combine it with analytics when evaluating traffic changes. That means a clean analysis needs grouped queries, page-level trends, and annotated dates for changes to content or the search surface. A sudden movement in clicks without a query-group view can produce the wrong diagnosis.

Track Citation Share as the percentage of relevant AI answers in a defined category that cite your domain. Pair it with Answer Presence, which shows how broadly the brand appears across the question set. Citation Count per day can show volume, but volume without a stable prompt universe can hide a loss on the questions closest to revenue. Share of Voice is useful when the same tracked questions reveal which competitors are being cited instead.

Do not assume an AI Overview click is identical to an old People Also Ask click. Google says it has observed that people who click from search results pages with AI Overviews are more likely to spend more time on a site. That is a useful directional statement from Google, not a published conversion benchmark for every publisher. Check your own engagement and conversion data before changing the value you assign to this traffic.

The reporting shift is simple. Move from asking whether a page won a People Also Ask placement to asking whether the brand is cited in the answers potential customers see, which page earns that citation, and whether the resulting visitor completes a meaningful action.

1. Stop using People Also Ask placement as the sole proxy for answer visibility.
2. Keep Search Console analysis grouped by query intent and annotated with known content changes.
3. Measure Citation Share and Answer Presence on a stable priority prompt set.
4. Validate traffic quality in your own analytics before assigning a commercial value to AI Overview clicks.

What should you watch next?

Teams should watch whether citations and supporting links remain visible in AI-generated People Also Ask answers, how often they appear by query type, and which source pages Google selects. Auspia noted that Google had not publicly confirmed the transition or published an official percentage for AI-generated People Also Ask answers. The 97% figure is useful as a dated market observation, but it should not be converted into a universal claim about every language, country, or query class.

Watch for variation before drawing conclusions. Test branded and non-branded questions, informational and commercial questions, and relevant locations where the business serves customers. Record the live result rather than assuming a result seen in one market applies everywhere. AI Overviews are conditional surfaces, and Google says they appear only when its systems determine they are additive to classic Search.

Watch for source churn, too. A citation that appears today can change after a page update, a competitor publication, or a Google system adjustment. That is why recurring prompt monitoring is more useful than a launch-week audit. It helps teams separate a temporary change from a durable visibility gap.

Finally, keep the editorial standard high. The response to fewer extractive snippets is not generic content written at greater volume. It is a stronger body of evidence: pages that answer a real question, show why the answer is credible, and connect to the rest of the topic. Rankings got you found. Citations help get you chosen.

For teams that need a durable operating model, the work is Citation Engineering: produce authoritative coverage, keep it fresh, and monitor whether the right answers cite the right source. It does not promise a citation count. It creates a disciplined way to compete for the answer surface that has replaced a familiar click path.

1. Monitor source links and citations in live results, not only whether an AI Overview appears.
2. Segment prompt tracking by intent, brand status, and relevant geography.
3. Treat the reported 97% figure as a dated third-party observation rather than an official Google percentage.
4. Refresh evidence-led pages and investigate citation losses against the exact query set that changed.

Sources

Source: AuspiaAI

Third-party tracking cited by Auspia reported People Also Ask AI Overview share moving from 12% in July 2025 to 97% in the first week of September 2026, based on a 19.2 million-query English-language sample. AuspiaAI, 2026-09-09

Source: Google Search Central

Google states that standard SEO practices remain relevant, that pages need to be indexed and snippet-eligible for supporting-link eligibility, and that no special AI Overview schema is required. Google Search Central, 2025-12-10