AIO Delta

Guide

What to do about pages that lose clicks to AI Overviews

2026-10-21

Once you've found which queries lost clicks while rankings held (see how to find them), the next question is what to do about them. There's no setting that brings the clicks back, and plenty of advice online promises more than anyone can deliver. This guide sticks to what Google has documented, what independent data shows, and a triage that keeps you from wasting a quarter rewriting pages that never mattered.

Step 1: Decide which pages are worth defending

Not every lost click is worth chasing. Ahrefs makes the point with an example: a commercial query like "email marketing tools comparison" may be worth defending, while "what is email marketing" probably isn't (Ahrefs).

Sort your affected pages into three groups:

  1. Revenue pages. Comparisons, alternatives, pricing, product-led how-tos. Their queries have led to sign-ups or sales. Defend these first.
  2. Audience pages. Guides that bring in newsletter sign-ups, links or repeat visitors. Worth improving if you have something original to add.
  3. Commodity pages. Definitions and facts that a summary answers completely. Accept some loss here, and spend your effort elsewhere.

Use your analytics conversion data, not traffic alone. Google says clicks from results pages with AI Overviews tend to be "higher quality", with users more likely to spend more time on the site (Google). Google hasn't published data behind that claim, so check your own conversion rates before and after.

Step 2: Fix the basics Google already asks for

Google's position is that AI features in Search run on the same ranking and quality systems as the rest of Search. The SEO basics still apply, and no special files, markup or schema are needed to appear in AI Overviews (Google, AI features and your website). For each defended page:

  • Make sure it's indexable and eligible to be shown with a snippet. Google says a page must be indexed and eligible for a snippet to be a supporting link in AI features.
  • Check that the title and description match what the searcher wants. Google's traffic-drop guide points to titles and snippets first when impressions hold but clicks fall (Google).
  • Make sure the important content is in text, with structured data that matches the visible page where you use it.

Step 3: Give people a reason to click

An AI Overview summarizes common knowledge well. It's worse at replacing things only you have. Google's AI optimization guide, published in May 2026, makes this the main recommendation: create "non-commodity" content. That means first-hand experience, original data or a distinct point of view, rather than restating what's already out there (Google, Optimizing for generative AI features).

For a page that lost clicks, ask what a summary can't give:

  • Your own data. Benchmarks, test results, prices you've checked, screenshots of real setups.
  • Tools and templates. A calculator, checklist, spreadsheet or sample file.
  • Depth on the next question. The summary answers "what is X"; your page answers "how do I do X in my situation" and "what goes wrong."
  • Clear, current specifics. Dates, versions, steps that someone actually followed.

Then make sure the title and snippet promise that thing, so the searcher can see why your link is worth more than the summary.

Step 4: Aim to be a cited source, not just a ranked one

Seer Interactive's data, from 3,119 queries across 42 organizations in Q3 2025, showed organic CTR of 0.70% for brands cited in an AI Overview, against 0.52% for brands that weren't (Seer, Sep 2025). Seer is careful to say this doesn't prove the citation causes the higher CTR. Still, cited pages consistently did better than uncited ones in its data.

Google doesn't offer a "get cited" lever, but some recent changes are worth knowing about:

  • Preferred sources. Since May 2026, users who pick your site as a preferred source can see it more in AI Overviews and AI Mode (Google, documentation updates). Google said in October 2026 that the number of people using preferred sources is more than seven times higher than at the start of the year (Google). If you have a loyal audience, it's reasonable to ask them to add you.
  • More prominent links. The same October 2026 announcement describes more inline links in AI Overviews and AI Mode, and a test showing official sites near the top for some searches.

Don't create pages for every possible fan-out query. Google's AI optimization guide specifically warns that mass-producing variations to chase AI features can fall under its scaled content abuse policy.

Step 5: Know your controls, and their cost

You have some control over how your content appears in AI features:

  • Snippet controls. nosnippet, data-nosnippet and max-snippet limit what Google can show from a page, including in AI features (Google). Restricting snippets can also make your regular result less appealing.
  • Search generative AI control. Since August 2026, Search Console has a setting that can exclude a site, or part of it, from Search generative AI features. Google says changes generally take a few days, and that the setting doesn't affect AI training; for that you use Google-Extended (Google, Search generative AI control).

Opting out isn't a free win. Excluded content stops earning impressions and clicks from those features, and the searcher still sees an AI answer, just one built from someone else's pages. Treat this as a business decision, for example for licensing reasons, not as a fix for falling CTR.

Step 6: Skip the shortcuts

A lot of "AEO" and "GEO" advice is easy to sell and hard to verify. Google's AI optimization guide says tactics such as llms.txt files or special "chunking" of content aren't used by Google Search (Google). Google's changelog adds that llms.txt isn't needed for Google Search (Google). It may still help other AI tools, but don't expect it to change your Google clicks.

Similarly, FAQ rich results stopped showing in Google Search in May 2026 (Google). FAQ sections can still help readers, but they won't bring back a rich result.

Step 7: Measure whether it worked

Changes to AI Overviews happen constantly, and CTR on AI Overview queries moves on its own. Seer measured a partial recovery in early 2026 across its client set (Seer, 2026). To tell your changes apart from Google's:

  1. Note the date of each change, page by page.
  2. Rerun the same Search Console comparison monthly, with the same thresholds.
  3. Keep a few affected pages untouched as a rough comparison group.
  4. Watch conversions from organic search, not just clicks.
  5. Check the Generative AI performance report by page to see whether AI-feature impressions changed (Google). It shows impressions only, not clicks.

Start with a list

All of this works better when you know which pages to work on. The free AIO Delta check takes a Search Console Performance export made with Compare dates. It shows the queries or pages where position and impressions held but CTR and clicks fell, with estimated missed clicks to help you prioritize. No login is needed. The numbers are estimates from your own data, not Google's.

FAQ

Should I rewrite every page that lost clicks to an AI Overview? No. Start with pages whose queries led to sign-ups, leads or sales. A definition page that never converted may not be worth defending, while a comparison or pricing page usually is.

Do I need special markup, schema or an llms.txt file to appear in AI Overviews? Google says no. Its documentation says no special files, markup or schema are needed for AI features, and its May 2026 AI optimization guide says tactics like llms.txt files aren't used by Google Search.

Should I opt my site out of AI Overviews? Only after weighing it carefully. Search Console's Search generative AI control can exclude your content from those features, but Google says that removes both the impressions and the traffic you get from them. It's a business decision, not a fix for CTR.

Does being cited in the AI Overview help? Studies by Seer Interactive found cited pages had higher CTR than uncited pages on AI Overview queries, but Seer says that doesn't prove the citation causes the difference.

How do I know whether my changes worked? Rerun the same Search Console comparison on a schedule with the same thresholds, and track conversions from organic traffic in your analytics. Compare against a few untouched pages so you can tell your changes apart from Google-wide shifts.

Sources