AN AI MODE ANSWER WITH AN OFFICIAL-PAGES BOX (ILLUSTRATIVE)
TOP OF THE ANSWER
Official page: book an appointment
Why it helps: the one place to complete this task
BELOW IT
The written AI answer
Inline links, news, reviews, everyone else
Decorative illustration only. Google has not published a design for this test.

Google says it will test a spot at the very top of its AI responses for official pages, and that matters most if you run the one site where a customer can actually finish a task. That’s the core of Google AI Mode official pages, an idea Google floated on October 15, 2026. Think booking, pricing, returns, applications, support. If you’re the place a person has to land to get it done, this test may point straight at you. Or at someone else with a similar name.

Below we cover what Google actually said, what it left out, what the click data says about links in AI answers, and a short checklist for making your task pages the obvious official answer. Treat the checklist as our read, not Google’s rules. For the wider picture, start with our AI search work, and see how optimizing for Google AI Mode changes once it’s the default.

7x
Growth in people using Preferred Sources since the start of 2026, per Google (no baseline given)
Google, Oct 15, 2026
8% vs 15%
Share of Google visits with a click on a regular result: with an AI summary vs without
Pew Research Center, July 22, 2025
1%
Visits where someone clicked a link inside the AI summary itself
Pew Research Center, July 22, 2025

What is Google testing with official pages in AI Mode?

Google said it will test a dedicated section at the top of an AI response that shows official web pages, with a line on why each link is useful. It’s meant for cases where there are “only one or two official places on the web” to complete a task or get exact information.

The announcement came from Hema Budaraju, Google’s VP of Product Management, in a post titled “More ways to connect with the web from Search.” It sat alongside a few other changes. Google says it has made “continued and substantial increases” to inline links in AI Mode and AI Overviews and will show them higher in responses more often. It’s also expanding news carousels that highlight Preferred Sources, adding thumbnails to links at the end of some AI responses, and testing clickable creator and publisher names on video results in AI Mode.

One thing worth saying plainly: it’s a test. The post uses future tense. PPC Land, which covered it on October 18, points out that Google gave no start date, no market and no test size. So nobody can tell you today that this box is live for your searches.

Which pages will count as “official,” and how will Google decide?

Google hasn’t said. It didn’t define “official” or give examples, and it didn’t explain how a page earns the label. Anyone telling you they know the rule is guessing.

What we can say is where the idea comes from. Google already puts a company’s own site up top when you search for that company by name. This test appears to stretch that habit from sites to tasks, so a search like “renew my license” or “book a table” might surface the one page that does the job. That second part is PPC Land’s reading, not Google’s wording, and we’d hold it loosely.

It also fits the direction we’ve seen elsewhere. When a brand name alone can trigger an AI Overview, what the answer says about you starts to matter a lot. We went through that in what your brand name now triggers in Google. The official-pages test is the task-level version of the same problem.

Will links in AI Mode and AI Overviews send me more traffic?

Google says it’s showing more links and placing them higher, but it published no click numbers, so nobody outside Google can say how much traffic that adds. The best public baseline is still older, and it isn’t encouraging.

Pew Research Center tracked the browsing of 900 U.S. adults in March 2025 and analyzed 68,879 unique Google searches, 12,593 of which produced an AI summary. On visits with an AI summary, people clicked a regular result link 8 percent of the time. Without one, it was 15 percent. Links inside the summary itself were clicked on 1 percent of visits.

Share of Google visits with a link click, by page type (percent)
Regular result, no AI summary

15%

Regular result, AI summary shown

8%

Link inside the AI summary

1%

Source: Pew Research Center, July 22, 2025. Browsing data from 900 U.S. adults, March 2025; 68,879 unique searches. This predates the October 2026 link changes, so it shows the starting point, not the effect of the update.

So the fair read is this. A top slot for an official page could matter more than any inline link, because it puts one clear action in front of a person who already wants to act. But “could” is carrying the sentence. Until Google shares data, plan for the traffic you can measure, not the traffic you hope for.

How do I make my page the official answer for a task?

Make it impossible to confuse. One task, one page, one obvious owner. Google hasn’t published criteria, so this is our checklist built on basic search hygiene and on how Google already treats brand queries.

  1. List your real tasks. Write down the five to ten things a person comes to you to do: book, pay, return, apply, cancel, get support, check status. These are your candidates.
  2. Give each task one canonical page. If three URLs answer “how do I return an item,” pick one, redirect or canonicalize the rest, and point your internal links at it.
  3. Put the action above the fold. The page should let someone start the task, not just read about it. A booking page with a booking form beats a blog post about booking.
  4. Say who you are, plainly. Use your full business name, address or service area, and contact details in text on the page, matching what’s in your Business Profile.
  5. Keep structured data honest. Mark up what’s visible on the page, and make sure it matches. Don’t add schema for things a visitor can’t see.
  6. Check that crawlers can reach it. Robots rules, login walls, heavy JavaScript and CDN settings can all hide your task page from Google even when it works fine for people.
  7. Clean up lookalikes. Old microsites, partner pages and third-party listings that describe your task can compete with you. Keep them accurate, or retire them.

None of this is new, and that’s sort of the point. If Google starts choosing official pages, the businesses that already did the boring work are the ones that get picked. Our guide on AI-ready architecture, schema and crawlability covers steps five and six in more depth.

What does Google say a page needs to appear in AI Mode links?

Google’s Search Central documentation says a page must be indexed and eligible to show in Search with a snippet. It adds that there are no additional technical requirements, and that you don’t need special files, AI text files or new markup to appear in AI Overviews or AI Mode.

The same documentation points to the usual fundamentals: let Google crawl the page (including through your CDN or host), link to it internally so it’s easy to find, give it a good page experience, keep the important content in text, match structured data to visible text, and keep Merchant Center and Business Profile information current. It also notes that AI Mode and AI Overviews can use “query fan-out,” running several related searches to build one answer, which can surface a wider mix of links than classic search.

That last point is useful for task pages. If the model branches into “how do I,” “how much,” and “where do I sign up,” each branch needs a page that answers it. A thin landing page that does one of the three leaves you out of the other two.

Does this only help big brands, and what should I do this week?

We don’t know yet, and Google’s wording suggests the test is narrow on purpose. It applies when there are only one or two official places to complete something, which describes plenty of small and mid-sized businesses as well as household names. A local clinic is the official place to book that clinic. The risk is for businesses whose task is also handled by aggregators, marketplaces or directories, because Google may see more than two candidates.

Here’s a simple week of work. Run the five or six task queries that matter to you in AI Mode and note who gets linked. Check whether your own task page appears at all, and whether a third party appears instead. Fix the thin spots from the checklist above. Then log what you see monthly, since the test could change, widen or quietly vanish.

One more option, if you want it: Google says Preferred Sources use grew more than 7x this year, though it gave no baseline and didn’t define “use.” If your audience reads your news or updates, our piece on the Preferred Sources play shows how to ask for the star.

One thing to check this week: open AI Mode, type the task your customers most often want to do with you, and see which page gets linked. If it’s not yours, that’s your first fix.

Key takeaway

Google is testing an official-pages box at the top of AI answers, but it hasn’t said how a page qualifies, where it will run or what it does to clicks. You can’t optimize for a rule that doesn’t exist, so spend the effort on the basics that would make you the obvious pick under almost any rule: one canonical page per task, a visible action, consistent business details, and clean crawl access.

Wondering which page Google would pick as yours?

We’ll check how AI Mode treats your top customer tasks, show you who gets linked, and tell you which pages to fix first.

Claim my free audit →

Sources

  • Hema Budaraju, Google, “More ways to connect with the web from Search,” Google Keyword blog, October 15, 2026.
  • PPC Land, “Google pushes links higher in AI answers as Preferred Sources use rises 7x,” October 18, 2026.
  • Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results,” July 22, 2025 (900 U.S. adults, March 2025; 68,879 searches).
  • Google Search Central, “AI features and your website,” documentation read October 2026 (page shows no publication date).
  • The section questions above are common query patterns for this topic; we did not verify search volume for each one.