AI Overviews on branded queries: how to protect your brand searches

Last updated: 10 October 2026 · Data from September 2026 tracking

What the data actually says

Two sets of numbers are doing the rounds, and they're easy to confuse.

The first is a share of queries. The second is a count of brands. A brand can show an AI Overview for one query and not another, so "93% of brands" doesn't mean "93% of brand searches." Some sites quoting 83% or 93% as one figure are mixing these up.

Google hasn't announced any change related to the increase. So treat this as an observed trend in tracking data, not a confirmed product decision.

What could go wrong for a brand

Branded searches are your most valuable ones. The person already knows you, and they usually click your site. An AI Overview at the top of that results page changes what they see first. These are risks worth checking for, not guaranteed outcomes:

• Wrong or outdated facts. Old pricing, a closed office, a discontinued service, or a mix-up with a similarly named company.
• Someone else's version of your story. The summary may lean on a review site, forum thread or competitor comparison instead of your own pages.
• Fewer clicks to you. If the summary answers the question, some people may not click, the same pattern behind zero-click searches. That's plausible for brand queries, but it hasn't been measured yet.
• Your brand missing from your own answer. The AI Overview may appear and cite other sites, but not you.

A 6-step protection plan

1. Measure first. List your top 20 to 50 brand queries: your name, name plus "reviews", "pricing", "login", "contact", "vs [competitor]", and your product names. Search each one in a signed-out browser window and fill in a table like the one below. This takes an hour and tells you whether you have a problem at all. If you also want to see how much traffic AI tools send you, here is how to track AI traffic in GA4.
2. Check your branded click data. In Search Console's Performance report, add a Query filter set to Custom (regex) with your brand name and common misspellings. Compare clicks, impressions, CTR and position for the weeks before and after late September. Google says traffic from its AI features is included in the Performance report under the Web search type, so this is the right place to look, but you can't separate AI Overview clicks from the rest.
3. Make your own pages the clearest source of brand facts. Write plain, current answers to the questions people ask about you: what you do, who it's for, where you are, what it costs or how pricing works, how to contact you. Keep key facts in normal text, not only in images or scripts. Google's guidance for AI features is just standard SEO, with structured data matching the visible text. Covering your brand's topics clearly and consistently also builds topical authority. Our guide on optimizing content for Google AI Overviews covers the content side.
4. Fix inconsistencies across the web. Make your name, address, phone, description and links match everywhere: Google Business Profile (our local SEO work covers this), LinkedIn, social profiles, directories and any Wikidata entry you are eligible for. Claim your Knowledge Panel if you have one. If an AI Overview states something wrong, use its feedback option, and fix the source it relied on. Be realistic: these are sensible steps, but there is no solid evidence yet showing how fast, or whether, they change an AI answer.
5. Cover the third-party space. If review sites, forums or comparison pages dominate your brand results, respond to reviews, keep your profiles current, and publish honest versions of the pages people search for, such as "[your brand] vs [competitor]" and "[your brand] pricing."
6. Re-test monthly and log it. Repeat the query check, note any changes, and keep dated screenshots. When a wrong answer disappears, you want to know what you changed. For a structured way to run these checks across AI tools, the GEO & AI Citation OS includes a prompt testing and results log.

Brand query audit template

Copy this into a sheet, one row per query, each checked in a signed-out window.

What Google lets you control (and what it doesn't)

Per Google's own documentation on AI features and your website:

• A page must be indexed and eligible to show with a snippet to appear in AI Overviews. There are no extra technical requirements, and no special schema or AI text file to add.
• The documented controls are nosnippet, data-nosnippet, max-snippet and noindex. Google doesn't describe a separate opt-out for AI Overviews. It says AI is built into Search.
• Google-Extended is for limiting AI training and grounding in some other Google systems. It is not a switch for Search's AI Overviews.

What not to do

• Don't panic over a tracker number. Wait for click data from your own Search Console before you change strategy.
• Don't add fake or keyword-stuffed brand FAQ pages. It won't earn a rich result (Google ended FAQ rich results in May 2026), and thin pages can fall foul of Google's spam policies.
• Don't edit Wikipedia or Wikidata dishonestly. Those communities enforce sourcing rules, and a promotional edit can backfire.
• Don't treat brand ads as automatic. Some brands pay to hold the top of their own results. Weigh the cost against your actual organic click data first.
• Don't promise clients a fix. Nobody outside Google can guarantee what an AI Overview says about a brand.

FAQ

Sources

• DemandSphere: AI Overviews on branded queries tripled in September
• Search Engine Land: Google AI Overviews jump from 26% to 80% of branded queries (includes Chris Long's 100-brand test)
• Google Search Central: AI features and your website
• Onely: How to Fix Incorrect AI Brand Information (practitioner guidance, effectiveness unproven)