Search your own company name on Google this week and there is a good chance the first thing you see is not your homepage. It is a paragraph written by Google's AI, built from Wikipedia, YouTube, LinkedIn and whatever else its systems decided was relevant, with your site somewhere underneath.
That used to be rare. In September 2026, several independent tracking datasets showed AI Overviews jumping from a minority of branded searches to most of them in a matter of weeks. Google has not announced anything, which means there is no documentation telling brand owners what changed or what to do about it.
This guide is for marketers and business owners who need a practical response rather than a hot take. It covers what the data really shows (and what it does not), which kinds of branded query carry the most risk, a step-by-step audit you can run in an afternoon, how to fix the pages Google pulls from, and how to measure whether branded clicks are actually leaking in Search Console.
What four datasets say about branded AI Overviews in September 2026
The story broke at the end of September through independent trackers rather than a Google announcement. Each source measured something slightly different, so it is worth putting them side by side before drawing conclusions.
| Source | What was measured | Headline figure |
|---|---|---|
| DemandSphere (published 1 October 2026) | Daily share of tracked branded keywords showing an AI Overview, all markets and devices | 26.12% on 1 September, peak of 90.48% on 27 September, 82.06% on 29 September |
| Ahrefs (published 2 October 2026) | 232.8 million US desktop crawls, July to September; 92.5 million were branded keywords | Branded monthly averages of 61.3% (July), 65.1% (August), 73.3% (September); 82.91% on 29 September |
| Ahrefs brand-name panel | Exact names of 100 brands from Interbrand's Best Global Brands 2025, checked weekly | 5 of 98 usable results on 1 July; 63 of 100 on 30 September |
| Chris Long, Nectiv (via Search Engine Roundtable) | A SerpApi scrape of 100 enterprise brand names | AI Overviews for 93 of 100 brands, flagged on 30 September |
Two things stand out. First, the jump was fast: DemandSphere's daily series sat below 60% until 26 September, then spiked. Second, the Ahrefs monthly numbers show the gap between branded and non-branded queries closing. Non-branded queries averaged 78.3% in September, so brand searches are no longer the protected category they were. For context, Ahrefs' own November 2025 study (a different sample) found AI Overviews on just 13.1% of branded queries.
“Google is starting to show AI Overviews for a ton of branded queries.”
— Chris Long, quoted by Search Engine Roundtable, Google AI Overviews Return For Many Large Brand Names
What nobody has is an official explanation. As Semrush's news write-up notes, Google has not announced a change related to branded queries, and it has not said whether the rollout is finished. Ahrefs itself cautions that the current situation may not last. Treat the numbers as a strong signal of direction, not a permanent baseline.
Why an AI Overview on your name is not the same as Google rewriting your brand
It is easy to read "83% of branded searches" and assume Google now answers every brand query with its own summary, pushing your site off the page. The data is more nuanced, and the nuance changes what you should do.
- Presence is not citation. DemandSphere counts an AI Overview as present regardless of whether the brand itself is cited in it. None of the trackers report how often the brand's own website is linked.
- Placement varies. In Ahrefs' 100-brand panel, 48 of the 63 AI Overviews sat in position one (average position 1.49). But Search Engine Roundtable's spot checks found most appearing mid-page or lower, with Adobe a notable exception at the top. Placement appears to depend on the brand and the moment.
- Competitors rarely appear, for now. Of 58 Ahrefs panel responses with readable text, only 2 (3.4%) discussed competitors: Pepsi's answer mentioned Coca-Cola, and Intel's mentioned TSMC and Nvidia. Only one verified competitor-owned website appeared in 57 source lists.
- Some names are excluded. Search Engine Roundtable reported that searches for Google's own name and for news publications did not show AI Overviews at the time of checking.
- The sample is big brands. Interbrand's list and "enterprise brands" are household names. Smaller businesses may see different behaviour, and no dataset yet breaks results down by industry, device or query type.
“Ahrefs doesn't report clicks on these queries or how often the brand's website is cited in responses.”
— Search Engine Journal, AI Overviews Have Spread To Most Big-Brand Searches
So the honest summary is this: AI Overviews now appear on most large-brand searches in the tracked samples, the sources they cite are mostly third-party platforms, and nobody has yet published click data showing the impact. That gap is exactly why you need your own audit and your own Search Console baseline, rather than an industry average.
Where third-party pages take over your brand story
The citation data from Ahrefs' brand panel (16 to 29 September, 57 AI Overviews with source lists) is the most useful part of the research for practitioners, because it tells you which pages are shaping what Google says about a brand.
| Cited domain | Share of branded AI Overviews citing it | Who controls the content | Your lever |
|---|---|---|---|
| Wikipedia | 71.9% | Volunteer editors | Neutral, sourced corrections via talk pages; never edit your own article directly |
| YouTube | 38.6% | Often you (official channel) or third parties | Accurate channel "About" text, video titles and descriptions; up-to-date explainer videos |
| 22.8% | You (company page), employees | Company page tagline, description, size, HQ, specialties | |
| Yahoo Finance | 15.8% | Data providers | Correct profile data at the source providers (relevant for public companies) |
| 10.5% | You (page) | Page "About" fields and pinned information | |
| Walmart | 7.0% | Retailer and sellers | Product listing content via your retail account |
Source: Ahrefs, 2 October 2026. Percentages are of 57 AI Overviews in a 100-brand panel and do not add to 100% because each overview cites several sources.
Search Engine Journal points out that citations to YouTube and LinkedIn may well be brand-managed pages, though Ahrefs did not count how many. That is good news: at least a third of the sources in this list are profiles you can edit today. It is also why entity work, such as keeping Wikidata and Wikipedia accurate for your brand, now has a visible payoff on your own SERP.
Seven kinds of branded query and how exposed each one is
No published dataset breaks branded AI Overviews down by query type yet. The framework below is our own working assessment, based on how AI Overviews behave on informational queries generally and on what we see when auditing client brands. Use it to decide where to look first, then confirm with your own searches.
| Branded query type | Example | Likely user intent | Risk if an AI Overview answers it | Audit priority |
|---|---|---|---|---|
| Bare brand name | "acme" | Navigate to site, or learn who you are | Outdated description; site pushed down | High |
| Brand + reviews / complaints / legit | "acme reviews", "is acme legit" | Trust check before buying | Old complaints summarised as current; unbalanced sentiment | Very high |
| Brand + pricing / cost | "acme pricing" | Compare cost | Out-of-date prices or plans quoted as fact | Very high |
| Brand vs competitor | "acme vs globex" | Shortlisting | Third-party comparison framing wins | High |
| Brand + product or feature | "acme invoicing app" | Capability check | Missing or discontinued features listed | Medium |
| Brand + login / support / contact | "acme login" | Pure navigation | Wrong phone number or support route | Medium |
| Brand + jobs / careers / news | "acme careers" | Research the company | Stale headcount, leadership or HQ details | Low to medium |
The pattern is simple: the closer the query is to a purchase decision, the more an inaccurate summary costs you. A wrong headquarters city is embarrassing. A wrong price, or a three-year-old refund complaint summarised as "customers report issues with refunds", loses deals. If your brand already suffers from AI describing your old positioning, expect the same lag to show up in these overviews.
The four risks that actually cost money
Not every branded AI Overview is a problem. Many are accurate and arguably helpful. These are the scenarios worth acting on:
- Factual errors. Google says so itself: its help page states that “AI Overviews can and will make mistakes.” Wrong founding dates are trivial; wrong pricing, availability, licensing or safety information is not.
- Stale information. Rebrands, mergers, discontinued products and changed policies are the classic failure. The model can only summarise what the sources say, and Wikipedia or an old press article may be years behind.
- Narrative set by others. Chris Long's warning, quoted by Search Engine Journal, was that "competitors/external sources can now control the narrative of your brand directly in the search results." Today the data shows competitors rarely cited, but review platforms, forums and comparison sites can still frame the answer.
- Click erosion on navigational searches. If a summary sits above your site, some searchers will get what they need without visiting. Ahrefs references a separate earlier study finding a 58% lower click-through rate for the top-ranking page when an AI Overview is present. That was not measured on branded queries, which is why you need your own numbers (covered below).
Branded AI answers also interact with what happens in chat assistants. If ChatGPT or Gemini are already steering buyers away, as discussed in our piece on AI chatbots talking buyers out of purchases, the Google overview is one more place the same flawed narrative can appear.
Auditing your branded AI Overviews in one afternoon
This is the process we run for clients. It needs a browser, a spreadsheet and access to Search Console. Allow two to four hours for a typical brand.
- Build your branded query list. Start with Search Console: open the Performance report, apply the branded queries filter if your property has it, and export the top 50 to 100 branded queries by impressions. Add the high-intent patterns from the table above even if volume is low ("brand + reviews", "brand + pricing", "brand vs [top competitor]").
- Search signed out, in a clean window. Use a private browser window, signed out of Google, in the country you sell to. Search Engine Journal recommends recording dated snapshots of the AI Overview and its cited pages; do this for every query on your list.
- Capture four things per query. Whether an AI Overview appeared, where it sat (top, mid-page, lower), the full text, and every cited URL. Screenshots plus pasted text is enough.
- Fact-check every claim. Mark each sentence as correct, outdated, wrong or missing context. Pay special attention to prices, products, leadership, locations and anything about complaints or legal matters.
- Classify each cited URL. Owned (your site), managed (your LinkedIn, YouTube, Facebook pages), editable via process (Wikipedia, Wikidata, directory listings), or uncontrolled (news, forums, competitor or review sites).
- Check your own ranking position on the same query. Semrush suggests comparing the queries that trigger AI Overviews with your organic rankings for the same terms. If you still rank first but the overview sits above you, the problem is click share, not visibility.
- Report bad answers to Google. Google's help page explains that each overview has thumbs-up and thumbs-down icons, and a thumbs-down lets you report it as inaccurate or problematic with details. It is not a guaranteed fix, but it costs nothing and leaves a record.
- Repeat monthly. AI Overviews change as sources change. A monthly snapshot gives you a before-and-after record when you fix a source page.
A simple log keeps the audit useful over time:
| Date | Query | AIO shown? | Position | Accuracy issue | Cited URLs (type) | Fix owner |
|---|---|---|---|---|---|---|
| Example: 14 Oct | acme pricing | Yes | Top | Quotes 2024 plan prices | Wikipedia (editable), old review blog (uncontrolled) | Marketing + PR |
| Example: 14 Oct | acme login | No | n/a | None | n/a | n/a |
| Example: 14 Oct | is acme legit | Yes | Mid-page | Cites 2022 refund complaint as current | Forum thread (uncontrolled), Facebook page (managed) | Support + reputation |
If you want this to run continuously rather than by hand, the same log can feed the tracking setup described in our guide to building an AI visibility dashboard.
Fixing the pages Google pulls from
You cannot edit an AI Overview. You can change what it is built from. Google's documentation is clear that there is no special trick involved:
“There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.”
— Google Search Central, AI features and your website
The same page explains that AI Overviews use a "query fan-out" technique, running several related searches across subtopics to find supporting pages. For a brand query, that means Google may pull your about page, your pricing page, a Wikipedia entry and a review site in one go. Your job is to make the accurate sources the clearest and most consistent ones.
Owned pages: make the facts easy to lift
- Put a plain-English company description in the first paragraph of your about page: what you do, for whom, where, and since when.
- Keep a current pricing page, even if it only gives ranges or "from" prices. If prices are custom, say so explicitly; silence invites stale third-party numbers.
- Publish a dated "what changed" or company news page after rebrands, acquisitions or discontinued products.
- Mark up your homepage with consistent Organization schema (name, logo, sameAs links to your official profiles). Google says no special schema is required for AI features, but consistent entity data helps every system that reads your site.
Managed profiles: the quickest wins
LinkedIn, YouTube and Facebook appear in a large share of the citations Ahrefs logged. Update the description fields so they match your about page word for word on the key facts. Delete or update old "About" text that mentions products you no longer sell.
Editable-by-process sources
Wikipedia and Wikidata need patience and neutral sourcing. Propose corrections on the talk page with independent references rather than editing your own article. For a full method, see our guide to fixing wrong or outdated brand information in AI.
Uncontrolled sources
Old forum threads and review sites cannot be edited, but they can be outweighed. Respond publicly to complaints with what changed, earn fresh reviews, and pitch updated stories to the publications that cover your sector. Our AI citation outreach workflow covers how to prioritise which publications to approach. If the issue is reputational rather than factual, it is worth involving an online reputation management team early.
One control to avoid: Google's documentation lists nosnippet, data-nosnippet and max-snippet as ways to limit what appears from your pages. Applying them to your own brand pages does not stop an AI Overview appearing; it only removes your accurate content from the inputs and leaves third parties to fill the gap.
Measuring branded CTR in Search Console without fooling yourself
Everyone wants to know whether this is costing clicks. Here is how to find out for your own brand, with the limits of each tool stated up front.
Use the branded queries filter, and know its limits
Google introduced the branded queries filter in November 2025, and Search Engine Journal reports it opened to all eligible sites in March 2026. It shows clicks, impressions, CTR and position for queries Google's AI-assisted system classifies as branded, including typos and unique product names. It is only available for top-level properties with enough query volume, and John Mueller has said queries cannot currently be added to the classification manually. That means your branded set will not match a vendor's keyword list, so compare like with like.
Compare CTR, not clicks
Semrush's advice here is right: compare branded-query CTR for a period before and after AI Overviews appeared, because branded click volume moves with demand (campaigns, seasonality, news). CTR normalises for that. We prefer four-week windows to one-week windows for smaller brands, because daily noise is high. For a deeper method on separating ranking loss from AI Overview loss, see reading stable positions with falling clicks and our walkthrough of using the branded queries filter to isolate update impact.
Know what the other reports cannot tell you
- Generative AI performance report: available globally for all sites as of 31 August 2026, according to Search Engine Journal. It shows how often your links appeared in AI Overviews and AI Mode, but it has no query dimension, so you cannot isolate brand searches. Mentions without a link are not counted.
- Standard Performance report: Google's documentation says traffic from AI Overviews and AI Mode is included under the "Web" search type, mixed in with everything else.
- GA4: visits from AI Overview links are categorised as Organic Search. The AI Assistant channel added in May 2026 does not capture them.
A worked example (hypothetical numbers)
Suppose a mid-sized software brand exports branded data for two four-week windows:
| Window | Branded impressions | Branded clicks | CTR | Avg. position |
|---|---|---|---|---|
| Before: 18 Aug to 14 Sep | 120,000 | 66,000 | 55.0% | 1.1 |
| After: 15 Sep to 12 Oct | 132,000 | 63,400 | 48.0% | 1.1 |
Clicks fell only 3.9%, which a dashboard might shrug off. But impressions rose 10% (a campaign drove more brand searches), and CTR dropped 7 percentage points with position unchanged. At the "before" CTR, 132,000 impressions would have produced 72,600 clicks; the brand actually received 63,400. That is roughly 9,200 missing branded clicks in four weeks, or about 13% of expected volume. With position flat, Semrush's rule of thumb applies: if rankings hold but CTR drops, investigate the AI Overview first. The next step is to check which of the highest-impression branded queries now show an overview, using the audit log above.
If those missing clicks were high-intent (pricing, login, demo), it is worth testing whether a brand search ad recovers them. Whether that is worth paying for depends on your competition; our guide on bidding on your own brand name walks through the decision.
How branded AI Overviews change your wider search plan
This shift reinforces a broader point: ranking first for your own name is no longer the same as controlling your brand SERP. Three adjustments make sense for most businesses:
- Report brand visibility and brand clicks separately. A brand can be "number one" and still lose clicks. Add AI Overview presence and accuracy to your monthly reporting alongside branded CTR.
- Give entity and reputation work an owner. Profiles, Wikipedia accuracy and review responses used to be nobody's job. They are now inputs to the first thing people see when they search your name.
- Treat Google and chat assistants as one problem. The same sources feed both. Our guide to SEO that works for Google and AI search covers the overlap, and zero-click marketing explains how to build brand recall when fewer searches end in a visit.
Questions brand managers are asking about AI Overviews on their name
Can I turn off AI Overviews for my brand name?
No. Google treats AI Overviews as a core Search feature, and there is no setting for site owners to opt a query out. Snippet controls only limit how your own content is used, which usually makes the answer worse, not absent.
Did Google officially announce this change?
No. As of early October 2026, Google had not announced a change related to branded queries, and the figures come from third-party trackers (DemandSphere, Ahrefs, and Chris Long of Nectiv).
Are competitors showing up in AI Overviews for my brand?
Rarely, based on current data. Ahrefs found only 2 of 58 readable panel responses discussed competitors, and one competitor-owned site among 57 source lists. Check your own "brand vs competitor" queries, where comparison content is more likely.
Which sources does Google cite most for brand names?
In Ahrefs' panel, Wikipedia (71.9%), YouTube (38.6%) and LinkedIn (22.8%) led. Several of those can be brand-managed profiles, so start your fixes there.
How do I know if branded clicks are dropping because of AI Overviews?
Compare branded CTR (not clicks) before and after in Search Console with the branded queries filter. If average position is stable but CTR falls on queries that now show an overview, the overview is the likely cause.
Does this affect small businesses too?
The published data covers large brands only. Smaller brands may see different rates, so run the audit on your own queries rather than relying on the industry figures.
Checking what Google’s AI says about your brand this week
Do three things in the next five working days: export your branded queries and baseline CTR from Search Console, run a signed-out audit of your top 20 branded and high-intent queries, and update the three managed profiles (LinkedIn, YouTube, Facebook) and your about page so the facts match. That alone covers most of what Google is currently citing.
If you would rather have specialists run the audit, fix the sources and set up ongoing tracking, our SEO team handles branded SERP audits as part of AI search work. Get a quote and we will review your brand SERP before the call.
Further reading on AI Overviews brand searches
- AI Chatbots Talked 57.5% of Users Out of Buying: What Brands Should Do
- The LLM Positioning Lag: Why AI Still Describes Your Old Brand
- Zero-Click Marketing: How to Execute and Measure It
Sources for this branded AI Overviews analysis
- Semrush: Google AI Overviews now appear for most major brands (6 October 2026)
- Ahrefs: AI Overviews Currently Appear for 83% of Branded Searches (2 October 2026)
- Search Engine Journal: AI Overviews Have Spread To Most Big-Brand Searches (9 October 2026)
- Search Engine Roundtable: Google AI Overviews Return For Many Large Brand Names (30 September 2026)
- Scalevise: Google AI Overviews Surge on Branded Queries, Data Shows (summary of DemandSphere data, 3 October 2026)
- Google Search Central: AI features and your website
- Google Search Help: Find information in faster and easier ways with AI Overviews
- Google Search Central Blog: Introducing the branded queries filter in Search Console (November 2025)
