AI Overviews hit 82% of tracked brand searches in September

Google showed an AI Overview on most of the brand-name searches DemandSphere tracks, according to data it published on 1 October 2026. The share of its tracked branded keywords returning one rose from about 26% in early September to 82.06% on 29 September, so the summary Google writes about your company now sits on your most valuable results page.

The September numbers

DemandSphere tracks branded keywords once a day per keyword, across all markets and devices, through its DemandMetrics platform. Its report on branded AI Overviews, written by Ray Grieselhuber, shows the share holding between roughly 26% and 35% for most of the month before it jumped to 69.21% on 26 September and 90.48% on 27 September. It stood at 80.23% on 28 September and 82.06% on 29 September. DemandSphere counted an AI Overview whether or not the brand itself was cited in it.

Nectiv’s Chris Long spotted the same pattern from a different direction. He posted that “93% of the brands I tested have AI Overviews appearing”, naming Reddit, Salesforce, Amazon and Adobe among them, as Search Engine Roundtable reported on 30 September. Barry Schwartz noted that most of these summaries appear below the company’s own listing, with Adobe an exception where the AI Overview sat above its main page. Google’s own name and news publishers were left out. Google has made no announcement.

Your brand search has a new narrator

For most companies the branded results page is the one they have controlled most closely. Someone types your name because they already know you, then gets your homepage, your sitelinks, perhaps a knowledge panel and your own ads. Until September that page was mostly assembled from things you wrote or paid for.

With an AI Overview, Google becomes the author. It writes a paragraph about your company from whatever pages it judges relevant, which can include a review site, a Reddit thread, an old press story or a competitor’s comparison page. A prospect checking you out before a sales call, a journalist confirming a fact or a candidate reading up before an interview may now get Google’s summary before they reach your site. If that summary carries an old price, a discontinued product or a complaint from 2023, Google has put it on the page you rely on most.

Ownership is the harder problem. Brand searches usually belong to SEO, the summary’s content belongs to brand and comms, and the sources it cites often belong to PR. Our view at Manual Focus is that the head of brand or comms should own the AI Overview on your brand name, with SEO supplying the monitoring, because what it says is a reputation question before it is a ranking one. Our positioning audit pipeline is a good starting point for comparing what Google says with what you want said.

For now Google mostly places the summary below your homepage listing, where it is less likely to take the click than it would be above it, though the Adobe result shows Google will put it on top for some brands.

Five checks for this week

1. Search your own brand terms. Run your company name, your main product names and “[brand] reviews”, “[brand] pricing” and “[brand] vs [competitor]” in a private window on mobile and desktop. Screenshot every AI Overview.
2. Read each summary for errors. Mark anything out of date, wrong or unflattering, and list the sources Google links beside it.
3. Fix the sources you control. Update your about page, product pages and Google Business Profile. Ask PR to approach review sites, directories and other third-party pages that carry wrong facts, and request Wikipedia corrections on the article’s talk page with your connection disclosed rather than editing it yourself.
4. Set a branded click baseline. In Search Console, filter for branded queries and compare click-through rate before and after 26 September. The generative AI features report, which we covered in our post on Lens data in Search Console, shows impressions for pages appearing in AI Overviews.
5. Repeat every fortnight. Keep the same query list and screenshots so you can see whether Google’s summary, its sources and its position move.

Gaps in the evidence

DemandSphere hasn’t said how many branded keywords it tracks or which brands they belong to, and says it is still working on breakdowns by device and market, so there is no UK figure yet. Chris Long’s test is a social post, not a published study. Neither source measures clicks, so the traffic effect is unknown. Google hasn’t confirmed the change, and a jump this sharp over a few days could be a test it scales back. The brands Long named are large global companies, which may not reflect what a mid-sized company sees.

If you want help auditing what AI search says about your brand, talk to us.

Set this up as a routine that runs every other Monday morning.

I run marketing for [brand], [one line on what you sell] at [website], selling
mainly in [market]. Our main competitor is [competitor].

Search Google for each of these queries, on mobile and desktop, without being
signed in:
- [brand]
- [main product name 1], [main product name 2], [add more]
- [brand] reviews
- [brand] pricing
- [brand] vs [competitor]

For each search, record whether an AI Overview appears, the full text of the
summary, every source Google links beside it, and whether it sits above or
below our own website listing.

Give me a table with the columns: Date | Query | Device | AI Overview shown?
(yes/no) | Position (above or below our listing) | Summary text | Sources
linked.

Then compare it with the previous run and list anything that changed: new or
dropped AI Overviews, changes in what the summary says, new sources, or a
change in position. If a search wouldn't load, or you couldn't see the
results as a normal searcher would, say so in that row rather than leaving it
blank.

I run marketing for [brand]. Below are the AI Overviews Google showed for
searches on our brand, with the sources it linked beside each one:
[paste the table from the first prompt, or attach screenshots].

Attached are our current facts: our about page, product pages and Google
Business Profile details [attach or paste].

For each summary, mark every statement that is out of date, wrong or
unflattering, and compare it with our current facts.

Give me a table with the columns: Query | Statement in the summary | Problem
(out of date / wrong / unflattering) | What our current facts say | Sources
Google linked beside it | Who fixes it (us: own pages or Google Business
Profile / PR: review site or directory / Wikipedia talk page).

Then draft:
- the updated wording for any of our own pages that is out of date
- a short, polite correction request for each review site or directory
- a Wikipedia talk-page request that discloses our connection to [brand] and
  asks for the change, rather than an edit to the article itself

Don't send anything. If you can't tell which source a statement came from,
say so rather than guessing.

DemandSphere's tracking showed AI Overviews on most branded searches from
26 September 2026. I want a baseline for our branded clicks either side of
that date.

Attached is a Google Search Console Performance export, filtered to queries
containing [brand] and [main product names], with daily clicks, impressions
and click-through rate from [start date] to [end date].

Compare the weeks before 26 September with the same number of weeks after it.

Give me a table with the columns: Query | Clicks before | Clicks after |
Impressions before | Impressions after | CTR before | CTR after | CTR change
(percentage points). Add a total row for all branded queries, and sort by
CTR change, largest drop first.

Only report what the data shows, and don't say why anything changed. If the
date range is too short for a fair comparison, or some branded queries look
like they've been filtered out, say so before the table.