Google AI Overviews Now Trigger on 83% of Branded Searches. Your Website Isn't the Risk.

AI Overviews now appear on 82.91% of US branded keyword searches as of September 29, 2026 — up from 61.3% in July. Chris Long tested 100 enterprise brands in September; 93 triggered AI Overviews. Wikipedia was cited in 71.9% of those responses. LinkedIn in 22.8%. Your own website was not in the top three citation sources.

That is not a GEO metric. That is a brand control metric. And for most companies, it is currently failing.

What the Numbers Show

Three independent data sources converged on the same finding in early October 2026:

Ahrefs (Ryan Law, October 2) measured the AI Overview trigger rate for branded keyword searches across the US desktop index. DemandSphere (October 1) ran a parallel dataset. Chris Long at Nectiv tested 100 enterprise brands in September and published the citation source breakdown.

The progression is steep:

That is a 21.6-point increase in two months. The shift is not gradual.

The top external citation sources inside branded AI Overviews, from Long's 100-brand test:

Why It Matters

1. Brand absence is now a scored gap, not a neutral condition. When 83% of branded searches trigger an AIO, a brand that does not appear in that AIO is being displaced — not ignored. The AI is answering the query with someone else's framing of your company.
2. Wikipedia hygiene is brand-critical, not optional. At 71.9%, Wikipedia is the single most influential source in branded AI Overviews. A Wikipedia article with outdated founding date, wrong employee count, or missing product description will be surfaced to 7 in 10 branded AIO responses. That is not an encyclopedia problem — that is a brand narrative problem.
3. YouTube and LinkedIn are citation vectors, not distribution channels. The marketing rationale for YouTube (38.6%) and LinkedIn (22.8%) has changed. They are no longer just audience channels. They are citation inputs to the AI layer that answers brand queries. An outdated LinkedIn company page with 2022 headcount is a liability, not an irrelevance.
4. The 7 brands that didn't trigger AIOs are the dangerous ones. In Long's 100-brand test, 7 enterprise brands did not trigger an AIO. That sounds safe. It means Google does not have enough structured data to generate a confident answer about those brands. That is a brand authority failure, not a win.

What to Do

1. Run a branded AIO audit this week.
   • Search your brand name (and 3–5 variants: product name, founder name, category + brand) in Google US.
   • Screenshot what appears in the AI Overview.
   • Note: which sources are cited? What claims does the AIO make about your company? Are they accurate?
   • If no AIO triggers — that is the finding. Lack of structured entity data is the likely cause.
2. Treat Wikipedia as tier-1 brand infrastructure.
   • If you have a Wikipedia page: audit it for accuracy. Founding date, HQ, employee count, product description, key people. Every inaccuracy there is an inaccuracy in 71.9% of branded AIOs.
   • If you don't have one: assess whether you meet notability criteria. If you do, getting one live is now a brand priority, not a PR nice-to-have.
   • Support Wikipedia accuracy with your own pages: structured Organization schema, consistent About page data, and third-party press citations all feed the entity graph that Wikipedia editors rely on.
3. Audit your LinkedIn company page for citation accuracy.
   • Update: employee count, founding year, tagline, specialties list, location.
   • Ensure your About section matches your website's Organization schema exactly — inconsistency gives AI systems a reason to doubt the entity match.
   • Note: LinkedIn is valuable for Google AIOs and ChatGPT (33% citation rate on how-to queries). It is near-zero for Gemini conversational (0.0%, AmICited June–July 2026 data). Prioritise for Google and ChatGPT contexts.
4. Publish a YouTube channel About page that is structured and entity-linked.
   • At 38.6% citation rate, YouTube is your second-highest brand citation vector.
   • Your channel About section should describe what you do in one clear sentence, include your website URL (for entity linking), and match the product/service language on your homepage.
   • Video titles and descriptions for brand-related content (product demos, founder interviews) should include your exact brand name — that is what Google indexes and cites.
5. Add Organization schema with sameAs to your homepage.
   • The sameAs array should list your Wikipedia page, LinkedIn company page, YouTube channel, and Wikidata entry.
   • This is the explicit signal telling AI systems these are all the same entity. Without it, Google has to infer the connection — and may fail.

Chad's Take

The uncomfortable thing most brand managers don't want to hear: your website is no longer where your brand story is being told at the first touchpoint. When someone Googles your company name, they now hit an AI-generated summary before your homepage — and that summary is built from Wikipedia, YouTube, and LinkedIn, not from your About page or your brand guidelines. You can have the best-designed homepage in your category and still lose the branded AIO to an outdated LinkedIn entry from 2021. Brand control has migrated off your owned channels. The brands that win in AI search are the ones that treat third-party entity data with the same rigour they previously applied to their own site — because that is where the summary gets written.

Sources

• Ahrefs / Ryan Law — AI Overviews branded keyword trigger rate analysis, US desktop index, September 29, 2026 measurement; progression data from July–September 2026. ahrefs.com
• DemandSphere — Parallel branded AIO coverage dataset, October 1, 2026.
• Chris Long / Nectiv — 100-enterprise-brand AI Overview trigger test, September 2026; citation source breakdown (Wikipedia 71.9%, YouTube 38.6%, LinkedIn 22.8%).
• AmICited — LinkedIn citation rate by surface (Gemini app 0.0%, Google AI Mode 3.3%, Google AI Overviews branded 22.8%, ChatGPT 33%), data window June 24 – July 23, 2026, 1,905 prompts.