The news
The average brand is mentioned in only 16.3% of AI answers, AthenaHQ said in a report released on September 29, exposing an AI search brand visibility gap just as a separate Emplifi and Semrush survey found 66% of consumers have used AI to research products.
AthenaHQ, a San Francisco company that sells generative engine optimization (GEO) tools, which help brands appear in chatbot answers, said its State of AI Search 2026 report draws on millions of AI-generated responses across eight leading large language models, naming ChatGPT, Gemini, Perplexity, Claude and Grok among them. It found that a brand's own website goes uncited in 84% of AI answers. Category leaders do far better, at 56.5%, roughly 3.5 times the average.
Much of what AI tools cite comes from other sites. Reddit accounted for 21.9% of off-page citations, meaning citations of sites other than the brand's own, AthenaHQ said. The models also behave very differently: Grok drew on an average of 27 distinct domains per response, compared with five for Gemini. Chief executive Andrew Yan said "AI search is a completely different ecosystem than traditional search."
On the same day, Emplifi and Semrush, an Adobe company, released a report finding that 66% of consumers have used AI tools to research products or discover gift ideas, and 11% start product research directly with an AI tool. The consumer figures come from an online survey of 1,638 people in the US and UK aged 18 to 65 who use social media regularly, conducted with Alchemer between April 14 and May 1, 2026.
Semrush traffic data in the same report shows AI referrals to 30 major US retailers have increased 190 times over the period studied. AI referral traffic to that group of retail sites now equals an estimated 8% of its social referral traffic, up from 0.04% three years earlier, the companies said. More than one in four consumers surveyed (26%) add the word Reddit to product searches to find real customer opinions.
AI is also moving into how small businesses reach customers. On September 29, Meta Platforms (META) expanded its AI agent Muse to small businesses, TechCrunch reported. Muse links to Instagram analytics, Facebook pages and Meta ad accounts, and integrates with tools including Shopify, Stripe, Klaviyo and Intuit QuickBooks. A free version comes with usage limits, and paid plans allow more use.
The numbers
- Average brand's share of AI answers mentioning it (AthenaHQ)
- 16.3%
- AI answers that do not cite the brand's own website (AthenaHQ)
- 84%
- Category leaders' figure (AthenaHQ)
- 56.5%, about 3.5x the average
- Reddit share of off-page citations (AthenaHQ)
- 21.9%
- Surveyed consumers who have used AI to research products or gifts (Emplifi)
- 66% of 1,638
- Growth in AI referrals to 30 major US retailers (Semrush)
- 190 times
Why CEOs should care
For CEOs and CMOs, these numbers turn AI visibility into a revenue question. If two-thirds of surveyed consumers already use AI to research products, and the average brand appears in roughly one AI answer in six, most brands are missing at the moment a buyer builds a shortlist. Ask your team for a baseline: for your top 20 buying questions, how often do ChatGPT, Gemini and other assistants mention you, and how often do they mention your closest competitor?
The citation data shows where to invest. Because a brand's own site goes uncited in 84% of AI answers and Reddit supplies more than a fifth of outside citations, visibility depends heavily on what customers, reviewers and forums say, not only on your website. Marketing leaders should make product pages clear and easy to cite, with specific facts, prices and specs, and fund reviews, community and press work, while following each platform's rules on disclosure.
For CFOs, expect a wave of pitches. AthenaHQ sells GEO tools and Semrush sells search analytics, so treat their figures as directional rather than independent. Before paying for any AI visibility service, ask how it measures mentions, which models and prompts it tracks, how often results change, and how it will show a link to traffic or sales.
The bigger picture
Search is splitting in two. A results page lists many links; an AI answer narrows the choice to a few named options, and AthenaHQ's data suggests category leaders already capture a far larger share of that attention. That favors incumbents with strong reputations and raises the cost of waiting for challengers. At the same time, platforms such as Meta are putting AI agents inside small businesses' own marketing and operations, so more of the path from discovery to sale now runs through AI tools on both sides.
What’s next
Watch the fourth-quarter holiday season for fresh data on AI referral traffic to retailers, whether AI assistants add formal programs or paid placements for brands, and whether GEO vendors agree on shared, auditable ways to measure visibility.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
Companies in this story
