Veuno finds AI assistants favor a narrow set of brands in category answers

Veuno's study of more than 5,400 AI answers found that AI assistants accurately identify brands by name nearly every time, but leave many companies out when buyers ask for the best option in a category. The gap is especially sharp for B2B companies, which appeared in just 35% of category answers versus 51% for consumer brands.

Why it matters: - The study suggests brand awareness does not translate into AI-era recommendation visibility. - Buyers using AI assistants for category research may see a small, repeated shortlist instead of a broad market view. - B2B companies appear to be at a bigger disadvantage than consumer brands in AI-generated answers.

What happened: - Veuno released a study on Oct. 20, 2026, examining how AI assistants answer buyer questions about companies and product categories. - The research tracked 100 prompts across ChatGPT, Claude, Microsoft Copilot, Google AI and Perplexity from Sept. 30 to Oct. 13, 2026. - The study produced more than 5,400 answers. - When users asked about a company by name, AI assistants accurately described the company in 99% of answers. - When users asked category questions, AI assistants named B2B companies in 35% of answers and consumer brands in 51%.

The details: - The median B2B company appeared in 18% of category answers. - The median consumer brand appeared in 56% of category answers. - Six of the 10 B2B companies appeared in fewer than one in four category answers. - Three of the 10 B2B companies appeared in fewer than one in 10 category answers. - Among consumer brands, four of 10 appeared in fewer than one in four category answers. - One of 10 consumer brands appeared in fewer than one in 10 category answers. - On 10 of the 80 category questions, the company did not appear in a single answer across 14 days and five engines. - On 26 of the 80 category questions, the company appeared in fewer than one answer in 10. - The study included 10 B2B companies in sectors including healthcare data, financial infrastructure, accounting AI, developer tools, employee benefits and construction data. - The study also included 10 consumer companies in personal finance, consumer health, travel, children’s education, food and home products. - Veuno selected venture-backed or privately funded U.S. businesses with 100 or fewer employees and estimated annual revenue between $1 million and $15 million. - None of the companies was a Veuno client. - Each company received one branded prompt and four unbranded category prompts. - A company counted as present when its name appeared in the answer text. - Perplexity results were based on a smaller number of completed runs.

Between the lines: - Veuno says AI assistants appear to rely on a short list of sources and companies for each category. - Selina Wragg, Veuno's founder and CEO, said being known and being recommended are different outcomes. - Wragg said AI assistants repeat what they read in comparison articles, review sites, analyst lists and community threads. - Wragg said companies that are absent from those sources can remain absent from AI answers, even if the brand is well known. - The pattern points to a visibility problem rooted in third-party coverage, not just brand familiarity. - Veuno's initial 40-prompt audits over the same two weeks showed an even lower starting point for many companies. - In those audits, B2B companies appeared in 1% to 2% of category answers and were absent from about two-thirds of category questions. - A consumer services brand in those audits appeared in 16% of category answers and was absent from about half. - Wragg said most B2B budgets still go to company websites and ads, while AI assistants are reading elsewhere.

What's next: - Veuno's findings suggest companies may need to widen their visibility efforts beyond owned channels if they want to appear in AI-generated category recommendations. - The study implies future marketing measurement may need to track not just brand recall, but whether AI assistants recommend a company in category searches.

The bottom line: - AI assistants can recognize a brand, but that does not mean they will recommend it when buyers ask for the best option.