Frank Vitetta Research Finds ChatGPT Is Seven Times Less Likely to Name Brands in Long Queries Than Short Ones

New AI search study analyzing 8,555 ChatGPT answers to buying questions shows that prompt length significantly affects brand visibility in AI responses

LONDON, UNITED KINGDOM, October 19, 2026 /EINPresswire.com/ -- New research from Frank Vitetta, an AI consultant based in London, has found that the length of a search query has a significant effect on whether ChatGPT names a brand in its response. The study, based on 8,555 ChatGPT answers to shopping and buying questions collected by LLM Scout over nine months, found that short questions produced brand mentions in nearly half of all answers, while long questions produced brand mentions in fewer than one in fifteen.

Short questions of approximately one line — the everyday "best X for Y" format — produced brand mentions in 46.7% of answers. Questions of 120 characters or more produced brand mentions in only 6.5% of answers, making a long query approximately seven times less likely to result in a brand being named.

A Step-by-Step Drop in Brand Visibility

The research identified a clear progression. Questions between 35 and 69 characters named a brand in 44% to 47% of answers — the highest-performing range and representing just over half of all answers studied. From 70 characters onward, the rate began to decline: 35% for questions between 70 and 89 characters, and 28% for questions between 90 and 119 characters. At 120 characters or more, only 75 of 1,147 answers included a brand name.

The pattern reflects a difference in how ChatGPT responds to different types of questions. Short "best X" questions tend to produce lists of brand names. Longer questions that describe a scenario and ask for advice tend to produce advice rather than specific brand recommendations.

Implications for AI Visibility Measurement

The findings carry a practical implication for brands that track their AI visibility scores — metrics that measure how frequently a brand appears in AI-generated answers. According to Vitetta, those scores can shift based on the mix of question lengths used in testing, without any change in the AI assistant's underlying behavior.

"Nothing about the assistant has changed. Only the mix has," Vitetta writes in the study.

"If you build test prompt sets to measure how assistants answer, the length mix of the set moves the result," he adds.

Removing the longest questions from the dataset caused the overall brand mention rate to rise from 36% to 40.5% — a meaningful shift driven entirely by the composition of the test set rather than by any change in ChatGPT.

Why the Research Matters for Brands

The study is relevant to the growing field of generative engine optimisation (GEO), AI optimisation (AIO) and answer engine optimisation (AEO) — disciplines focused on helping brands appear in AI-generated responses as consumers increasingly turn to ChatGPT, Gemini, and similar tools for purchasing decisions rather than traditional search engines.

Vitetta recommends that brands request their AI visibility scores broken down by short and long questions, test the longer and more detailed queries that real customers are likely to type into AI assistants, and maintain consistent question sets across reporting periods to ensure comparable results over time.

The full study, including methodology, data charts, and limitations, is available at https://frankvitetta.net/field-notes/prompt-length-brand-mentions-chatgpt/

About Frank Vitetta

Frank Vitetta is an AI consultant based in London who helps teams prepare their data for AI applications. He writes about how AI assistants such as ChatGPT find and name brands. The research data was collected using LLM Scout (https://llmscout.co). More information is available at www.frankvitetta.net.

Media Relations
Frank Vitetta