Query length dictates whether ChatGPT mentions a specific brand in shopping answers, according to research published by London-based AI consultant Frank Vitetta, einnews.com reported. Short queries around one line long yield brand recommendations nearly half the time, whereas queries stretching past 120 characters produce brand mentions in fewer than seven percent of responses.
Prompt Length Changes Brand Visibility Rates
The study analyzed 8,555 ChatGPT answers to shopping and buying questions collected by LLM Scout over a nine-month period. Short, everyday queries formatted as “best X for Y” produced brand mentions in 46.7 percent of answers. In contrast, queries of 120 characters or more resulted in brand names appearing in only 6.5 percent of responses, making a long prompt roughly seven times less likely to name a brand.
Brand visibility dropped in a steady progression as prompt length increased. Questions spanning 35 to 69 characters captured brand mentions in 44 percent to 47 percent of answers, representing the highest-performing range and accounting for just over half of all evaluated prompts. Once query length crossed 70 characters, mention rates fell to 35 percent for prompts between 70 and 89 characters, and dropped to 28 percent for those between 90 and 119 characters. Out of 1,147 answers generated by prompts of 120 characters or more, only 75 included a brand name.
ChatGPT Responds to Detailed Scenarios with Advice
The disparity stems from how the AI assistant processes different query structures. Short “best X” prompts typically trigger lists of commercial brand names, while longer queries describing specific scenarios prompt the model to deliver general advice rather than direct product recommendations.
These mechanics affect generative engine optimization (GEO), AI optimization (AIO), and answer engine optimisation (AEO). Brands increasingly track their AI visibility scores to measure how often assistants name them during consumer purchasing research.
Query Mix Skews AI Visibility Scores
According to Vitetta, tracking metrics can fluctuate based on the specific mix of prompt lengths used in testing, independent of any underlying changes to the AI assistant’s behavior. “Nothing about the assistant has changed. Only the mix has,” Vitetta wrote in the study. When researchers removed the longest questions from the dataset, the overall brand mention rate climbed from 36 percent to 40.5 percent.
To maintain accurate tracking over time, Vitetta recommends that brands request visibility scores broken down by short and long questions. He also advises companies to test detailed queries that mirror actual consumer habits and to keep test prompt sets consistent across reporting periods.
Frequently Asked Questions
What data collection tool did the study use?
The research gathered data using LLM Scout over a nine-month period, compiling 8,555 total ChatGPT answers to shopping and buying questions.
Who conducted the research on prompt length?
Frank Vitetta, an AI consultant based in London who helps organizations prepare data for AI applications, authored the study.
How much less likely is a brand to be named in a long prompt?
Queries of 120 characters or more are approximately seven times less likely to result in a brand name compared to short, one-line searches.
Where can the full research methodology be accessed?
The complete study, including data charts and limitations, is available through Frank Vitetta.