What are Category Prompts?
Category prompts are unbranded prompts about a product or service category, phrased the way a buyer who has never heard of you would ask them. Examples are "best CRM for small businesses", "top payroll software for Dutch companies" or "which e-bike is best for commuting". They show which brands AI assistants recommend when the user has no brand in mind.
They are the opposite of branded prompts, which include your name ("is [brand] any good") or compare you directly with a named competitor. Branded prompts are useful for checking how you are described, but they almost always mention you, so they say little about whether new buyers discover you.
Most category prompts follow a few patterns: best-of lists ("best [category] for [audience]"), use-case questions ("tools for [task]") and problem-led questions ("how do I [solve a problem]").
Why Category Prompts matter
Category prompts reflect the moment of discovery, when buyers build their shortlist. Appearing in these answers is the AI-search equivalent of ranking for a generic keyword in classic SEO: it is where new customers find you.
They also keep your metrics honest. A prompt set dominated by branded questions inflates your mention rate and AI share of voice. If your share of voice comes close to 100 percent, your prompts are probably too branded.
How to build a set of Category Prompts
List your categories, audiences and use cases, then combine them into questions such as "best [category] for [audience]" and "[category] for [use case]".
Add problem-led questions that buyers ask before they know your category exists.
Use the words customers use in sales calls, reviews and support tickets, not your internal jargon.
Include local and language variants if you sell in several markets, such as Dutch prompts for the Netherlands.
Keep a smaller set of branded prompts separate and report the two groups apart.
Trackbase Prompt Tracking runs your category prompts daily across 11 AI models and shows how you rank against competitors on each of them. To turn the results into a report, see 7 GEO metrics that actually matter.
Examples of Category Prompts
A B2B payroll software vendor tracks "best payroll software for small businesses in the Netherlands" and "how do I automate payroll for 50 employees".
A Dutch bicycle webshop tracks "beste e-bike voor woon-werkverkeer" in Dutch alongside its English equivalent, "best e-bike for commuting".
A cybersecurity startup tracks "how to protect a small business from phishing", a problem-led prompt where AI assistants name vendors in the answer.
A hotel group tracks "quiet hotels near Amsterdam Central Station" to see which properties AI assistants suggest to travelers.
Frequently asked questions
What is the difference between category prompts and branded prompts?
Category prompts do not contain your brand name and test whether AI assistants recommend you when a buyer is still exploring options. Branded prompts include your name and show how you are described once someone already knows you. Both are useful, but they answer different questions, so track and report them as separate groups.
How many category prompts should I track?
Enough to cover your main categories, audiences and use cases; a few dozen is a practical start. Each prompt should represent a real buying question. It is better to track fewer prompts that matter commercially than hundreds of variations of the same question. Expand the set as you learn which topics drive visibility and which competitors appear.
Should category prompts be in the language of my market?
Yes, if you sell in several countries. AI assistants can recommend different brands for the same question in Dutch, German or English, partly because they draw on different sources for each language. A Dutch company selling locally should track Dutch prompts first and add English prompts for international buyers.