Glossary
What is Buying prompt?
A buying prompt is a real purchase-intent question a customer asks an AI assistant — e.g. 'best CRM for small agencies' or 'cheapest GEO tool in Spanish'. GEO works by knowing which buying prompts matter in your category and ensuring AI answers name you in them.
Why buying prompts matter
Buyers describe needs to assistants as questions with constraints, such as their team size, country or must-have feature. Those prompts are the unit of GEO measurement. Unlike search keywords, they come with no reliable volume data, so the prompts you choose to track decide which brands look like your competitors and whether your results mean anything.
How to build a set of buying prompts
Start from evidence: sales calls, support tickets, review sites, search suggestions and the questions customers ask before buying. Add the persona and constraint the way a buyer would, for example 'for a ten-person agency' or 'in Mexico'. Cover the whole journey: category discovery ('best X for Y'), comparisons ('A vs B'), alternatives ('alternatives to A') and problem-led questions ('how do I do Z').
Keep most prompts unbranded, meaning they do not contain your brand name, because those show whether engines bring you up unprompted. Write them in the language your buyers actually use, since answers differ by language.
A hypothetical example
Imagine an invoicing app for freelancers in Chile. Useful prompts might be 'best invoicing software for freelancers in Chile', 'how do I send electronic invoices as a freelancer' and 'alternatives to' a well-known competitor. Each represents a different stage: discovery, education and switching. A prompt like 'is [our brand] good?' belongs in a separate branded group, because it names you by construction and would inflate the results.
Common mistakes
- Tracking only prompts that include your brand name, which flatters the results.
- Using generic prompts such as 'best software', which no real buyer would type and which name whoever is biggest.
- Writing prompts like keyword strings instead of natural questions.
- Adding many near-duplicates, which inflates the count without adding information.
Common questions
How many buying prompts should I track?
Enough to cover your main intents, but quality beats quantity. A few dozen well-chosen prompts are more useful than hundreds of near-duplicates, while too few make the results noisy.
Are buying prompts the same as keywords?
No. Prompts are longer, conversational and carry context, and there is no dependable search-volume figure for them. Treat them as representative samples of buyer intent, not as demand estimates.
See if AI assistants recommend your brand
Run a free scan to see which AI engines mention your brand and who they cite instead of you.
Scan my site free