Track 30 to 50 real buyer questions to start: category questions ("best X for Y"), competitor questions ("X alternatives", "X vs Y"), problem questions, pricing questions and questions about your own brand. Choose them from what buyers actually ask your sales team, in their words, not from your keyword list.
The five kinds of prompts
| Type | Example | Why it matters |
|---|---|---|
| Category | "Best CRM for a 5-person agency" | Where AI picks winners |
| Competitor | "HubSpot alternatives for small teams" | Buyers already shopping |
| Problem | "How do I stop losing leads in spreadsheets" | Early research, tool fits naturally |
| Pricing | "Cheapest CRM with email sequences" | High intent |
| Brand | "What is [your brand]? Is it good?" | What AI says when buyers check you |
A good starting split is about 40% category, 25% competitor, 20% problem, 10% pricing and 5% brand.
How to find them
- Sales calls and support tickets: the questions people asked before buying.
- Reddit and communities: how people describe the problem.
- "People also ask" in Google for your category.
- Ask the AI itself: "What questions do buyers ask before choosing a [category]?"
How many is enough
- 30 to 50 gives a stable picture for one product. Seenlense generates 30 to 50 from your category and competitors, and you edit them before the first scan (Seenlense is our product).
- More prompts help once you track several segments, countries or products.
- Fewer than 15 is too noisy to read trends from.
Keep the list stable
Change prompts rarely. If you rewrite the list every week, you cannot tell whether visibility changed or the questions did.
FAQ
How many prompts should I track?
30 to 50 for one product to start. Add more for extra segments or markets.
Should I include my brand name in prompts?
Only in a few brand questions. Most prompts should be unbranded, because that is where AI decides whether to name you at all.