Creating Prompts - How to create effective monitoring prompts

To create effective monitoring prompts in Thirdeye, follow these principles:

1. Use Natural Language

   • Write prompts the way a real customer would ask.

   • ✅ Good: “What’s the best CRM for small businesses?”

   • ❌ Poor: “CRM software small business recommendation query”

2. Be Specific with Context

   • Narrow the prompt to industry, audience, or feature focus.

   • ✅ Good: “Best project management tool for marketing agencies”

   • ❌ Poor: “Best project management tool”

3. Match User Intent

   • Align prompts with the actual goals or problems customers have.

   • ✅ Good: “Looking for an alternative to Salesforce that’s more affordable”

   • ❌ Poor: “Why is [Your Brand] the best CRM?”

4. Avoid Generic or Unrealistic Prompts

   • Skip vague or exaggerated questions that AI wouldn’t realistically answer.

   • ✅ Good: “Which email marketing tool works best for e-commerce stores?”

   • ❌ Poor: “What’s the absolute best software in the universe?”

5. Stay Neutral, Not Promotional

   • Don’t force the brand name into the prompt unnaturally.

   • ✅ Good: “Compare Slack vs Microsoft Teams for remote team collaboration”

   • ❌ Poor: “Why should everyone use [Your Brand]?”

6. Cover Different Prompt Types

   • Recommendation prompts: Best tools for a use case.

   • Comparison prompts: Brand vs competitor analysis.

   • Use case prompts: Industry or problem-specific.

   • Alternative prompts: Competitor replacement scenarios.

By applying these best practices, you ensure prompts return realistic, competitive, and actionable insights about how AI platforms position your brand.