The right prompt gives you a deal name, an owner, and a dollar amount. The wrong one gives you a high-level framework about pipeline health. The difference isn’t the AI model—it’s the way you ask. It’s 7:47am on Monday. Your pipeline review kicks off at 8. You have thirteen minutes to figure out which deals are at risk, which reps are off pace, and whether you’re actually on track to hit this quarter’s target. You type into your AI tool: “How’s my pipeline looking?” What you get back is a 400-word think piece on the drivers of pipeline health—interesting, but useless in the moment. You close the tab, open HubSpot, and start drilling into deals one by one, exactly what you were trying to avoid. The issue isn’t the AI; it’s the prompt. And once you fix the prompt, a bigger issue appears: the AI is still guessing at your actual pipeline. It’s working off an outdated CSV, without your metric definitions, without your historical context. It doesn’t know what “committed” means in your sales process or why your Stage 3 doesn’t match anyone else’s Stage 3. This article shares eleven prompts that actually work, pulled from the Databox Prompt Library. They’re designed for HubSpot CRM and Pipedrive. Copy, paste, and use them in your next pipeline review. TL;DR: Vague AI prompts yield vague answers. The remedy is specificity: a clearly named object, a defined time frame, and a concrete decision trigger. The eleven prompts below come from the Databox Prompt Library. They are the exact questions…