Grade every lead the same way before a human calls

Most teams qualify leads by gut feel. Whoever reads the inquiry first decides whether it deserves a call, and two people will make two different calls on the same lead. Good-fit prospects get slow replies while bad-fit ones eat up calendar slots.

This pack fixes the first step. You define five weighted criteria: ICP fit, problem match, authority, budget signal, and urgency. The prompt scores each new inquiry against them, justifies every score by quoting the lead's own words, flags anything unknown, and hands your team a total, a recommended action, and two questions to ask in the first five minutes of the call. The human still decides. The scorecard just guarantees every lead got the same first look.

The recipe walks you through four steps. First, fill in the scorecard: five criteria, weighted to total 100, plus score bands that map to actions like book a call, nurture, or decline. Second, paste the prompt into ChatGPT, Claude, or Gemini and replace the bracketed sections with your details. Third, feed it a lead: the full form fill, email, or LinkedIn message.

The fourth step matters most. Before you trust the grades, run the prompt on ten past leads whose outcomes you already know. If your real best customers don't score near the top, your weights are wrong, so adjust and re-run. The recipe also tells you how to catch the model inventing facts about a lead, which is the failure mode that quietly ruins scoring systems.

Copy-paste scoring holds up at a handful of leads a week. It breaks when leads arrive faster than someone remembers to paste them, when scores need to land in your CRM on their own, or when the grade should trigger the follow-up itself: a booking link for high scores, a nurture sequence for the middle band. That is a built system with intake, scoring, routing, and logging, and building those is what we do. If your leads are already stacking up, book a free AI opportunity audit and we'll map exactly what that system looks like for your pipeline.