Lead-Scoring Model Builder · Marketing Tool Stack

Most lead scoring models are a pile of guesses nobody revisits. This prompt makes your AI build one that is grounded in your ICP and easy to maintain.

What this isA reusable prompt that turns your ICP and funnel into a documented fit-plus-intent scoring model.

You give itYour ICP, lifecycle stages, and the fields you can score on.

You get backFit and intent point tables, an MQL threshold, plus decay and negative-scoring rules.

Paste this into Claude, ChatGPT, or any capable model. Fill in the bracketed parts and it does the rest.

You are a marketing operations analyst. Help me build a lead scoring model for [COMPANY], a [B2B SaaS / your industry] company.

Our ICP: [ideal customer: company size, industry, buyer role].
Our funnel stages: [list lifecycle stages, e.g. Subscriber, Lead, MQL, SQL, Opportunity].
Data we can score on: [available fields: job title, company size, industry, page views, content downloads, email engagement, demo requests].

Build a two-part model:
1. Fit score (0 to 50): firmographic and demographic match to the ICP. List each attribute, the points, and the reasoning.
2. Intent score (0 to 50): behavior and engagement. List each signal, the points, and a decay rule for stale activity.

Then:
- Define the combined threshold that makes a lead an MQL, and explain why.
- Flag any signal that looks like noise and should not be scored.
- Suggest one negative-scoring rule (for example student or competitor emails).
- Keep it maintainable. No more than 8 fit attributes and 8 intent signals.

Ask me for anything you need before you start.

When to use it

When you are setting up scoring for the first time, or auditing a model that has drifted and nobody trusts anymore.

How to use it

Fill the bracketed parts with your real ICP and fields, paste it in, and answer the follow-up questions. Then pressure-test the output against your last 20 closed-won deals before you build it in your MAP.