> ## Documentation Index
> Fetch the complete documentation index at: https://docs.caard.net/llms.txt
> Use this file to discover all available pages before exploring further.

# ICP profiles and scoring

> Score every lead against your ideal customer profile.

An ICP (ideal customer profile) describes who your best customers are. Caard scores every lead against your ICP profiles so you know who to follow up with first.

## Reading a score

Each lead gets an overall fit rating:

| Fit          | Default threshold |
| ------------ | ----------------- |
| Strong fit   | 80+               |
| Good fit     | 60+               |
| Moderate fit | 40+               |
| Weak fit     | 20+               |
| No fit       | below 20          |

The score breaks down across four dimensions — **Company**, **Prospect**, **Geo**, and **Behavioral** — with an AI-written rationale explaining the result and flagging missing data. View it on any lead's detail page, or browse all scores at **Leads → ICP → Scoring**.

<Tip>
  Filter the lead list by ICP fit after an event to work the strongest leads first.
</Tip>

## Building a profile

Create profiles at **Leads → ICP → Builder**. A profile includes:

* **Criteria** — weighted conditions, marked must-have or nice-to-have
* **Disqualifiers** — conditions that automatically rule a lead out
* **Dimension weights** — how much Company, Prospect, Geo, and Behavioral each count (they sum to 100%)
* **Thresholds** — the score cut-offs for each fit level

You can run a **Core** profile alongside **Experiment** profiles that test a hypothesis with a decision date and success threshold.

## The ICP guide

Not sure where to start? **Leads → ICP → Guide** walks you through it:

1. **Learn** — short guidance on what makes a strong ICP: anchor on lifetime value, use verifiable signals, keep must-haves few.
2. **Your company** — describe what you sell, to whom, and your pricing.
3. **ICP document** — upload or paste any existing ICP, positioning, or sales playbook.
4. **AI draft** — Caard generates a complete profile from your inputs, ready to review and save.

## Rescoring

When you change a profile, existing scores go stale. A banner offers to **rescore** all leads — progress is shown while it runs, and you can trigger a rescore manually at any time.

## ICP analytics

**Leads → ICP → Analytics** shows fit distribution, criteria hit rates, score trends over time, source breakdowns, and how ICP fit correlates with pipeline stages — useful for checking whether your profile actually predicts wins.