---
name: clarafold-icp-research
description: "Runs Clarafold's ICP Research Workflow: Researches a target segment, identifies decision-makers and structures account intelligence into a reusable profile. Use when the user wants to define or refine an ideal customer profile (ICP), research a target segment, map a buying committee, or decide which accounts not to target."
---

# ICP Research Workflow

The ICP Research Workflow is a seven-step process for turning closed-won, closed-lost and stalled deal history into a documented ideal customer profile: firmographic and behavioural criteria, a buying committee map by role, a buyer problem statement in the buyer's own language, and explicit exclusion criteria. It takes roughly 2–4 hours per segment.

## When to use

Use this skill when the user wants to define or refine an ideal customer profile (ICP), research a target segment, map a buying committee, or decide which accounts not to target.

## Ask the user for these inputs first

- Closed-won and closed-lost history
- Current customer list
- Target segment hypothesis

If an input is missing, say what is missing and what the output will lack without it. Do not fill gaps with invented data.

## Tools this workflow assumes

- CRM export
- Enrichment provider
- LinkedIn
- AI research assistant
- Spreadsheet

Use whatever the user has available (uploaded exports, connected tools, web search). Say which tools you actually used.

## Steps

1. **Pull the evidence.** Export closed-won, closed-lost and stalled deals. The ICP is a pattern in data you already have, not an opinion.
2. **Isolate the firmographics.** Size, sector, geography, tech stack and business model of accounts that closed well and renewed.
3. **Find the behavioural markers.** What was happening at the account when it bought. Triggers matter more than attributes.
4. **Map the buying committee.** Who initiated, who evaluated, who signed, who blocked — by role, not by name.
5. **Write the problem statement.** Articulate the problem in the buyer's language, drawn from call notes and lost-deal reasons.
6. **Build the exclusion list.** Define who is a poor fit and why. Knowing who to skip is half the value.
7. **Structure the output.** Produce a profile a researcher, a copywriter and a rep can each use without asking follow-up questions.

Work through the steps in order. Show the result of each step briefly before moving on, so the user can correct course.

## Output

Produce one document with a section for each of these:

- Documented ICP with firmographic and behavioural criteria
- Buying committee map by role
- Buyer problem statement in their language
- Explicit exclusion criteria

## Rules

- Use only evidence the user provides or that you can cite. Label assumptions as assumptions.
- Never invent metrics, company names, contacts or quotes.
- Prefer specific, checkable statements over general advice.

---

Published by Clarafold (https://clarafold.com/workflows/icp-research). Free to use and adapt.