ChatGPT Prompts for ICP Research: Define Your Ideal Customer Profile
ChatGPT prompts for ideal customer profile (ICP) research and definition. Firmographic, technographic, and behavioral attributes that predict customer success.
The Prompt
Act as a go-to-market strategist who has helped 50+ B2B companies sharpen their ICP to accelerate pipeline.
Build an Ideal Customer Profile (ICP) framework for:
Company: {your company name and product}
Product description: {what you sell and the core problem it solves}
Best current customers: {describe 3-5 of your happiest, most successful customers — industry, size, role who bought, why they love it}
Worst churn cases: {describe 2-3 customers who churned or struggled — what was different about them}
Average deal size: {ARR or deal value}
Sales cycle length: {average weeks/months from first contact to close}
Output:
1. ICP definition document:
- Firmographics (company size range, industry verticals, growth stage, geographic focus)
- Technographics (tools they use that signal readiness for your product)
- Behavioral signals (what actions indicate they need you NOW — trigger events to monitor)
- Role profile (the primary buyer: title, typical priorities, what they read, what they fear)
- Negative ICP (who to NOT sell to — the profile that burns resources and churns)
2. ICP scoring model (5 criteria, each scored 1-3, max 15):
- Criteria 1-5 with scoring rubric
3. Lead qualification checklist (7 yes/no questions a rep can use in 5 minutes)
4. Go-to-market implications:
- Which channels reach this ICP most efficiently
- Which content topics resonate with this ICP's priorities
- Which competitors does this ICP evaluate alongside you
Constraints:
- ICP must be specific enough to disqualify 60%+ of the market — too broad is useless
- Negative ICP must be honest — don't just list fictional bad customers
- Scoring model criteria must be observable before a call (firmographic/technographic signals)
- Lead checklist questions must be answerable in 5 minutes of LinkedIn + website research
Variables to fill in
-
{best customers}3-5 of your happiest, most successful customers — describe them specifically -
{churn cases}2-3 customers who left or struggled — what made them different -
{product description}What you sell and the core problem it solves -
{deal data}Average deal size and sales cycle length
How to use this prompt
- Run this with your sales and CS teams to validate the ICP against both acquisition and retention data
- Use the ICP scoring model to prioritize your current pipeline
- Share the negative ICP with your SDR team to reduce time spent on bad-fit prospects
- Revisit this prompt every 6 months as your product and customer base evolves
An ICP that doesn’t disqualify is not an ICP
The purpose of an ICP is to tell you who NOT to sell to as much as who to pursue. If your ICP describes 80% of the market (‘B2B companies that could benefit from better productivity’), it provides no targeting value. The prompt’s constraint — specific enough to disqualify 60%+ of the market — forces the precision that makes ICPs actually useful.
Churn data is your most honest ICP input
Customers who churned reveal the limits of your product’s fit. If your worst churn cases were all enterprise companies while your best customers are mid-market, that’s a critical ICP signal. Most companies build ICPs from their best customers alone — and miss the negative ICP pattern that’s causing them to waste cycles on deals they’ll eventually lose.
Technographics predict buyer readiness better than firmographics
A company using Salesforce, HubSpot, and Slack is a different buyer than a company on spreadsheets — even if both are the same size and industry. The tech stack signals sophistication, readiness to integrate, and willingness to pay for software. The prompt’s technographic section helps you identify the tech signals that predict your best customers.
Related free tools
- Free AI Prompt Generator — build structured prompts instantly
- AI Token Counter — estimate API costs
- AI ROI Calculator — calculate the business value of AI adoption