Overview
Identifying your Ideal Customer Profile (ICP) is crucial for refining your marketing and sales strategies. However, your approach to defining ICPs should fundamentally differ based on whether you've reached product-market fit.
This guide covers strategic principles for defining ICPs at different stages, then shows how to implement them in Unstuck Engine using structured parameters.
Two paths to ICP definition
Path 1: Pre-product-market fit (hypothesis-driven)
If you have not reached product-market fit, you lack win rate and net dollar retention data to validate your ICP.
Your ICP is a hypothesis based on:
Initial market research
Competitor analysis
Strategic assumptions
Target customer interviews
Founder intuition
Key principle: Treat every aspect of your ICP as unproven until you have closed deals and retention data. Be prepared to pivot quickly as you learn what actually works.
In Unstuck Engine: Start with broad Score mode parameters. Test multiple ICP hypotheses simultaneously. Monitor which ICP generates meetings, but remember—meetings don't validate an ICP. Only win rates and retention prove fit.
Path 2: Post-product-market fit (data-driven)
If you have reached product-market fit, you have win rate and net dollar retention data that proves which customer segments succeed with your product.
Your ICP is proven by:
Historical win rates by segment
Net dollar retention (NDR) by customer cohort
Customer lifetime value (LTV)
Time to value
Product usage patterns
Churn rates by segment
Key principle: Extract your ICP from actual customer data, not assumptions. Your best customers reveal your real ICP through their behavior and retention.
In Unstuck Engine: Analyze your existing customers in your CRM. Identify common characteristics of high-NDR, high-LTV customers. Configure ICP parameters to match these proven patterns using Score mode with weights proportional to correlation with success metrics.
How to approach ICP definition based on PMF status
Pre-PMF: Start broad, test hypotheses
Create 2-3 ICP hypotheses based on research
Use minimal Filter parameters (1-2 maximum)
Use Score mode for most criteria to maintain flexibility
Set Score weights conservatively (10-15 range)
Activate Signals and generate prospects for each ICP
Track meetings, demos, and proposals—but don't lock in your ICP yet
Wait for 10+ closed deals and 3-6 months of retention data
Rebuild your ICP based on actual winning patterns, not your original hypothesis
Critical mistake: Don't lock in your ICP based on which segments book the most meetings. Meetings ≠ product-market fit. Only wins + retention = validated ICP.
Post-PMF: Extract from proven data
Export your customer list from CRM
Calculate win rate, NDR, time to close, LTV, and CAC by customer segment
Identify patterns in top-performing cohorts (industries, sizes, roles, technologies, etc.)
Configure ICP parameters to match proven patterns:
Filter mode only for 90%+ correlation to success
Score weight 20-30 for 70%+ correlation
Score weight 10-20 for 50%+ correlation
Validate that new high-scoring prospects convert at expected rates
Re-calibrate quarterly based on latest data
Four foundational principles
1. Treat it as a hypothesis (until proven otherwise)
Your ICP remains a hypothesis until validated by win rates and net dollar retention data.
Validation criteria:
Minimum 10 closed deals matching ICP
Minimum 6 months retention data
Win rate significantly higher than non-ICP deals
NDR above your target threshold (typically 100%+)
In Unstuck Engine: Monitor Records filtered by ICP Score ranges. Track which score bands actually convert and retain. Adjust weights until ICP Score correlates with real business outcomes.
2. Focus on differentiation
Your ICP should emphasize characteristics that differentiate successful customers from unsuccessful ones, not just any possible attribute.
Pre-PMF: Differentiate based on market assumptions and research
Post-PMF: Differentiate based on actual win rate and NDR data
In Unstuck Engine: Only add parameters that meaningfully separate high-performers from low-performers.
3. Maximize specificity (but validate it)
Aim for high specificity, but ensure each specific criterion is validated by data.
Pre-PMF: Start specific based on hypothesis, but be ready to broaden if you're not getting traction
Post-PMF: Be as specific as your data supports—if 80% of high-NDR customers have characteristic X, weight it heavily
In Unstuck Engine: Layer multiple Score mode parameters. High-scoring prospects match many specific criteria.
4. Avoid unnecessary conditions
Over-filtering excludes potential good-fit prospects.
Pre-PMF: Use extreme caution with Filter mode—you don't yet know what's truly necessary
Post-PMF: Use Filter mode only for characteristics that have 90%+ correlation with success or failure
In Unstuck Engine: Start with Score mode for most parameters. Only promote to Filter mode after proving the requirement is non-negotiable.
Validating your ICP
Pre-PMF validation milestones:
10 closed deals – Earliest you can start seeing patterns
20 closed deals – Can identify preliminary trends
50 closed deals – Can build statistical confidence
6 months retention data – First meaningful retention signal
12 months retention data – Strong retention validation
Don't declare PMF or "proven ICP" before these milestones.
Post-PMF validation metrics:
Track these metrics by ICP Score band (e.g., 0-20, 21-40, 41-60, 61-80, 81-100):
Win rate by ICP Score: Target: 2-3x higher win rate for scores 80+ vs. scores 0-40
NDR by ICP Score: Target: NDR 120%+ for scores 80+, under 100% for scores 0-40
Time to close by ICP Score: Target: 20-40% faster for scores 80+ vs. scores 0-40
CAC by ICP Score: Target: 30-50% lower CAC for scores 80+
If these correlations don't exist, your ICP is still a hypothesis—keep iterating.
Implementation in Unstuck Engine
Step 1: Identify non-negotiable requirements (Filter mode)
List 2-3 absolute requirements that prospects MUST meet.
Pre-PMF: Use sparingly—you may be wrong about what's truly required
Post-PMF: Only use if 90%+ of successful customers match this criterion
Examples:
"Must be B2B"
"Must have 50+ employees"
"Must be Director+ level"
Step 2: Identify strong preferences (Score mode, weight 15-25)
List characteristics that indicate better fit but aren't dealbreakers.
Pre-PMF: Focus here—most parameters should be strong preferences
Post-PMF: Weight based on correlation with win rate and NDR (70%+ correlation)
Examples:
"Prefer US-based companies"
"Prefer SaaS companies"
"Prefer recent funding"
Step 3: Identify nice-to-haves (Score mode, weight 5-15)
List characteristics that slightly improve fit.
Pre-PMF: Test these but don't over-weight
Post-PMF: Weight based on correlation (50-70% correlation)
Examples:
"Nice if they use specific technologies"
"Bonus if they're in certain industries"
Step 4: Identify exclusions (Exclude mode)
List characteristics that disqualify prospects.
Pre-PMF: Use very sparingly—you might be excluding future best customers
Post-PMF: Use for characteristics with 90%+ correlation with churn or failed deals
Examples:
"Never target competitors"
"Exclude certain countries"
Example ICP and Persona configurations
Example 1: Mid-Market SaaS Sales Leaders
Target: Sales leaders at mid-market B2B SaaS companies with recent funding
ICP configuration:
Target Market: B2B (Filter mode)
Industry: "SaaS, software, technology" (Score mode, weight 20)
Headcount Range: 51-200, 201-500 (Score mode, weight 15)
GTM Motion: PLG, Hybrid, SLG (Score mode, weight 20)
Latest Funding Round Type: Series A, Series B, Series C (Score mode, weight 25)
Latest Funding Round Year: 2022-2024 (Score mode, weight 20)
Country of Headquarter: "United States, United Kingdom, Canada" (Score mode, weight 10)
Persona configuration:
Seniority Level: Director, VP, CXO (Filter mode)
Function: Sales, Executive (Filter mode)
Job Title: "VP Sales, CRO, Chief Revenue Officer, Sales Director" (Score mode, weight 20)
Headline: "revenue, sales, growth, ARR" (Score mode, weight 15)
Total Years of Experience: 8-70 range (Score mode, weight 10)
Experience Keywords: "scaled, grew, built, revenue" (Score mode, weight 15)
Example 2: Cybersecurity MSPs
Target: Managed service providers in cybersecurity using Microsoft security stack
ICP configuration:
Industry: "managed service provider, MSP, cybersecurity, security" (Score mode, weight 25)
Headcount Range: 11-50, 51-200 (Score mode, weight 15)
Technologies Used: "Microsoft Defender, Azure Security Center, Microsoft 365" (Score mode, weight 25)
Latest Funding Round Type: Seed, Series A (Score mode, weight 15)
Country of Headquarter: "United States, Canada" (Score mode, weight 10)
Persona configuration:
Seniority Level: Manager, Director, VP, CXO (Filter mode)
Function: Executive, Sales, Business Development (Filter mode)
Job Title: "CEO, Founder, VP Sales, Director Business Development" (Score mode, weight 20)
Headline: "MSP, managed services, cybersecurity, IT services" (Score mode, weight 20)
Total Years of Experience: 5-70 range (Score mode, weight 10)
Example 3: Enterprise Product Leaders at PLG Companies
Target: Senior product leaders at product-led growth B2B SaaS companies
ICP configuration:
Target Market: B2B (Filter mode)
Industry: "SaaS, software" (Score mode, weight 20)
Headcount Range: 201-500, 501-1000 (Score mode, weight 20)
GTM Motion: PLG, Hybrid (Score mode, weight 30)
Signup Mechanism: Free Trial, Freemium (Score mode, weight 25)
Pricing Model: Subscription (Score mode, weight 10)
Founded Year: 2015-2022 (Score mode, weight 15)
Persona configuration:
Seniority Level: Director, VP, CXO (Filter mode)
Function: Product, Executive (Filter mode)
Job Title: "VP Product, CPO, Chief Product Officer, Director Product" (Score mode, weight 25)
Headline: "product, activation, onboarding, growth, PLG" (Score mode, weight 25)
Total Years of Experience: 10-70 range (Score mode, weight 10)
Experience Keywords: "product-led, activation, onboarding, PLG" (Score mode, weight 20)
Skills: "Product Management, Product Strategy, Product-Led Growth" (Score mode, weight 15)
Example 4: Robotics Engineers with Startup Experience
Target: Senior engineers in robotics/IoT who have founded or joined early-stage startups
ICP configuration:
Industry: "robotics, automation, IoT, hardware, manufacturing" (Score mode, weight 25)
Headcount Range: 11-50, 51-200 (Score mode, weight 15)
Latest Funding Round Type: Seed, Series A (Score mode, weight 20)
Founded Year: 2018-2024 (Score mode, weight 15)
Persona configuration:
Function: Engineering, Product (Filter mode)
Seniority Level: Senior, Manager, Director (Filter mode)
Job Title: "Robotics Engineer, Automation Engineer, IoT Engineer" (Score mode, weight 25)
Headline: "robotics, automation, IoT, embedded systems, hardware" (Score mode, weight 25)
Total Years of Experience: 8-70 range (Score mode, weight 15)
Experience Keywords: "founded, co-founder, startup, entrepreneurship" (Score mode, weight 25)
Skills: "Robotics, Automation, IoT, Embedded Systems, Machine Learning" (Score mode, weight 20)
Example 5: Healthcare IT Decision Makers
Target: IT and compliance leaders at healthcare organizations needing HIPAA solutions
ICP configuration:
Industry: "healthcare, hospital, medical, clinic, health system" (Score mode, weight 25)
Headcount Range: 201-500, 501-1000, 1001-5000 (Score mode, weight 20)
Certifications & Compliance: "HIPAA, SOC 2, HITRUST" (Score mode, weight 25)
Founded Year: 2000-2020 (Score mode, weight 10)
Persona configuration:
Seniority Level: Director, VP, CXO (Filter mode)
Function: IT, Executive (Filter mode)
Job Title: "CIO, CISO, VP IT, IT Director, Chief Information Officer" (Score mode, weight 25)
Headline: "healthcare IT, HIPAA, compliance, health technology, medical IT" (Score mode, weight 25)
Total Years of Experience: 10-70 range (Score mode, weight 15)
Skills: "Healthcare IT, HIPAA Compliance, IT Security, Health Information Systems" (Score mode, weight 20)
Example 6: Early-Stage Fintech Founders
Target: Technical founders at early-stage fintech companies
ICP configuration:
Industry: "fintech, financial services, payments, banking, lending" (Score mode, weight 25)
Headcount Range: 11-50, 51-200 (Score mode, weight 15)
Latest Funding Round Type: Seed, Series A (Score mode, weight 25)
Latest Funding Round Year: 2023-2024 (Score mode, weight 20)
Founded Year: 2020-2024 (Score mode, weight 20)
Persona configuration:
Seniority Level: Owner, CXO (Filter mode)
Function: Executive, Engineering, Product (Filter mode)
Job Title: "CEO, CTO, Founder, Co-Founder" (Score mode, weight 30)
Headline: "founder, fintech, payments, financial technology" (Score mode, weight 25)
Total Years of Experience: 5-70 range (Score mode, weight 10)
Experience Keywords: "founded, built, launched, startup" (Score mode, weight 25)
Skills: "Fintech, Financial Services, Entrepreneurship, Product Development" (Score mode, weight 15)
Example 7: Marketing Leaders at High-Growth Companies
Target: Senior marketing executives at fast-growing B2B companies
ICP configuration:
Target Market: B2B (Filter mode)
Headcount Range: 51-200, 201-500 (Score mode, weight 20)
Latest Funding Round Type: Series A, Series B, Series C (Score mode, weight 25)
Latest Funding Round Year: 2022-2024 (Score mode, weight 25)
Monthly Website Visits: 10,000-500,000 (Score mode, weight 20)
GTM Motion: PLG, Hybrid, SLG (Score mode, weight 15)
Persona configuration:
Seniority Level: VP, CXO (Filter mode)
Function: Marketing, Executive (Filter mode)
Job Title: "CMO, VP Marketing, Chief Marketing Officer" (Score mode, weight 25)
Headline: "demand generation, marketing, growth, B2B marketing" (Score mode, weight 20)
Total Years of Experience: 10-70 range (Score mode, weight 10)
Experience Keywords: "scaled, grew, demand generation, pipeline" (Score mode, weight 20)
Skills: "Demand Generation, Marketing Strategy, B2B Marketing, Growth Marketing" (Score mode, weight 15)
Monitoring and iteration
Track these metrics by ICP Score band:
Pre-PMF metrics:
Meeting booking rate by ICP
Demo completion rate by ICP
Proposal rate by ICP
Win rate by ICP (need 10+ deals minimum)
Early retention signals (30-90 days)
Post-PMF metrics:
Win rate by ICP Score bands (target: 2-3x higher for scores 80+ vs. 0-40)
NDR by ICP Score bands (target: 120%+ for scores 80+)
Time to close by ICP Score (target: 20-40% faster for scores 80+)
CAC by ICP Score (target: 30-50% lower for scores 80+)
When to iterate:
Re-calibrate your ICP when:
Win rates don't correlate with ICP Score
You're excluding too many good prospects
You're including too many poor fits
Market conditions shift significantly
Pre-PMF: Update immediately when you learn from closed deals
Post-PMF: Re-calibrate quarterly based on latest data
FAQs
How do I know if I've reached product-market fit?
You've reached PMF when you have 20+ paying customers, NDR consistently above 100%, predictable win rates, and customers achieving measurable value.
How many parameters should I use?
Start with 5-7 parameters. Most effective ICPs have 5-10 parameters total. More than 15 often indicates over-engineering.
Should I create one ICP or multiple?
Pre-PMF: Start with 2-3 ICP hypotheses to test different segments
Post-PMF: Create multiple ICPs when your data shows distinct successful segments
Most customers use 3-5 ICPs.
How often should I update my ICP?
Pre-PMF: Update immediately when you learn from closed deals or churn
Post-PMF: Re-calibrate quarterly based on latest win rate and NDR data
What's more important: ICP Score or Engagement Stage?
Both matter:
ICP Score tells you WHO fits (likelihood to succeed long-term)
Engagement Stage tells you WHEN they're ready (likelihood to buy now)
Prioritize high ICP Score + high Engagement Stage (A/B) for best results.
Should I weight parameters equally or differently?
Pre-PMF: Start with relatively equal weights (10-15)
Post-PMF: Weight proportional to correlation with win rate and NDR:
70-90% correlation → weight 25-30
50-70% correlation → weight 15-20
30-50% correlation → weight 5-15
What if my high-scoring prospects aren't converting?
Your ICP parameters aren't predictive of actual success. Re-analyze your closed deals and adjust parameters based on what actually correlates with wins and retention.