Every B2B company has an Ideal Customer Profile. Most of them are wrong.
Not wrong in the sense that they describe imaginary customers. Wrong in the sense that they're too broad, too static, and too disconnected from the actual decisions they should inform. They're documents created once during a strategy offsite, referenced in pitch decks, and ignored in practice.
Your ICP should be the single most important filter in your go-to-market operation. Every marketing pound, every sales call, every product decision should pass through it. If it doesn't, you don't have an ICP, you have a wish list.
The Common Mistakes
Mistake 1: Too Broad
"We sell to B2B SaaS companies with 50-500 employees."
That's not an ICP. That's a market. An ICP should be specific enough that your team can look at any prospect and immediately say "yes, this is us" or "no, this isn't." If the answer is "maybe," your ICP isn't doing its job.
A useful ICP adds layers of specificity:
- Industry vertical: Not just "SaaS" but "vertical SaaS in healthcare, fintech, or logistics"
- Growth stage: Not just "50-500 employees" but "Series A-C companies that have raised $5M-$50M and are growing headcount 20%+ YoY"
- Organisational maturity: "Has at least one dedicated marketing hire but no marketing ops function"
- Technology indicators: "Uses HubSpot or Salesforce, has implemented at least one marketing automation tool"
- Pain signals: "Currently running outbound manually, experiencing declining response rates"
Mistake 2: Based on Aspiration, Not Evidence
Your ICP should be derived from your best existing customers, not from the customers you wish you had. This seems obvious, but it's staggeringly common to see ICPs that describe a market segment the company has never successfully sold into.
Run this analysis:
- List your top 20 customers by revenue, retention, and satisfaction (NPS or equivalent)
- Find the common attributes across these customers
- That cluster of attributes is your ICP
If your best customers are $5-20M revenue professional services firms, but your ICP says "enterprise technology companies with 1,000+ employees," you have a disconnect. Your ICP is aspirational, not evidential.
Mistake 3: No Negative Definition
An ICP should explicitly state who you don't sell to. This is as important as defining who you do sell to, because it prevents your team from chasing bad-fit deals that will churn, drain support resources, and distort your product roadmap.
Negative ICP criteria might include:
- Companies below a revenue threshold (they can't afford you and will churn)
- Industries with regulatory complexity you can't support
- Companies with a tech stack that doesn't integrate with your product
- Geographies you can't service effectively
- Companies in active M&A (decision-making freezes during acquisition)
Mistake 4: Static Document
Markets change. Your product evolves. Your competitive landscape shifts. An ICP created 18 months ago may not reflect who your best customers are today.
ICPs should be reviewed quarterly and updated based on:
- Win/loss analysis (who are you winning and losing, and why?)
- Churn analysis (which customers leave, and what do they have in common?)
- Expansion analysis (which customers grow with you, and why?)
- Market shifts (new segments emerging, existing segments contracting)
Building an ICP That Works
Step 1: Analyse Your Best Customers
Pull data on your top-performing customers. "Top-performing" means:
- High revenue: They pay you well
- Low churn risk: They renew reliably
- High NPS: They'd recommend you
- Low support burden: They don't drain your team
- Expansion potential: They buy more over time
Find the 15-20 customers that score highest across all five dimensions. These are your ICP exemplars.
Step 2: Identify the Clusters
Look for patterns across your exemplars:
- Firmographic: Industry, size, geography, growth rate
- Technographic: Tech stack, digital maturity, tools in use
- Behavioural: How they found you, how long the sales cycle was, what content they consumed
- Organisational: Team structure, decision-making process, budget ownership
You'll likely find 2-3 distinct clusters. These become your ICP segments, not one monolithic profile, but a small set of specific, actionable profiles.
Step 3: Validate With Win/Loss Data
Cross-reference your ICP clusters against your win/loss data:
- Do prospects matching your ICP win at a higher rate than non-ICP prospects?
- Are ICP deals faster to close?
- Do ICP customers retain better?
If the answer to all three is yes, your ICP is valid. If not, adjust until it is.
Step 4: Operationalise It
An ICP is only useful if it changes behaviour. Here's how to operationalise it:
In marketing: Only run campaigns targeting ICP accounts. Score leads based on ICP fit. Gate access to high-cost resources (demos, trials, consultation) behind ICP qualification.
In sales: Build ICP fit into your lead scoring model. Train reps to qualify against ICP criteria in discovery. Require manager approval for non-ICP deals above a certain discount threshold.
In product: Prioritise feature requests from ICP customers. When ICP and non-ICP customers want different things, choose ICP every time.
In customer success: Allocate CSM resources proportionally to ICP fit. Your best CSMs should manage your best-fit customers.
Step 5: Build the Feedback Loop
Every quarter, run the ICP review:
- Are we winning more ICP deals than last quarter?
- Are ICP customers retaining better?
- Has anything changed in the market that affects our ICP?
- Should we add, remove, or modify any ICP segments?
Document changes and communicate them to every customer-facing team.
The Courage to Say No
The hardest part of a strong ICP is the discipline to say no. No to the prospect who doesn't fit but has budget. No to the RFP from a big-name company in the wrong industry. No to the channel partner who wants to bring you into deals outside your sweet spot.
Every non-ICP deal you close is a future churn risk, a support burden, and a distraction from the customers you serve best. The short-term revenue feels good. The long-term cost is real.
The best companies have the courage to walk away from bad-fit revenue. It's counterintuitive, but saying no to the wrong customers is how you say yes to growth.
The Bottom Line
Your ICP isn't a slide in your investor deck. It's the operating system for your entire GTM motion. If it's broad, static, and disconnected from evidence, it's not helping you, it's just making you feel like you have a strategy.
Do the work. Analyse your best customers. Build specific, evidence-based profiles. Operationalise them across every team. And have the discipline to live by them, even when it means walking away from revenue.
That's how you build a GTM motion that compounds. And compounding is how you win.