Most ICP definitions I see are useless. They read something like: "Mid-market SaaS companies with 50 to 500 employees in North America." That describes thousands of companies and tells your BDR nothing about who to prioritize or what to say.
A real ICP is a targeting weapon. It should tell you exactly which companies to go after, which people to contact, what to say to them, and when to reach out. Here is the framework I use to build ICPs that actually drive outbound results.
The Four-Layer ICP Model
I define ICPs across four layers, each one adding more precision to your targeting.
Layer 1: Firmographics. This is the baseline: company size (revenue and headcount), industry or vertical, geography, funding stage, and business model. Most teams stop here. That is the mistake.
Layer 2: Technographics. What tools does the company use? What tools are they missing? A company using Salesforce but not using a sales engagement platform is a different prospect than one already running Salesloft. Technographic data tells you about their maturity, their budget, and their likely pain points.
Layer 3: Trigger Events. Static data tells you who could buy. Trigger events tell you who is likely to buy right now. I track: leadership changes (new VP Sales, new CRO), funding rounds (Series A through C), hiring signals (posting for BDRs or sales ops), product launches, geographic expansion, and competitive switches.
Layer 4: Pain Signals. This is the most nuanced layer. What observable evidence suggests this company has the problem you solve? Job postings for roles that indicate gaps, negative Glassdoor reviews about sales processes, public comments from leaders about growth challenges, or declining market share.
Building Your Scoring Model
Once you have defined your four layers, assign weights. Not every signal is equally predictive. In my experience, trigger events and pain signals are 3 to 5 times more predictive than firmographics alone.
Create three tiers. Tier 1 accounts match on all four layers, meaning they are the right size, use complementary technology, have a recent trigger event, and show pain signals. These get fully personalized outreach. Tier 2 accounts match on two to three layers. These get semi-personalized sequences with specific hooks. Tier 3 accounts match on firmographics only. These get automated but targeted campaigns.
The distribution should look something like 10 to 15% Tier 1, 30% Tier 2, and the rest Tier 3. If your Tier 1 list has 500 accounts, you are probably not being selective enough.
From ICP to Messaging
Your ICP should directly inform your messaging. Each layer maps to a messaging element.
Firmographics inform context: "As a Series B SaaS company scaling past 100 employees..." Technographics inform relevance: "I noticed you are using Salesforce but do not have a dedicated sales engagement tool..." Trigger events inform timing: "Congrats on the recent funding round. Most teams at this stage..." Pain signals inform empathy: "I have seen teams in your position struggle with..."
The more layers you can reference in a single message, the more relevant and personalized it feels. But do not force it. One or two specific references is enough to differentiate you from the generic "I help companies like yours" emails flooding every inbox.
Validating Your ICP
Your ICP is a hypothesis until data proves it right. Run your outbound campaigns for 30 to 60 days and then analyze the results by ICP tier and layer.
Look at reply rates by tier. If Tier 1 is not significantly outperforming Tier 3, your scoring model needs work. Look at which trigger events and pain signals correlate with positive responses. Double down on what works and cut what does not.
Your ICP should evolve every quarter. Markets shift, your product changes, and new competitors emerge. The teams that treat ICP as a living document outperform the ones that set it and forget it.
Common ICP Mistakes
Defining ICP by who you want to sell to instead of who actually buys. Look at your closed-won deals and reverse-engineer the profile. You might be surprised.
Making your ICP too broad because you are afraid of missing opportunities. A narrow ICP with 80% hit rate will always outperform a broad ICP with 5% hit rate. You can expand later once you have a working motion.
Ignoring negative ICP criteria. Knowing who not to target is just as valuable. If companies under 20 employees never close, exclude them from your list and save the effort for accounts that convert.