ICP Definition Framework

Step 1: Analyze Your Closed-Won Data

Start with your CRM. Export every closed-won deal from the past 12 to 24 months. For each deal, capture: company name, industry, employee count, revenue, deal size, sales cycle length, win rate, expansion revenue, and churn status. You want at least 50 closed-won accounts to identify meaningful patterns. If you have fewer, supplement with your highest-engaged prospects and pipeline deals.

Segment your closed-won deals into tiers. Tier 1 accounts are your best customers — highest ACV, fastest close, lowest churn, most expansion. Tier 2 are solid customers. Tier 3 are customers you wish you had not closed — high support burden, long sales cycle, price sensitive, likely to churn. Your ICP should describe Tier 1 precisely and exclude Tier 3 explicitly.

This tiering exercise often reveals uncomfortable truths. You may discover that the industry you have been targeting hardest actually produces your worst customers, while a segment you ignored contains your highest-value accounts. Let the data lead, not your assumptions.

Step 2: Enrich with Firmographic Data

Firmographic data describes the structural characteristics of a company: industry, sub-industry, employee count, revenue, funding stage, location, growth rate, and organizational structure. Use enrichment tools like Apollo, ZoomInfo, or Clearbit to append firmographic data to your closed-won list.

Look for clusters in the data. Do your best customers share a specific revenue range? Are they concentrated in certain sub-industries? Do they tend to be post-Series A but pre-Series C? Are they growing at a specific rate? The goal is to identify the firmographic attributes that correlate most strongly with being a Tier 1 customer.

Go beyond basic demographics. Firmographic signals like recent funding rounds, headcount growth rate, new office openings, and hiring patterns for specific roles (e.g., "just hired a VP of Sales" or "growing their SDR team") are powerful ICP indicators. A company that just raised a Series B and is hiring 5 SDRs is in a fundamentally different buying mode than a company with flat headcount.

Step 3: Layer in Technographic Signals

Technographic data reveals what tools and technologies a company uses. This is one of the most underutilized ICP dimensions and one of the most powerful. If your product integrates with Salesforce, knowing which prospects use Salesforce (versus HubSpot or Pipedrive) is critical targeting intelligence.

Use tools like BuiltWith, Wappalyzer, or HG Insights to identify the tech stacks of your best customers. Look for patterns: do your Tier 1 customers disproportionately use a certain CRM, marketing automation platform, or sales engagement tool? Do they run a certain e-commerce platform, analytics suite, or cloud infrastructure?

Technographic signals also reveal sophistication level. A company using basic tools may not be ready for your product. A company with an advanced, integrated stack may indicate a tech-forward buyer who values automation and is more likely to adopt your solution. Build technographic inclusion and exclusion criteria into your ICP definition.

Step 4: Incorporate Intent Data

Intent data captures buying signals — indicators that a company is actively researching solutions in your category. First-party intent data comes from your own properties: website visits, content downloads, webinar attendance, pricing page views. Third-party intent data comes from platforms like Bombora, G2, or TrustRadius that aggregate research behavior across the web.

Intent signals dramatically improve ICP precision. A company that matches your firmographic and technographic criteria AND is actively researching solutions in your space is exponentially more likely to convert than a company that matches your demographic profile but has no active buying intent. Intent data turns your ICP from a static profile into a dynamic, timing-aware targeting engine.

Combine intent sources for maximum signal strength. A prospect who visited your pricing page, was seen researching competitors on G2, and recently posted a job listing for a role that would use your product is showing intent from three independent sources. That convergence of signals makes them a top-priority prospect.

Step 5: Build Your ICP Scoring Model

With firmographic, technographic, and intent data mapped, build a scoring model that assigns numerical values to each ICP attribute. Weight the attributes based on their correlation with Tier 1 outcomes. For example, if being in a specific sub-industry correlates most strongly with high ACV and low churn, that attribute gets the highest weight.

A simple scoring model might look like: Industry match (0-20 points), revenue range match (0-15 points), tech stack fit (0-15 points), growth rate (0-10 points), funding stage (0-10 points), intent signals (0-20 points), persona match (0-10 points). Total score out of 100. Accounts scoring 80+ are Tier 1 targets. Accounts scoring 60-79 are Tier 2. Below 60, do not pursue.

Operationalize this scoring model in your CRM and outbound tools. Use enrichment APIs to automatically score new leads and accounts as they enter your pipeline. This turns ICP from a slide deck concept into a working system that prioritizes your team's time and effort on the highest-value opportunities.

Step 6: Validate, Iterate, and Refine

Your ICP is a hypothesis until proven by results. After deploying your ICP scoring model, track how each ICP tier performs across your pipeline. Measure conversion rates at every stage: reply rate, meeting booked rate, opportunity creation rate, win rate, ACV, and time to close. If Tier 1 accounts are not outperforming Tier 2 by a meaningful margin, your model needs adjustment.

Revisit your ICP quarterly. Markets shift, your product evolves, and your competitive landscape changes. What defined your ideal customer six months ago may not be accurate today. Each quarter, re-run the closed-won analysis, update your enrichment data, and recalibrate your scoring weights. This continuous refinement is what separates companies with high-converting outbound from companies that spray and pray.

Related Resources

ICP definition is the first step in building a GTM engine. Explore these related guides to see how your ICP feeds into downstream systems:

• What does a GTM Engineer do? — Learn how GTM engineers operationalize ICPs into working systems.
• GTM Engineering Framework — See where ICP definition fits in the broader GTM engineering methodology.
• GTM Engineer Tools — Explore the enrichment and data tools used for ICP analysis.
• Pricing — See what it costs to have GTM11 build your ICP framework.