ICP, Segment, PMC, PMF, Beachhead:what they actually mean and when each one matters

Founders use these terms interchangeably. They shouldn’t, because each one answers a different question.

Every B2B SaaS company uses the terms ICP, segment, and beachhead. Most use them loosely. “Our ICP is mid-market SaaS companies.” “We’re expanding into a new segment.” “Healthcare is our beachhead.”

The problem isn’t the vocabulary. It’s that using these terms interchangeably creates real strategic confusion. Each concept answers a different question and operates at a different level.

This guide breaks down what each term actually means, how they relate to each other, and why getting them right helps you choose your next market.

ICP vs. Segment vs. Beachhead

If you Google “ICP vs segment vs beachhead,” you’ll find a dozen different interpretations. Some frameworks treat ICP as a sub-level within a segment. Others use ICP and target market interchangeably. Some don’t distinguish between beachhead and ICP at all.

There’s no universally agreed-upon hierarchy. What matters is that your team uses these terms consistently and that the distinctions are clear enough to make decisions with.

I used to treat the ICP as the segment you focus on. Working through this with clients changed my mind: the segment is where you play, and the ICP is a filter inside it.

What follows is how I define and use these concepts with the B2B SaaS leadership teams I work with. It’s not the only valid interpretation, but it works in practice, because each term gets one job: market, segment and target segment say where you play, the ICP says which accounts.

The Hierarchy: Market → Segment → Target Segment → ICP

1. What defines a Market? (PMC)

A market, or PMC (product market combination), is defined by two things:

• Problem domain (where the pain exists)
• Market category (where buyers evaluate alternatives).

The problem domain is the pain your buyers experience. The market category is the shelf they put you on when comparing options: the set of competitors they evaluate you against.

For example: the problem domain might be “internal information chaos.” The market category is “knowledge management platform”. When the buyer evaluates your product, they compare you to other KM platforms, not to project management tools or knowledge bases (though buyers may enter through those adjacent categories, the evaluation and references are anchored in the primary category).

The rule: any material change to the problem domain or the market category creates a new market. A new market means a new adoption curve. PMF resets completely. Your proof doesn’t transfer. Only the relationship can: solve a new problem for your existing customers and their trust comes with you.

2. What defines a Segment?

A segment is a subdivision within your market based on criteria that affect buying behavior and whole product requirements. Segments are defined by primary criteria: the dimensions where buying dynamics fundamentally change (more on this below).

Within intranet software, “upper SMB desk-based companies” and “enterprise desk-based organizations” are different segments. Different buyers, different decision processes, different competitive dynamics, different whole product needs. Same market.

3. What is a Target Segment?

Your target segment is the segment you’re actively pursuing right now. It’s a strategic focus choice, not a structural category.

The discipline: one target segment at a time per market. You might have multiple segments mapped. You actively pursue one.

4. What’s a Beachhead?

Your beachhead, a term from Geoffrey Moore’s Crossing the Chasm, is your first target segment: the segment you chose to dominate before expanding. What makes it a beachhead specifically is urgency: buyers in this segment have a compelling reason to act now. There’s a clear buying trigger. Budget is allocated or easy to justify. The pain can’t wait.

Multiple segments might fit your product. The beachhead is the one where buyers will buy this quarter, not eventually. The beachhead only changes if PMF fails.

5. What is an ICP (Ideal Customer Profile)?

Your ICP is the profile of the accounts inside your target segment that are most likely to be a strong fit. It doesn’t decide where you play. It decides which accounts there are worth your time, and which ones to disqualify.

The test: can a company belong to your target segment and still be a poor fit? If yes, the criteria that disqualify it belong in your ICP.

An ICP describes conditions, not company types. For example: the problem keeps coming back, it matters when it does, and someone is able and required to act on it. You can’t search for those conditions directly. You find them through proxies you can target (supporting criteria) and confirm them in the first conversations (qualifying criteria).

Each target segment gets its own ICP. What predicts a strong fit in upper SMB isn’t what predicts it in enterprise.

An ICP is about fit, not timing. A trigger event tells you an account is ready now. It doesn’t make it a better fit.

6. What is PMF (Product-Market Fit)?

PMF is the proven ability to win repeatably within a segment, among the accounts that match your ICP. You prove it segment by segment, not across the market as a whole. Having PMF in upper SMB says nothing about whether you have PMF in enterprise. Each segment has its own curve.

Three Types of Criteria and Why the Distinction Matters

Most ICP frameworks lump all criteria together. That’s a mistake, because different criteria serve fundamentally different purposes and confusing them leads to unfocused targeting and messaging that doesn’t resonate.

1. Primary Criteria: Define the Segment

Primary criteria determine where buying behavior fundamentally changes. Change a primary criterion and you’re in a different segment: different proof requirements, different positioning, often a different sales motion.

Primary criteria must be targetable. You need to be able to build a list, run campaigns, and tell sales who to call based on these dimensions. If you can’t filter for it externally, it can’t be primary, no matter how strongly it predicts the win.

Limit yourself to two primary dimensions. Every primary dimension multiplies your segment count and your GTM complexity. Two dimensions with three values each gives you nine segments. Add a third dimension and you’re at twenty-seven. No team of 20-80 people can build a distinct GTM motion for twenty-seven segments.

One primary dimension is ideal for early-stage companies: maximum focus, fastest reference density. Most successful beachheads are built around one dominant dimension, usually workflow or buyer role.

Two is common and workable — specific enough to win repeatably without fragmenting your GTM across too many combinations.

Three is the outer limit, and for most companies at this stage it means you’re either over-segmenting or some dimensions are actually supporting criteria masquerading as primary.

The test

The test: if you change this criterion, does the buying behavior change fundamentally? Different evaluation process? Different competitive alternatives? Different whole product requirements? If yes, it’s primary. If no, it’s supporting.

Which dimensions are primary depends on your market. There’s no universal template. In knowledge management software, the primary dimensions might be company size × industry. Because a 200-person tech company and a 200-person logistics company with mostly field workers need fundamentally different things from an intranet. In cybersecurity, it might be industry × regulatory environment. In developer tools, tech stack × team size.

Most founders have never tested this deliberately. They inherited segmentation from their CRM setup and never validated whether those dimensions actually predict different buying behavior. A bracket of “50-500 employees” might be hiding two completely different buyers that split at 200.

Supporting Criteria: Sharpen the ICP

Supporting criteria add useful context within segments but don’t require a separate GTM motion. Relaxing or changing them doesn’t put you in a new segment, it widens or narrows your focus within the same one.

Here’s the tricky part: the same dimension can be primary in one market and supporting in another. Geography is supporting for a company where Dutch and German buyers behave identically, but primary for a company where regulatory or cultural differences change the entire buying process. For most European SaaS companies, geography is primary: language, buying culture and the references buyers trust all change at the border. Industry is supporting when all your buyers have the same workflow regardless of vertical, but primary when healthcare buyers need fundamentally different things from tech buyers.

There’s no fixed list of “supporting dimensions.” The classification depends on your market. The test is the same as for primary criteria: does changing this dimension require a different GTM motion, different positioning, different proof, different sales process? If yes, it’s primary. If you can serve both with the same motion and the same proof, it’s supporting.

The rule: changing, relaxing, or adding supporting criteria widens or narrows your execution focus and changes your ICP, but does not create a new segment or reset PMF, unless it turns out the dimension actually changes buying logic or whole product requirements. If it does, you’ve discovered it was primary all along.

Common Segmentation Dimensions

Qualifying Criteria: Predict the Win and Shape the Messaging

This is the layer most ICP frameworks miss entirely.

Qualifying criteria strongly predict outcomes but cannot be targeted externally, they’re only observable through sales engagement. They determine whether a prospect within your primary segment is likely to close.

Examples

Fit (belongs in your ICP):

• Assigned owner — one person accountable for getting it live
• DIY tolerance exhausted — current state described as broken
• Failed previous attempt — tried solving it, didn’t stick
• Reference resonance — they see themselves in your proof

Timing (is now the moment):

• Trigger event — something changed that made the status quo unacceptable
• Leadership mandate — someone senior named this a priority
• Real timeline — a deadline, not “exploring options”
• Budget identified or easy to justify

Deal (can this one close):

• Champion exists — someone internal advocating for the purchase
• Decision-maker accessible — not buried behind a research committee

Only the fit criteria belong in your ICP. Timing and deal criteria change from month to month. Fit doesn’t.

Pick three to five that predict closed-won vs. closed-lost in your data. The rest are noise.

These might be the strongest predictors in your data. The difference between a greenfield buyer and a replacement buyer can mean the difference between a 40% win rate and a 5% win rate, within the exact same primary segment. But you can’t build a target list of “companies that don’t have a knowledge management software yet.” You discover it in the first conversation.

Here’s what most frameworks miss: qualifying criteria are the primary input for your messaging.

You don’t message primary criteria. Nobody writes a homepage that says “for companies with 150-400 employees that are desk-based.” You message the situation the buyer is in when they buy and that situation is described by qualifying criteria.

The homepage speaks to: information is scattered everywhere. Someone finally got told to fix it. There’s a deadline. The current approach is visibly broken. That’s leadership mandate + assigned ownership + timeline + DIY tolerance exhausted: all qualifying criteria, turned into language that makes the buyer say “this is exactly us.”

This creates a feedback loop: the qualifying patterns you discover in closed-won analysis become the language of your messaging, which attracts more buyers in the same situation, which increases the proportion of qualified deals in the pipeline.

Three Layers, Three Functions

Why It Matters: Choosing Your Next Market

Map your market this way and it does more than keep the vocabulary straight. It helps you choose your next market.

Most of what founders call their next market is really a new segment. The map tells you which one it is, and what it will cost. When you move, how much of what you’ve built carries over depends entirely on what you’re changing.

Most expansion frustration comes from underestimating which type of change you’re making.

The founder who says “we’re just expanding into professional services”, while staying in the same size band — might think they’re making a small move. If industry is a supporting criterion in their market, that’s 80-90% transfer. If it’s actually a primary criterion, because professional services firms buy fundamentally differently — it’s 40-60%. The same move, classified differently, requires completely different resourcing.

The founder who says “we’re going upmarket and entering a new geography” is changing two things simultaneously. If both are primary, that’s 10-20% transfer. If company size is primary but geography is supporting, it’s 40-60%. The classification determines the investment.

Going upmarket on its own is one primary change: a partial rebuild on paper. In practice it often feels like starting over, because bigger buyers want references from companies they see as peers.

That’s why getting the primary vs. supporting distinction right matters more than any other segmentation decision.

What’s an adjacent segment?

A segment where your current proof and whole product give you a structural advantage. Your next bowling pin. You’re not targeting it yet, but when you’re ready to expand, this is where the proof travels most easily.

The Practical Test

Here’s how to apply this to your company right now:

1. Define your PMC

What’s the problem domain? What market category do buyers put you in? If you’re not sure about the category, ask your buyers what they compared you to. That’s your category.

2. Identify your primary segmentation dimensions

Look at where buying behavior changes most. Test whether your current CRM segmentation reflects how buyers actually behave or whether it was inherited and never validated.

3. Choose your target segment

Of the segments you’ve mapped, where is the urgency highest and where can you win fastest? That’s where you focus. If you’re early, that’s your beachhead.

4. Build the ICP for that segment

Which targetable dimensions sharpen your focus without changing the buying dynamic? These are the ones you can relax when you want to grow within a segment.

5. Identify your qualifying criteria from closed-won analysis

What situational factors predicted whether a deal closed? These become both your qualification gate and your messaging foundation.

6. Map your expansion options against the PMF transfer table

For each one, identify what changes: supporting, one primary, or both. Resource accordingly.

Where I can helpMarket & Customer Intelligence