ICP Before TAM: Why Market Sizing Comes Second

Most go-to-market plans run in the wrong direction. A team sizes the market first, finds a number large enough to justify the effort, then reverse-engineers a customer profile to match it. The result is an ICP that describes an outcome the TAM analysis needed rather than a customer the product actually serves. "B2B SaaS companies with more than 50 employees" is not an ICP. It is a census segment. And building a go-to-market strategy on a census segment is why so many product teams end up with broad positioning, confused sales motions, and pipeline that looks healthy until the conversion data arrives.

The correct sequence runs the other way. You read the product's DNA first. The DNA dimensions tell you who the buyer is, what awareness level they enter with, what pain they are actively experiencing, and where the beachhead sits. That is your ICP. Once you have a real ICP, TAM is a calculation, not a discovery.

The Sequence Most Teams Run

The conventional approach treats market sizing as a precondition for strategy. A founder or product leader starts with total addressable market, usually via top-down analysis: industry reports, category revenue figures, some percentage-of-wallet estimate. This produces a large number. The large number gets put in a deck. Then the team works backward to define who they are selling to.

The problem is structural, not motivational. TAM analysis is not methodology-neutral. Every TAM calculation contains embedded assumptions about the customer. When you size the market for "sales intelligence software," you are already assuming something about who buys sales intelligence software, at what deal size, through what motion. Those assumptions are usually borrowed from the category, not derived from the product.

So the ICP that emerges from this process is a rationalization dressed as a derivation. It matches the market the analysis already assumed rather than the customer the product is actually built for.

This matters operationally. A rationalized ICP produces messaging that sounds right in a deck and fails in a sales call. It produces a qualification framework that flags the wrong prospects as good fits. It produces a content calendar full of topics the actual buyer does not care about because the content was built for the census segment, not the person.

What the DNA Read Actually Produces

The DNA read is a structured analysis of the product's characteristics across four dimensions. Each dimension is not a preference or a positioning choice. Each one is a constraint the product architecture has already decided, whether the team acknowledges it or not.

The four dimensions are: the buyer-user map, the complexity level, the growth motion, and the beachhead. Reading them correctly gives you the ICP before you open a single market report.

Buyer-User Map

The buyer-user map describes whether the person who purchases the product is the same person who uses it. Most B2B products sit somewhere on a spectrum between "same person buys and uses" and "executive buys, team uses, buyer never touches the interface."

Where a product sits on this spectrum determines the ICP in two ways. First, it defines which job title matters for the ICP. If the buyer is a VP of Sales and the users are individual account executives, the ICP profile must account for both, but they are not interchangeable. The VP's pain is a portfolio problem. The AE's pain is a workflow problem. The messaging, the proof points, and the entry motion are different for each. Second, the buyer-user map determines what "awareness" means in the ICP. A buyer who never uses the product has a different awareness curve than a buyer who is also the primary user. The end-user buyer finds the product through their own frustration. The executive buyer finds it through a business case someone else builds.

A product that conflates these two in its ICP ends up with a profile that is neither person. The sales team pitches the business case to someone expecting a workflow demo, or pitches the workflow to someone who needs a business case. The symptom is a long sales cycle with high late-stage drop-off: the buyer kept moving through the funnel until they realized they were looking at the wrong product for their actual decision context.

Complexity Level

Complexity determines the minimum viable sophistication of the buyer. A product that requires significant integration work, custom configuration, or organizational change to deliver value is not a product for a buyer who lacks the operational infrastructure to implement it. The ICP must specify the floor, not just the ceiling.

Teams routinely write ICPs that describe the ideal customer at their best: fully staffed, technically capable, organizationally aligned. The real ICP includes the minimum conditions under which the product can deliver value. That minimum is what qualifies a prospect. The ideal is what makes them a great customer after they are in.

Complexity also tells you something about sales motion. High complexity products almost always require a champion who understands the implementation path, which means the ICP must include a technical or operational buyer, not just an economic one. If the ICP does not name this person, the sales process will discover them halfway through a deal and stall.

Growth Motion

Growth motion tells you how the product wants to spread. Product-led growth assumes individual users adopt first and create organizational pressure upward. Sales-led growth assumes executive sponsorship precedes user adoption. Hybrid motions exist but they require different ICP specifications for each leg.

A product with PLG motion has an ICP anchored to the individual user. The buyer is often the user. The entry point is a free tier or a trial. The pain is personal and immediate. The awareness is high because the user found the product themselves.

A product with sales-led motion has an ICP anchored to the economic buyer. The entry is a discovery call. The pain is organizational. The awareness may be lower because the buyer was not actively searching for this category before the outbound touch.

Teams that mix up the growth motion in the ICP end up with sales processes misaligned to the product's actual adoption pattern. A PLG product with a sales-led ICP creates friction at the exact moment the product should be selling itself. A sales-led product with a PLG ICP sends salespeople after individual contributors who cannot buy.

Beachhead Definition

The beachhead is the narrowest coherent segment where the product can win completely before expanding. It is neither the total market nor the eventual market. It is the segment where the product has the strongest fit, the clearest differentiation, and the shortest path to a referenceable win.

The beachhead constrains the ICP vertically, by industry or use case, before horizontal expansion becomes relevant. A product built for compliance-heavy environments in financial services has a beachhead ICP that includes regulatory context as a qualifier. Remove that qualifier and the ICP expands to include companies where the product's primary differentiator is irrelevant.

Most founders resist a narrow beachhead definition because it feels like leaving money on the table. The correct framing is that a broad beachhead ICP is not money on the table. It is diluted positioning that makes the table smaller for everyone.

Why "B2B SaaS Companies" Is Not an ICP

A segment description is not a profile. "B2B SaaS companies with 50 to 500 employees" tells you the industry and a headcount band. It tells you nothing about the buyer's job title, their current pain state, their awareness of the problem category, the internal stakeholders involved in a purchasing decision, or the operational context that makes them ready to buy now rather than in six months.

An ICP built at this level of abstraction produces an addressable universe that is technically correct and practically useless. Every B2B SaaS company with 50-500 employees is in the target. But most of them are not ready to buy. Some of them do not have the problem. Some have the problem but lack the budget authority at the job title level you are targeting. Some have the problem, the budget, and the authority, but the internal priority queue puts the purchase in the next fiscal year.

None of these distinctions are visible at the segment level. They only become visible when the ICP specifies the conditions of readiness, not just the conditions of membership.

A functional ICP specifies the job title, the pain that is acute right now, the event or trigger that surfaced the pain, the awareness level of the solution category, the decision-making structure, and the operational minimum required to get value from the product. "Director of Revenue Operations at a Series B SaaS company that recently missed a quarterly forecast and is evaluating their pipeline visibility" is an ICP. "B2B SaaS companies" is a filter.

Deriving the ICP from Each DNA Dimension

Each DNA dimension produces a specific ICP constraint, and the four constraints together produce the full profile.

The buyer-user map gives you the job title hierarchy. It tells you who signs, who champions internally, and who uses the product day-to-day. Each of these roles has a different pain statement and a different message requirement. The ICP must name all three if they are distinct, and must specify which role is the primary entry point for the sales motion.

Complexity gives you the qualification floor. It translates into two or three minimum conditions a prospect must meet before they can get value: team size in the relevant function, technical infrastructure already in place, or organizational maturity on a specific process. Below this floor, prospects are not bad fits. They are not ready yet. The ICP should name the floor explicitly so the qualification process can apply it consistently.

Growth motion gives you the awareness tier. A PLG product's ICP includes buyers who are actively aware of the problem and actively searching for solutions, because those are the people who find the product through the free tier. A sales-led product's ICP can include buyers with lower awareness, because the outbound motion creates the conversation. The awareness tier in the ICP determines where content investment goes, how much education the sales cycle requires, and what the close rate benchmark should realistically be.

Beachhead gives you the vertical and use-case constraint. It narrows the ICP to the segment where the product wins most cleanly, and it provides a sequencing rationale for expansion: once the beachhead is saturated, which adjacent segment shares enough characteristics to apply the same motion with minimal adjustment.

Stack these four outputs and you have an ICP that is derived from the product rather than borrowed from the category. It is specific enough to generate a qualified prospect list and a coherent sales process.

Building TAM from the ICP

Once the ICP is defined at this level of specificity, TAM becomes a calculation. You count the buyers who fit the profile, estimate the deal size, and multiply. The number may be smaller than a top-down estimate from an industry report. That is the point.

A TAM built from a real ICP has three properties that a top-down TAM does not. First, it is reachable. Every company in the universe fits the profile, so in principle every company in the universe can be targeted, qualified, and sold to through the defined motion. There is no hidden non-buyer population inflating the number. Second, it is defensible. You can explain the methodology in one paragraph without relying on percentage-of-industry assumptions. The number comes from counting buyers, not from shaving a category total. Third, it generates useful product questions. If the ICP-derived TAM is smaller than the ambition requires, that is strategic information. It tells you which constraint is binding: Is the beachhead too narrow? Is the complexity floor too high? Is the buyer-user map limiting the addressable motion? Each constraint can be examined and relaxed deliberately, with a clear sense of what trade-off comes with the relaxation.

The top-down TAM hides all of this. It gives you a number without a mechanism, and a mechanism without a number is the thing you actually need to build the go-to-market plan.

Sequencing the Work in Practice

The sequence is: DNA read first, ICP derivation second, TAM calculation third.

The DNA read takes a few hours for a product team that already knows their product well. It surfaces faster if there is existing customer data to reference. Patterns in the current customer base reveal which DNA dimension is doing the most work in the current ICP, even if that ICP was never formally derived.

The ICP derivation takes longer because it requires the team to commit to specificity they have been avoiding. The resistance is usually strongest around the complexity floor and the beachhead constraint. Both feel like restrictions. Both are actually clarifications that make every downstream decision cheaper.

The TAM calculation is the last step and the fastest one. With a well-specified ICP, it is an afternoon of work, not a research project. The research project happened in the ICP derivation phase, which is where it should have been all along.

Teams that run this sequence produce positioning that survives the first real sales conversation. Teams that run the reverse sequence produce positioning that survives the first investor meeting and fails the second customer call.