Why one ICP outperforms multiple segments for B2B software companies

"The concept of having multiple ICPs is sort of an oxymoron," Rob said on Youroast Ep1. "It's kind of like it can't be true. Like you can't have multiple things that are the ideal, right? It's like you have one true love - you can't have multiple true loves." [Ep1, 4:10]

The logic: multiple ICPs is a contradiction in terms. The ideal customer is singular by definition. You can have secondary customers - profiles that share enough characteristics that the same message reaches them. But the ideal - the customer for whom your product is the most specific, most obvious, most valuable answer - can only be one profile.

Why it matters practically: a claim specific enough to resonate deeply with buyer A will feel slightly wrong to buyer B. A claim broad enough to reach both A and B will resonate shallowly with both. Shallow resonance produces:

• Longer evaluation cycles
• Higher churn (buyers who were adjacent to the ICP, not the ICP)
• A sales team that cannot qualify quickly because the ICP criteria are too broad to filter against
• Positioning that gets squishier every quarter as more adjacent buyers are accommodated

Rob is not absolutist. He describes what he calls a parabolic view: broad experimentation is legitimate very early, narrow focus is required once signal exists, expansion is legitimate again at scale.

Very early stage (pre-signal):

"When you're really early and you don't really know who your ideal customer is, there's an argument to be made where you should actually be pretty broad. Let's try multiple. Let's throw a bunch of stuff against the wall. You're really looking for signal." [Ep1, 5:00]

The critical qualification on mindset: "Your mindset should not be, hey, I'm trying to maximize the effectiveness. Your mindset should be, I'm trying to figure out who the ideal one is." [Ep1, 5:15]

Broad experimentation is a discovery mode. It is not a GTM strategy.

Once signal exists:

"From like five hundred K to even like ten million, I think you should be really at least eighty percent focused, if not one hundred percent focused on that one ICP in where you're putting your marketing and sales spend." [Ep1, 5:40]

The signal that ends the experimental phase: you can identify which customer type converts fastest, stays longest, and refers most consistently. When that is clear, the experimental phase is over.

The most common version of the ICP mistake is not companies that never found their ICP. It is companies that found a working ICP, saw adjacent opportunities, and expanded before genuinely dominating the core.

"We just see companies prematurely scaling beyond their core where there's so much more room for growth to truly dominate in an ICP segment space," Rob said. "But while they're doing that, they see all these silver edges of opportunity at the fringes and they start chasing them way too early." [Ep1, 6:00]

The cost of premature expansion is not just missed revenue in the new segments. It is degraded positioning in the core segment:

• The hero claim gets vague enough to speak to everyone
• Sales starts pitching different buyer types with different objections
• The playbook that worked for the original ICP stops working because it is being adapted in real time
• The differentiation gets, in Rob's words, "all squishy"

"You're far better off, in our opinion, to have a dominant ownership over that one ICP. Like everyone truly in that space should know who you are, should understand your product. You should be owning ten, twenty percent, if not more of that market share before you think about truly expanding." [Ep1, 6:20]

Rob is clear about the conditions under which pursuing multiple segments genuinely makes sense: resources to play multiple games simultaneously without degrading execution in any of them.

"The only time where really multiple segment ICP comes into play is when you have the resources to play multiple games. These are companies, in our opinion, that are doing well over ten, twenty million. Often what we see, the ones who come to us, they're doing much more than that - they're over a hundred million in revenue." [Ep1, 7:00]

At that scale, companies have:

• Budget to maintain distinct positioning per segment
• Team to run distinct messaging and sales motions
• Organizational structure to keep segments from diluting each other

Startups have none of those three. Running multi-segment GTM with a lean roster and a low budget is, in Rob's framing, "kidding themselves."

Rob's practical advice for founders who know they should narrow but find it uncomfortable:

"Be uncomfortably narrow and just truly own that niche. There's no reason why you can't build a million dollar business in a very micro niche. That was our thesis when we started Fletch. Just think one ideal customer, one avatar, one channel, one tactic - really focus." [Ep1, 29:00]

The discomfort itself is a signal. When a positioning statement starts to feel too specific - when adjacent segments think "that's cool but not quite for me" - you are probably close to the right level of specificity.

Most founders stop one step before this point because the narrowness is uncomfortable. The market they excluded feels like lost revenue. It is actually the thing that makes the included buyer trust the claim.

The test: adjacent segments should think "cool, but not quite for me." If every adjacent segment thinks "this might be for me," you are not specific enough.

A website designed for multiple ICPs is a website that converts none of them well.

Every section that tries to speak to two buyer types simultaneously makes a compromise. The hero claim gets vaguer. The proof blocks get more generic. The CTA becomes less urgent because it is not addressed to a specific person with a specific problem.

A website designed for one ICP - the buyer your product solves the most specific problem for - earns more demos from that buyer. Adjacent segments who have similar problems self-select in without being explicitly targeted. But the primary ICP sees themselves first and feels understood in a way generic messaging never achieves.

Signs your website is designed for multiple ICPs:

• The hero has multiple claims stacked together
• "Built for X, Y, and Z teams" appears in the first section
• The proof blocks reference multiple entirely different use cases with no dominant thread
• Your team describes the ICP differently from each other

Read website redesign prerequisites for the team survey that surfaces ICP alignment gaps before any design sprint begins.

Run the team survey. Ask everyone who will give feedback during your next design sprint or launch: what does the product do, who is it for, what problem does it solve? Separately, no discussion first. Compare the ICP descriptions. If three people describe three different buyers, the ICP has not been decided.

Read the companion articles:

• Fix positioning before you redesign - the full positioning framework before a sprint
• Three litmus tests to check your B2B software positioning - the specific tests including the team survey above
• The Fletch positioning framework for B2B software companies - how Fletch maps existing categories versus new categories

Watch the full episode: Youroast Ep1 with Rob Kaminski.