Census Segments: your target market list counts companies, it doesn't find buyers

Ask ten B2B founders to name their target segments and nine of them will read you a filter. Healthcare. Two hundred to a thousand employees. US-based. Running Salesforce. That's four dropdowns in a prospecting tool, and their two closest competitors pulled the same four dropdowns this morning. Everybody's looking at the same list and calling it strategy.

You define target market segments for a B2B company by grouping buyers around the problem they're trying to escape, not the attributes they happen to share. A real segment carries three things a firmographic filter can't see: a trigger that put the problem on the CEO's desk this quarter, an old way they're currently defending, and a villain they'd name out loud. Same industry, different trigger, different segment.

That distinction is the whole game, and almost nobody runs it. Most companies segment their market the way a census taker works a neighborhood. They count what exists. Counting isn't segmenting.

What's actually wrong with the way most B2B companies segment?

They build census segments. A census segment is a group defined entirely by observable attributes: industry, headcount, revenue band, geography, funding stage, tech stack. Every one of those facts is true. Not one of them tells you what to say.

A census tells you who exists. A segment is supposed to tell you who moves. Those are different jobs, and firmographics only do the first one.

Here's why this feels productive while it's failing. Census segments are easy to defend in a board meeting. They come with numbers attached. Total addressable market, account counts, penetration rates. Nobody argues with a filter. Meanwhile the message written for that segment has to speak to a healthcare company that just lost a payer contract and a healthcare company that just raised a Series C, and those two people have nothing in common except a NAICS code.

April Dunford put the underlying idea plainly: "Positioning is context setting for products." A segment is the context. If your segments are census rows, you have no context to set, and you end up writing one flat message and spraying it at everyone. That's usually the real reason your marketing keeps attracting the wrong customers.

Why does firmographic segmentation fail harder now than it did five years ago?

Two things changed, and they compound.

The first is that the data got free. Firmographic targeting used to be an edge because assembling the list was hard work. Now every competitor in your category buys the same enrichment, runs the same filters, and lands in the same inbox on the same Tuesday. When everyone can build the identical list, the list stops being a strategy and becomes table stakes. The differentiation has to move upstream, into how you cut the market in the first place.

The second is that AI became the first analyst in the buying process. Buyers now open ChatGPT or Claude and describe a situation, not a category. They type things like "our billing team keeps growing but collections aren't improving, who fixes that." That's a trigger, an old way, and a villain in one sentence. It is a perfect description of a real segment.

Now ask what the engine has to match that against. If your site describes your market as verticals and employee bands, there's nothing on the page that resembles the question. You're organized by census and the buyer is asking by situation, so the retrieval simply doesn't happen. This is the same failure mode underneath narrative identity work generally: brand is the new backlink, and in AI search a clear, consistent brand narrative is what gets a company cited, the way backlinks once drove search rankings. A vertical list isn't a narrative. It's a taxonomy.

“A census tells you who exists. A segment tells you who moves. Most B2B companies have spent three years perfecting a headcount.”

How do you tell a real segment from a census segment?

Run three questions against each segment you currently claim. Answer them out loud, in a sentence, without hedging. If a segment can't survive all three, it isn't a segment. It's a filter with a name on it.

1. 1What happened in their world that made this urgent in the last ninety days? This is the trigger. A contract renegotiated, a regulation landing, a competitor repositioning, a funding round with new board expectations, a leader hired or fired. No trigger means no timing, and no timing means your pipeline is guesswork.
2. 2What are they doing instead of buying from you, right now, and who built it? This is the old way, and it's almost never nothing. It's a spreadsheet the CFO depends on, two people they hired last year, an incumbent tool nobody loves but everyone knows. You can't write a new way until you can describe the old way with respect.
3. 3Who or what would they blame if you asked them at a bar? This is the villain. If nobody in the segment can name an enemy, the pain isn't sharp enough to fund a purchase this year.

Those three answers are the same three anchors the Magnetic Messaging Framework runs on: villain framing, an old-way / new-way contrast, and a promised-land outcome. That's not a coincidence. Segmentation and messaging are the same act performed at different altitudes, which is why building a positioning strategy always stalls when the segments underneath it are census rows.

Here's what the difference looks like on paper, using a healthcare revenue-cycle example.

Same companies. Same filter. Completely different message, and only one of those two columns produces a sentence a CFO would repeat to a colleague.

What pattern shows up across 200+ B2B companies?

Across more than two hundred messaging and homepage audits at PitchKitchen, the segmentation failure shows up the same way almost every time. The company claims six to nine segments. The deck has an industry slide where the only thing that changes between versions is the logo strip and one adjective. Marketing built those segments from the CRM. Nobody built them from a conversation.

Then you re-cut the same customer base by trigger, and the count collapses. Six segments usually become two or three. One of those is quietly producing the majority of closed-won revenue, and it was never on the slide because it cuts across industries. It was invisible to a filter, so it stayed invisible to the strategy.

There's a second pattern worth stealing, and it costs nothing. The sales team already knows the real segments. Reps carry an informal segmentation in their heads that never made it into a document, because nobody asked. Ask two of your best reps which deals feel easy and which feel like pushing rope, then ask what was true about the easy ones. You'll get triggers back, not firmographics. That informal map is usually closer to the truth than anything in the CRM, and it's the same reason your messaging stops working when you move upmarket: the trigger changed and the filter didn't notice.

What does re-segmenting actually look like in practice?

Take a healthtech company in the $5M-$75M range, the kind PitchKitchen works with every week. Five segments on the slide, cut by care setting: hospitals, ambulatory, post-acute, payers, digital health. Reasonable. Defensible. Producing a pipeline where nothing moved and every deal ended in "not now."

We ran the three questions across their last twenty closed-won deals. The care-setting cut turned out to be noise. What actually predicted a deal was a single trigger appearing in eleven of the twenty: a new CFO or VP of Finance arriving in the previous two quarters and being handed a margin number they hadn't set. That person shows up in hospitals and in post-acute and in payers. The census cut had scattered them across five slides so thoroughly that nobody had ever seen them as one group.

The re-segmentation didn't change who they sold to. It changed what they said and when they showed up. One segment defined by a trigger, one villain everyone in it recognized, one old way to name with respect. The homepage stopped listing care settings and started describing a situation. That's the move, and it's why running a competitive positioning analysis before you've fixed your segments tends to produce a very tidy answer to the wrong question.

What does this mean for your next planning cycle?

Don't start by adding segments. Start by auditing the ones you have against the three questions. Most companies find they have fewer real segments than they thought and one they didn't know existed.

Two guardrails before you cut. First, a segment you can't reach is a fantasy, so every trigger-based segment still needs an observable proxy you can actually target on, which is where firmographics finally earn their keep. They're the delivery mechanism, not the definition. Second, fewer segments is usually the right answer, but that isn't the same as niching down, and the two decisions get confused constantly. If you're weighing the narrower move, we've written about what niching down actually costs.

The test I'd run this week is smaller than a planning offsite. Pull your last twenty closed-won deals, write the trigger next to each one in a sentence, and count how many distinct triggers you get. If the answer is two or three, you just found your real segments. If the answer is twenty, you don't have a segmentation problem, you have a message that isn't doing any qualifying work at all, and your homepage is where that shows up first. The Brand Signal Score, PitchKitchen's free homepage messaging diagnostic at pitchkitchen.com/brand-signal-score, will show you exactly which part of it is failing.

This is just truth. PitchKitchen builds Magnetic Messaging Frameworks for founder-led B2B companies in the $5M-$75M range, and the segmentation conversation is where most engagements actually start, because a message can't be sharp when the group it's aimed at was assembled by a dropdown. Get the segments right and the message writes itself. Get them wrong and you'll spend another year making a census sound urgent.