Your Customers Don't Have Segments — They Have Attention Signatures

Demographics are dead. Needs-based segmentation is the replacement. Neither one can explain why two people who want the same thing make opposite decisions.

I was at a conference recently where a very senior marketer stood on stage and said, with visible conviction, “We know our customer. She’s 28 to 35, urban, household income above £50k, and she cares about sustainability.”

I wanted to stand up and ask whether she cares about sustainability the way a Thinker cares about sustainability — where she wants the carbon audit, the supply chain map, and the ISO certification number — or the way a Socialiser cares about sustainability, where she wants to know that people like her care about it too, and the brand makes her feel part of something. Because those are not the same customer. They do not make decisions the same way, they do not even process information the same way, and if your segmentation model cannot tell you which one is looking at your landing page right now, you are spending money on the wrong assumption.

The industry knows this, by the way. It has known it for a while. McKinsey’s 2026 consumer outlook segments wellness buyers by need states now — “maximalist optimisers,” “wellness shirkers,” and the like — which is a long way from age and income. Experian layers behavioural and psychographic data on top of demographics because the demographic layer alone was, in their polite corporate language, “insufficient.” And Roland Berger’s new consumer playbook argues that loyalty itself is fragmenting by psychological profile, which is their way of saying that the old categories have stopped working.

They are all moving in the same direction: away from who the customer is and toward what the customer needs. Which is better. But it is still not enough, because “what they need” is the output — the architecture underneath it, the cognitive machinery that determines how they evaluate what they need, is the variable that actually explains the decision.

The Four-Chunk Bottleneck Nobody Talks About

Here is something that should have restructured marketing the day it was published.

Nelson Cowan’s working memory model (1988, 2001) established that the human Focus of Attention holds approximately four chunks of novel information at once. Four. Your beautifully crafted value proposition, your six-bullet feature list, your brand story, and your social proof carousel are competing for four slots — and the rest is noise. The brain does not store it for later; it simply does not process it.

This is a hard constraint, and you cannot design around it. You can only work with it. The question that matters — the one most marketing teams never ask — is which four chunks get in. Because what enters those slots is not random; it is determined by an attention allocation pattern that is shaped by motivational orientation, cognitive tempo, and the appraisal filters that determine what the brain treats as signal versus noise.

I call this an attention signature. It is not a personality type, it is not a segment, and it is not a persona with a name like “Busy Mum” or “Ambitious Professional.” It is the default processing architecture of a particular kind of mind, and STAR identifies four of them. Every product, every campaign, and every landing page is being evaluated through one of these four architectures, whether you have designed for it or not.

Same Product. Four Brains. Four Completely Different Decisions.

Let me make this concrete. A premium electric SUV — same car, same website, same showroom, same salesperson. Four customers walk in, and four completely different cognitive journeys begin.

The Socialiser walks in and asks who else is driving one. Their Central Executive — the component of working memory that decides what gets processed — scans for social context before anything else: who owns one, what the community looks like, whether their family will enjoy it, and whether this is a car that people like them drive. They chunk information relationally, which means the spec sheet is not irrelevant but it is not what gets processed first. What gets processed first is whether the product activates their relational schemas — the mental models built from years of reading group dynamics, social norms, and who is connected to whom. A testimonial from someone they identify with will outperform a battery range chart, not because they are unsophisticated, but because their architecture treats social proof as primary data. Feed a Socialiser a spec-heavy landing page with no human element and you have built the marketing equivalent of a room with no heating — they will leave, they will not be able to articulate why, and they will just know something felt wrong.

The Thinker walks in and asks for the engineering spec. Their Central Executive scans for evidence, logic, and structural integrity: what is the battery range under real-world conditions (not lab conditions), what did the Euro NCAP tests actually show, and how does the regenerative braking system work at the thermodynamic level? Social proof is noise to a Thinker, and so is aspirational copy — they build deep, hierarchical schemas organised by causal chains and systematic principles, which means they want the white paper, the comparison table, and the third-party teardown. If your marketing cannot survive their scrutiny, they will draw the correct conclusion: that your product probably cannot either. The failure mode is vague aspiration, because “the future of driving” means nothing to a Thinker if you cannot explain the mechanism. Their cognitive load spikes not from complexity — they love complexity — but from imprecision; vague claims are static, and static is what makes a Thinker close the tab.

The Adventurer walks in and asks for the keys. Their Central Executive scans for possibility: what can this thing do, where can it take them, and what opens up that was not open before? They chunk by opportunity, and their schemas are action-oriented and procedural, built from sequences and outcomes rather than relationships or logic. The test drive is the spec sheet, and the experience is the evidence — hand an Adventurer a PDF and you have fundamentally misunderstood how their brain allocates its four precious chunks of Focus of Attention. They will not read it, not because they are lazy, but because their architecture does not prioritise static information delivered through a passive channel. The failure mode is over-specification: too many details, too many constraints, too much “here is how it works” and not enough “here is what it makes possible.” An Adventurer’s cognitive load spikes from constraint rather than complexity, so give them the keys and get out of the way.

The Realist walks in and asks about the warranty. Their Central Executive scans for risk: what happens when the battery degrades, what the warranty coverage actually says, whether the charging infrastructure is reliable in practice rather than theory, what the resale value looks like after three years, and what the known failure modes are. They chunk by reliability, and their schemas are built from evidence, precedent, and track records — which means they are not drawn to novelty or community but to the warranty page, the safety rating, and the maintenance schedule. The information that most marketing teams bury because they assume nobody reads it? The Realist reads it, because the Realist’s attention signature elevates it to the top of the processing queue, while the aspirational brand story that your creative team spent three weeks on does not even register. The failure mode is uncertainty: unproven claims, vague promises, and “trust us.” The Realist does not trust. The Realist verifies. Increase the ambiguity in your information environment and you do not intrigue a Realist — you lose them.

The Part That Actually Matters for Your Next Campaign

Here is the insight, and it is the one that almost nobody in marketing has fully internalised.

The same message, delivered in the same format, imposes different cognitive costs on different attention signatures. A message that activates a Thinker’s analytical schemas is processed efficiently by the Thinker, but that exact same message, reaching a Socialiser, consumes Focus of Attention capacity without activating any relevant schema — it is not rejected, it is not disliked, it is simply not processed, because the brain treats it as noise and moves on.

This is not a preference difference. It is a processing difference. And it is the mechanism that explains every piece of research that says “personalisation works” without being able to tell you why it works for some people and not others.

McKinsey has moved to need-state segmentation, Experian has layered behavioural data on top of demographics, and Roland Berger argues for psychologically specific loyalty strategies. They are all circling the same insight: the consumer is not a category but a processing system. But none of them have answered the question underneath, which is why needs-based segmentation works in the first place. STAR answers that question, because the four types are not labels — they are attention signatures, default patterns of cognitive allocation shaped by motivational need, processing tempo, and appraisal filters. The cognitive architecture underneath the intent.

What You Should Actually Do About This

If your segmentation model does not account for how different architectures process the same information, you are optimising the wrong variable. You are A/B testing button colours when you should be building four versions of the same page, each engineered for a different attention signature: the Socialiser needs the community, the Thinker needs the evidence, the Adventurer needs the experience, and the Realist needs the safety margin. Not as nice-to-have variations, but as structural requirements for their brain to process your message at all.

The next time someone tells you their segmentation model is psychographic, ask them one question: does it account for the cognitive architecture underneath?

If the answer is no, you are still spending money on the wrong assumption.

David Chadderton spent his twenties and thirties teaching people how to make life-or-death decisions at forty thousand feet. He now applies the same principles to consumer psychology, which, depending on the brief, can feel equally high-stakes. He’s the creator of the STAR Framework and the author of The STAR Framework: Rewriting the Rules of Consumer Engagement (NYC Big Book Award 2025), The STAR Operating System: Decode Mindset, Understand Motivation, Transform Human Behaviour, and Dear Algorithm, It’s Not Me, It’s You. By day, a Chief Marketing Officer. By night, a behavioural science obsessive who writes The Unoptimised Human because he can’t stop thinking about why people do what they do.