Why Your Personalisation Strategy Is Looking in the Wrong Direction

The most expensive assumption in marketing is that past behaviour predicts future behaviour. Here’s what to replace it with.

Most personalisation is backwards-looking.

It says: “You did this before, so we’ll assume you’ll do it again.” It builds segments from recency and frequency, serves recommendations based on what was clicked last time, and optimises for the pattern it already recognises. And it works, up to a point. For reordering, habit-driven purchases, and low-involvement decisions, behavioural personalisation is perfectly adequate.

But it breaks down completely when the customer’s context changes. And context changes are exactly when good personalisation matters most.

The Behaviour Trap

Here’s the problem with building your entire personalisation strategy on clicks and purchase history: behaviour is a symptom, not a cause.

When someone books the cheapest room available, that’s behaviour. But the reason behind it could be budget constraint, risk aversion, a desire for simplicity, or a dozen other things. The behaviour looks identical in your data. The motivation is completely different. And the motivation is what determines what they’ll do next.

Most organisations have built sophisticated systems for tracking what people do. Very few have built anything to understand why they do it. Which means the personalisation they serve is technically impressive and psychologically shallow.

You can see this playing out in real time across almost every sector. The student who booked the cheapest room in first year because budget was the primary constraint becomes the student who books the premium studio in second year because autonomy is now the driver. Same person, different motivation, completely different decision architecture. Click history won’t catch that transition. It’ll just keep recommending cheap rooms to someone who’s moved on.

What You Have to Unlearn

The unlearning required here is fundamental, and it starts with one assumption: that past behaviour is the best predictor of future behaviour.

It often isn’t. People change context. They change life stage. They change priorities. The person who bought the budget option last year might be buying the premium option this year, not because they’ve been upsold, but because what they need from the product has changed. Their motivational profile has shifted. The behavioural data hasn’t caught up yet, and if you’re relying on it exclusively, it won’t catch up until after the decision has already been made.

This is the hard part. The data infrastructure most organisations have built is optimised for tracking behaviour, not understanding motivation. The dashboards, the segments, the automated journeys, the recommendation engines, all of it is calibrated to what happened, not why it happened. Changing that isn’t a matter of adjusting a few settings. You’re not just changing the algorithm. You’re changing the question you’re asking of the data.

Instead of “what did this person do?” the question becomes “what drives this person’s decisions?” That shift sounds subtle. It is not. It changes everything downstream.

From Behaviour to Motivation

A genuinely psychology-led approach to personalisation starts from a different place entirely. Instead of segmenting by behaviour (“early bookers vs late bookers,” “high spenders vs low spenders,” “frequent vs lapsed”), it segments by motivational driver.

In practice, this means understanding that a customer’s decisions are shaped by a relatively stable set of psychological needs, even when their behaviour looks inconsistent. The person who booked early last year and late this year might be the same motivational type operating in different contexts. Or they might have genuinely shifted motivational profile. Either way, treating them as the same behavioural segment misses what’s actually happening.

What does motivation-led segmentation look like? It means identifying the psychological drivers that actually determine how people make decisions, things like whether someone is primarily motivated by security, autonomy, belonging, or competence. These drivers are more stable than behaviour, more predictive of future decisions, and more actionable from a messaging perspective.

The student who is security-motivated responds to guarantees, certainty, social proof from trusted sources, and messaging that reduces anxiety. The student who is autonomy-motivated responds to choice, flexibility, independence, and messaging that emphasises freedom. Same product. Same price point. Completely different psychological entry points.

If your personalisation is built on behaviour alone, you’ll send both of them the same message and wonder why the conversion rate is mediocre. If it’s built on motivation, you’ll send different messages to different psychological profiles and see the conversion lift that comes from actually speaking to the reason people decide.

The Conversion Lift Nobody Talks About

There’s a reason this approach produces better results, and it’s not complicated. When you speak to the actual reason someone makes a decision, rather than the pattern of when they tend to make it, you’re operating at a different level of the decision architecture.

Behavioural personalisation is essentially pattern-matching. “People who did X also tended to do Y, so let’s recommend Y.” It works when the pattern is stable. It fails when the context shifts.

Motivation-led personalisation is reasoning from cause. “This person is driven by security, so they’ll respond to messaging that reduces uncertainty.” It works when the context shifts, because the motivation is still there even when the behaviour changes.

The conversion lift comes from relevance at the right level. Not just “this is relevant to what you bought before,” but “this is relevant to why you make decisions.” That’s a fundamentally different proposition, and customers respond to it because it feels like being understood rather than being tracked.

Why Most Organisations Won’t Do This

If motivation-led personalisation is more effective, why isn’t everyone doing it?

Three reasons. First, it requires a different kind of data. Not just transactional and behavioural data, but psychographic and attitudinal data. That’s harder to collect, harder to structure, and harder to act on. Most data teams aren’t set up for it.

Second, it requires a different kind of expertise. Data scientists can build behavioural segments. Building motivational segments requires behavioural science, the ability to understand human decision-making at a psychological level. That’s a capability most organisations don’t have in-house and don’t know how to hire for.

Third, and this is the real blocker, it requires admitting that the personalisation infrastructure you’ve already built is solving the wrong problem. That’s a hard conversation to have with a board that’s invested millions in a customer data platform that’s optimised for behavioural tracking. The sunk cost is real, and the organisational inertia is powerful.

But the organisations that make the shift will have a significant advantage. Not because the technology is better, but because the question is better. “What drives your decisions?” is a more useful question than “what did you click on last time?” And building your personalisation strategy around the better question produces better outcomes across every metric that matters.

The Unlearning Is the Work

The hardest part of moving to a psychology-led approach isn’t the technology. It isn’t the data. It isn’t even the expertise.

It’s the unlearning.

It’s letting go of the comfortable assumption that more data equals more understanding. It’s accepting that a dashboard full of behavioural metrics might be telling you what happened without telling you why. It’s recognising that the segmentation model you’ve been refining for years might be grouping people by the wrong criteria.

That’s uncomfortable work. But it’s the work that matters, because the alternative is continuing to personalise based on a model that was built for a world where context was stable and behaviour was predictable. That world doesn’t exist anymore, if it ever did.

The organisations that figure this out first will be the ones that personalise in a way that actually feels personal. Not because they tracked more clicks, but because they understood more about why people click in the first place.

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.