You know what your customers do. But do you know why?

At the end of our  first segmentation piece “before you commission a segmentation, ask whether you actually need one“, we argued that before commissioning a segmentation, it is worth being very clear about the business decision it needs to support. When segmentation is the right answer, the next question is what kind of understanding you need it to create.

Today, businesses have access to more behavioural data than ever. We can see what people buy, how often they buy it, which channels they use, what they click on, how they move through an experience and, in many cases, how valuable they are to the business.

All of that can be incredibly useful. Behaviour gives us something concrete to work with and can reveal patterns that people themselves may struggle to articulate.

But there is a difference between knowing what somebody does and understanding why they do it.

That difference matters in segmentation, because two people can behave in apparently similar ways for completely different reasons. Equally, people who currently behave very differently may share needs or motivations that make them part of the same future opportunity.

If we only look at the behaviour we can see today, there is a risk that we become very good at describing the market as it already exists, without necessarily understanding what might change it.

What can behavioural segmentation tell you?

Behavioural segmentation groups people according to what they actually do. Depending on the category, that might include purchase frequency, spend, product usage, channel choice, engagement, loyalty, content consumption or any number of other observable behaviours.

There is real strength in that. Behaviour is tangible, often readily available and, importantly, can make segments easier to identify and target later. If the business needs to understand differences between its current customers, prioritise valuable groups or tailor activity based on established patterns of behaviour, it can be an extremely useful lens.

It can also challenge what people tell us about themselves. We do not always have perfect recall, and what we say we do is not necessarily what happens in practice. Looking at real behaviour can help close that gap.

So, this is not an argument for looking beyond behavioural data because behaviour somehow matters less. Quite the opposite. At 2CV, behavioural understanding is an important part of how we build a fuller picture of people. The question is whether behaviour on its own gives the business enough understanding to make the decisions it needs to make.

Sometimes it does. Sometimes it doesn’t.

Behaviour tells you what happened, but not always what drove it

Think about two customers buying the same product at roughly the same frequency and spending broadly the same amount. From the data, they might look very similar.

But one might be buying because the product makes life easier and removes something they find stressful. Another might be deeply engaged with the category and actively seeking the best experience available. One may have arrived through habit, while the other has made a very deliberate choice.

The behaviour looks similar. What sits underneath it is not.

That matters because those people may respond very differently when the market changes, a competitor appears, prices rise or a new proposition becomes available. They may need different things from the brand, and the opportunities they represent could be quite different too.

We see the same thing in customer journeys. In one project with a global IT business, what had been treated as a single customer journey turned out to contain four distinct journey types, shaped by different triggers, levels of urgency and decision-making needs. That clearer understanding helped the business target customers more effectively and identify new routes to growth.

The point is not that behaviour was wrong. The behaviour was part of the evidence. It simply did not tell the whole story.

Behavioural data can only show you what you can see.

Behavioural data is only as complete as what a business can observe and connect. A brand might know a great deal about what people buy from it, how they use its digital platforms or how frequently they engage, but much less about what those same people do with competitors or elsewhere in the category.

Even within your own data, there are limits. A till receipt tell you what someone bought, but not what they picked up and put back, what alternatives they considered, or whether the eventual choice was habitual or carefully deliberated. In the same way, a business may have extremely rich digital data simply because it is easy to capture, but much less visibility of what happens in retail or other offline environments. That fragmentation can also have consequences inside the organisation, with different teams building views of the same audience depending on the data they can see.

That matters when segmentation is being used to find growth. A highly engaged category consumer might look relatively unimportant in your customer database simply because they currently prefer a competitor, while people who do not buy from you at all may be almost invisible. If we are not careful, the segmentation can start reflecting the data that happens to be available rather than the market the business is actually trying to understand.

The important thing is not to segment on what is measurable simply because it is measurable. We need to understand what is meaningful for the decisions the business needs to make.

What does needs-based segmentation add?

Needs-based segmentation looks beneath observable behaviour to understand what people are trying to achieve, what matters to them and what shapes the choices they make.

Depending on the challenge, that can include functional and emotional needs, motivations, attitudes, tensions, priorities or the context in which decisions are being made.

It is also worth remembering that needs and behaviours are not always fixed to an individual. The same person can make very different choices depending on the occasion, context or what they are trying to achieve at that moment. In some categories, understanding those occasions can be just as valuable as understanding differences between people.

Looking at needs and motivations alongside behaviour can give a segmentation more explanatory power. Rather than simply telling you that Group A buys more often than Group B, it can help you understand what is driving those differences and whether they represent something that the business can meaningfully respond to.

It can also help expose opportunity that is not obvious from existing customer data.

If the segmentation is built entirely around existing behaviour, it can naturally pull the organisation towards the audiences, products and patterns it already knows. Understanding needs and motivations can widen the lens, helping identify people who may not currently buy from you, or whose current behaviour understates the opportunity they represent.

That sits closely with how we think at 2CV. In a complex market, the challenge is rarely a shortage of possible audiences or data. It is understanding which needs, people and opportunities genuinely matter, so that the business can decide where to focus with confidence.

The same behaviour can hide very different opportunities

One of the reasons this matters so much is that behaviour is often shaped by circumstance.

People buy what is available to them. They respond to price, distribution, habit, convenience and the choices a category currently gives them. What they do today is therefore partly a reflection of the market they are operating in today.

That makes behavioural data an excellent record of what is happening. But if the question is where future opportunity might come from, we sometimes need to understand what is not yet happening too.

Someone may rarely buy from a category because the current offer does not meet an important need. Another person may appear highly engaged simply because there are few alternatives. A group that looks relatively low value today could represent significant future potential if a particular barrier were removed or a need were addressed differently.

Those things are much harder to see if we only group people around their existing actions.

This is where motivations, needs and attitudes can help us see beyond the current shape of the market. They give us a way to understand not just where people are now, but what might cause them to move.

It is rarely behaviour VERSUS needs

It is tempting to frame this as a choice between behavioural segmentation and needs-based segmentation. In practice, we think that is often the wrong debate.

The richest segmentations tend to come from combining different sources of understanding and being clear about the role each one plays.

Behavioural data might show us what people actually do and help make the eventual segments easier to find again. Qualitative research can help us explore the context around those behaviours, uncover language and tensions, and understand motivations we may not have known to ask about. Quantitative research can then establish how widely those needs or attitudes are shared and how they connect with behaviour, value and other commercially important outcomes.

Each lens answers a slightly different question.

That combination can be particularly powerful because it avoids two extremes. On the one hand, we do not want a segmentation that is beautifully human but impossible for the organisation to identify or act on. On the other, we do not want something that is highly precise in describing customer behaviour but gives teams very little ideas of what those customers actually need from them.

Combining these different lenses does not mean putting every available variable into the segmentation. Behavioural data and needs-based understanding can both offer valuable truths, but the inputs still need to earn their place. The aim is to focus on the measures that genuinely distinguish between people in a way that matters for the business, rather than adding complexity simply because the data is available.

Why combine qualitative and quantitative research in segmentation

Qualitative and quantitative research play different but highly complementary roles in segmentation.

Qualitative work is particularly valuable when we are still working out what differences might matter. It allows us to explore how people think about a category in their own terms, understand the context around their choices and uncover needs or motivations that may not be obvious from existing data.

That can prevent the quantitative stage from being constrained by the assumptions the business already holds. If we only measure what we already know to ask, we run the risk of reproducing the existing view of the market rather than discovering something new.

Quantitative research then helps establish structure. It allows us to understand how needs, attitudes and behaviours relate to one another, how common different patterns are and whether the differences we are seeing are meaningful enough to support a useful segmentation.

The value comes from the combination. Qualitative understanding helps us ask better questions, while quantitative evidence helps us understand how those differences play out across the wider market.

The balance will not be identical for every project. It should depend on the business challenge, the data that already exists and what the segmentation ultimately needs to support. But starting with a broad understanding of people usually gives us a better chance of finding differences that matter, rather than simply differences that are easy to measure.

What makes a segmentation meaningful?

For us, a meaningful segmentation reveals differences that matter both to people and to the business.

That sounds obvious, but it is an important test. A dataset will usually contain plenty of ways to divide people. The question is whether those divisions help explain something important and create a basis for different decisions.

Sometimes behaviour will be at the heart of that. Sometimes needs and motivations will do more of the heavy lifting. More often, the useful answer will draw from both.

What matters is resisting the urge to segment around whichever data happens to be easiest to access. The starting point should still be the business question, followed by the kind of understanding needed to answer it.

That should shape both who the segmentation needs to represent and what information it is built around. For a retention or cross-selling question, existing customers and their behaviour may provide much of what is needed. But if the ambition is to identify future growth across a category, defining the universe around current customers, or relying too heavily on what they do today, can constrain the opportunity before the exercise has even begun.

The universe, data sources and defining inputs of a segmentation should be carefully considered based on the business question and the decisions the segmentation needs to support. Ultimately, the goal is not more data, but more useful understanding.

From understanding the right people to designing the right segmentation

The strongest segmentations give a business something it did not have before: a clearer way to understand people, explain meaningful differences and see where opportunity lies.

Behaviour is a vital part of that picture. But when we combine what people do with a deeper understanding of what they think, feel and need, the segmentation can move from describing the market to helping explain it.

That richer understanding gives the business a much stronger foundation for deciding where to focus and what different audiences might need from it.

But having the right ingredients is only part of the challenge. The next question is how you design the segmentation around the decisions it will actually need to support, and that starts long before the fieldwork begins.