Customer Segmentation: Stop Marketing to Everyone

A customer segment is a distinct group of people or organisations that share common characteristics and respond similarly to a given marketing message or offer. Done properly, segmentation tells you who to prioritise, what to say, and where to spend, before you commit budget to any of it.

Most companies say they do segmentation. Very few actually do. What they have instead is a demographic profile dressed up as strategy, and a media plan that ignores it entirely.

Customer segmentation sits at the centre of go-to-market strategy. If you are thinking about how to structure your growth approach more broadly, the full picture is covered in the Go-To-Market and Growth Strategy hub, which pulls together the strategic frameworks that connect segmentation to positioning, channel selection, and commercial outcomes.

Why Most Segmentation Work Produces Nothing Actionable

I have sat through more segmentation presentations than I can count. The format is almost always the same: a research agency presents six personas with names like “Ambitious Alex” and “Cautious Carol”, each with a mood board, a list of media habits, and a paragraph about their values. The room nods. The deck gets filed. The media plan goes out the following week targeting 25-54s with a household income above a certain threshold.

The problem is not the research. The problem is that the segmentation was never designed to drive a decision. It was designed to demonstrate that a decision had been thought about.

Useful segmentation starts with a commercial question, not a research brief. The question is not “who are our customers?” The question is “which customers, if we won more of them, would most improve our business performance?” That reframe changes everything about what you measure and what you prioritise.

When I was running iProspect, we grew the agency from around 20 people to over 100 across several years. A significant part of that growth came from being deliberate about which client segments we pursued and, equally importantly, which ones we stopped pursuing. Winning a client in the wrong segment, one that stretched the team, eroded margin, or required capabilities we did not have, cost more than losing them. Segmentation in an agency context is not just about marketing. It shapes resourcing, hiring, and commercial model decisions.

The Four Segmentation Approaches Worth Using

There is no single correct way to segment a customer base. The right approach depends on what you are trying to decide. These four frameworks each have a different use case.

Demographic and Firmographic Segmentation

This is the most common starting point. For consumer brands: age, gender, income, life stage, geography. For B2B: company size, industry vertical, revenue, headcount, geography. It is useful as a filter, not as a strategy. Knowing your customers skew 35-54 and earn above the median tells you something about where they might be reachable. It tells you almost nothing about what they need or why they would choose you.

Demographic segmentation tends to be overweighted in media planning because it is what targeting platforms are built on. That is a platform convenience, not a strategic insight.

Behavioural Segmentation

This segments customers by what they actually do: purchase frequency, category spend, channel preference, product usage patterns, switching behaviour. It is more predictive than demographics because behaviour is a revealed preference rather than a stated one. Someone who buys your product every six weeks and has done so for three years is a fundamentally different customer from someone who bought once during a promotional period. Treating them identically in your marketing is a waste of budget on one end and a missed retention opportunity on the other.

Behavioural data also helps you identify the customers most likely to churn before they do. That is a more valuable use of segmentation than most acquisition-focused models acknowledge.

Needs-Based Segmentation

This groups customers by the underlying problem they are trying to solve, regardless of which product or service they currently use to solve it. It is the most strategically powerful form of segmentation and the hardest to execute because it requires qualitative depth, not just data analysis.

A financial services company might find that two customers with identical demographics have completely different needs: one wants certainty and protection, the other wants growth and control. Those needs require different products, different messaging, and different sales approaches. BCG’s work on financial services segmentation illustrates how needs-based models surface commercial opportunities that demographic cuts miss entirely.

Value-Based Segmentation

This ranks segments by their actual or potential commercial value to your business: lifetime value, margin contribution, referral rate, cost to serve. It is the segmentation approach most directly connected to P&L decisions, and the one most marketing teams are least equipped to do because it requires finance data they often do not have access to.

Value-based segmentation is where prioritisation becomes explicit. If segment A generates three times the lifetime value of segment B but requires similar acquisition investment, you have a clear resource allocation signal. Most marketing budgets are not built this way. They are built on historical spend patterns and internal politics.

How to Build a Segmentation Model That Actually Gets Used

The test of a segmentation model is not whether it is intellectually coherent. It is whether it changes what the sales team says on a call, what the creative team puts in an ad, and what the media team bids on. Most models fail that test.

Here is the process that tends to produce something usable.

Start With the Commercial Hypothesis

Before you touch any data, write down the commercial question you are trying to answer. “Which customer type, if we doubled our share of wallet with them, would have the biggest impact on EBITDA?” or “Which segment is currently underserved by our category and could be won without a price war?” These questions shape what variables you need in your model and what output would actually be useful.

Combine Quantitative and Qualitative Data

Transaction data tells you what customers did. Qualitative research tells you why. You need both. A segment that looks homogeneous in your CRM might contain customers with very different motivations and switching triggers. Relying on one data source produces a model that is either precise but shallow, or rich but unscalable.

The depth of qualitative work required depends on category complexity. In a low-involvement category, a handful of interviews might be enough to pressure-test a quantitative model. In a high-consideration B2B sale, the qualitative layer is often where the real insight lives.

Validate Against Actual Revenue

Map your proposed segments back to your existing customer base and revenue. If your top segment by strategic priority represents 8% of current revenue, that is useful to know. It might mean there is a growth opportunity, or it might mean your current GTM model is misaligned with where you want to play. Either way, it surfaces a decision.

Limit the Number of Segments You Actively Pursue

This is where most segmentation projects go wrong. The research produces eight viable segments. The business tries to address all eight simultaneously with differentiated messaging and tailored offers. The result is eight mediocre executions instead of two excellent ones.

Prioritisation is not a failure of ambition. It is the point of the exercise. BCG’s commercial transformation framework makes the case that focused resource allocation against fewer, higher-value segments consistently outperforms spread-thin approaches across a broader set.

Where Segmentation Breaks Down in Practice

There are a few failure modes I have seen repeatedly, across agency clients and in-house teams alike.

Segmentation Designed to Confirm Existing Strategy

The brief goes out after the strategy has already been agreed. The research is designed to validate a decision that has already been made, not to inform one that is still open. The segments that emerge conveniently support the existing plan. This is not segmentation. It is expensive decoration.

I saw this at a large financial services client where the segmentation project ran for four months and produced a model that looked almost identical to the one they had been using for six years. When I asked whether the brief had been written with an open hypothesis, the answer was essentially no. The brief had been written to justify a channel investment already committed to in the annual plan.

Segments That Cannot Be Reached

A segment is only useful if you can identify and reach the people in it. A needs-based segment defined by attitudinal characteristics is intellectually satisfying but operationally useless if there is no media, data, or channel strategy that can find those people at scale. The segmentation model has to be built with activation in mind from the start, not retrofitted to media planning after the fact.

This is one of the reasons behavioural segmentation has become more commercially useful as first-party data capabilities have improved. Behavioural signals are increasingly addressable in paid media, CRM, and on-site personalisation in ways that attitudinal profiles simply are not.

Segments That Change Faster Than the Model

A segmentation model built in 2020 may not reflect customer reality in 2025. Category dynamics shift. New entrants change switching behaviour. Economic conditions alter what customers value. A model built during a period of low interest rates and high consumer confidence will produce different segments from one built during a period of cost pressure and category consolidation.

Most companies treat segmentation as a project rather than a capability. They invest heavily in a model, use it for three years, and then repeat the exercise. A better approach is to build lighter, faster validation loops that flag when segment behaviour is drifting from the model assumptions. Tools like continuous feedback mechanisms can help surface behavioural shifts before they become strategic blind spots.

Segmentation in B2B: The Additional Complexity

B2B segmentation has a layer of complexity that consumer models do not face: the buying unit is rarely a single person. You might be segmenting companies by firmographic and needs-based criteria at the account level, while simultaneously having to think about the different roles within the buying committee, each of which has different priorities, different objections, and different influence over the final decision.

A CFO evaluating a software purchase is not the same segment as the IT director evaluating the same purchase, even if they are at the same company. The CFO is solving a risk and cost problem. The IT director is solving an integration and maintenance problem. Treating them identically in your marketing, as most B2B companies do, produces generic messaging that speaks to neither.

Forrester’s research on go-to-market struggles in complex B2B categories consistently points to misalignment between how companies segment their market and how buying decisions are actually made inside target accounts. The segmentation model assumes a rational, single-decision-maker process. The reality is messier, slower, and more political.

Account-based marketing emerged partly as a response to this problem. By treating the account as the unit of segmentation and building programmes that address multiple stakeholders within it, B2B marketers can align their model more closely with how complex purchases actually happen. The challenge is that it requires more content, more coordination, and more patience than most marketing teams are structured to deliver.

The Link Between Segmentation and Positioning

Segmentation and positioning are not separate exercises. They are two sides of the same strategic decision. You cannot define a credible brand position without knowing which segment you are positioning for. And you cannot prioritise a segment without understanding whether you have, or can build, a defensible position within it.

The failure to connect these two is one of the most common strategic errors I see. A brand will invest in a segmentation model that identifies a high-value segment, then brief the creative team on positioning work that makes no reference to it. The creative team produces positioning based on what the brand wants to say about itself, not what the priority segment needs to hear. The two pieces of work never connect.

Good segmentation work should produce a clear brief for positioning: here is who we are trying to win, here is what they currently believe, here is what they need to believe differently in order to choose us. That brief makes positioning work faster, more focused, and more likely to produce something that actually moves behaviour.

Segmentation as a Growth Lever, Not Just a Planning Tool

Most companies treat segmentation as an input to the annual plan. Build the model, set the priorities, brief the agencies, run the campaigns. Repeat next year. That is a reasonable use of segmentation, but it is not the only one.

Segmentation can also be used dynamically, as a growth lever that shapes how you allocate budget and attention in real time. If you have a clear model of segment value and segment behaviour, you can make faster decisions about where to increase investment when performance data signals opportunity, and where to pull back when it signals diminishing returns.

This is where segmentation connects to the broader growth strategy conversation. Forrester’s intelligent growth model argues that companies which build segmentation into their operating rhythm, rather than treating it as a periodic research exercise, make better resource allocation decisions over time. The model becomes a shared language for commercial decisions, not just a marketing artefact.

I would add one caveat from experience: the value of a dynamic segmentation model depends entirely on the quality of the data feeding it. A model built on clean, validated first-party data will improve over time as you learn which segments respond to which interventions. A model built on patchy CRM data and third-party assumptions will compound its errors. The investment in data infrastructure is not separate from the investment in segmentation strategy. It is the same investment.

There is also a growth hacking dimension to segmentation that is worth acknowledging. Many of the most well-documented growth hacking examples from high-growth companies involve identifying an underserved micro-segment and building a product or channel strategy specifically for them, before expanding. Dropbox’s early focus on technically sophisticated early adopters, Airbnb’s initial concentration on specific event-driven demand, these were segmentation decisions before they were growth strategies. The segment came first. The growth model followed.

Segmentation is also directly relevant to pipeline and revenue planning. Research on GTM team performance consistently shows that teams with clearly defined segment priorities generate more qualified pipeline than those working with broad, undifferentiated targeting. That is not a surprising finding, but it is one that many organisations have not operationalised.

If you are working through how segmentation connects to your broader commercial model, the Go-To-Market and Growth Strategy hub covers the full strategic architecture, from market prioritisation through to channel strategy and performance measurement, in a way that keeps the commercial logic connected throughout.

What Good Segmentation Looks Like in Practice

A useful segmentation model has five characteristics. It is built around a commercial question, not a research template. It combines behavioural data with qualitative depth. It produces a prioritised list of segments, not an exhaustive taxonomy. It is connected to positioning and channel strategy from the outset. And it has a defined review cadence so it does not become stale.

The companies I have seen use segmentation most effectively tend to share one trait: they treat it as a commercial capability, not a marketing deliverable. The output is not a deck. It is a set of decisions about where to play, what to say, and how to win. Those decisions get embedded in briefs, in budget allocation, in sales training, and in product roadmaps.

That is a higher bar than most segmentation projects are set up to meet. But it is the only bar worth setting.

There is a version of marketing that works as a blunt instrument: spend broadly, message generically, optimise for the metrics that are easiest to measure. It can produce short-term results, particularly in categories with strong demand and weak competition. But it does not compound. It does not build preference. And it does not tell you anything useful about where your next period of growth is coming from.

Segmentation, done properly, is the alternative to that. It is the decision to be specific about who you are trying to win and why, before you spend a pound on reaching them.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions