Segmentation Part Two

Summer 2021

In Part 1, we established that segmentation divides customers into groups such that within a segment customers are similar but between segments they differ. That’s always interesting but segmentation is a time consuming and costly exercise so it’s worth doing only if it will help you to achieve one or more clearly defined objectives. These could include helping you to give customers a better experience, to communicate with them more effectively or to sell them more.

But defining the objective(s) still isn’t enough. You must also be clear how segmentation will help and what criteria it would need to meet to be helpful. For example, we might say that our objectives are to:

  1. Deliver a better experience to customers by tailoring it to meet the exact preferences of different segments

  2. Communicate more effectively with customers through segment-specific messages.

Almost regardless of the objectives a worthwhile segmentation exercise would have to pass the following tests:

  1. Understand: Does it improve your understanding of customers’ needs and preferences?

  2. Identify: Does it identify customers by segment?

  3. Reach: Does it enable you to reach customers by segment with tailored messages?


As we identified last time, there are many ways of dividing customers into groups but most of them will not help us to make customers more satisfied. Since the definition of a great customer experience is one that meets our customers’ needs perfectly, a segmentation exercise would fall at the first hurdle if it didn’t improve our understanding of customers’ needs and preferences. A really deep dive into customers’ preferences and the way they make choices will almost always demonstrate that your customers are not all the same and will provide a basis for dividing them into groups that would enable us to develop segment-specific experiences that would improve customer satisfaction. 


Many segmentation exercises pass the “understand” test by doing lots of research, gathering loads of data and using it to divide customers into groups labelled with catchy names based on characteristics the data say they share. This will often enable the organisation to design several experiences or levels of service calculated to appeal to different segments. But that’s of little business value if you can’t identify which segments your real customers fall into. You therefore need to ensure that any deep dives into understanding customers’ needs and preferences also gathers detailed profiling information that will enable you to identify which segments your real customers belong to.

Sometimes this will be easy. A tour operator, for example, might profile its customers according to family life cycle with such categories as young singles, young childless couples, young families, families with older children, younger empty nesters, older singles, older couples.

But what if you’re a housing association, a local authority or a bank? Family life cycle would certainly help the housing association but most demographic variables would not be good differentiators. Dividing customers into segments by gender and age bands would be poor predictors. Males in their 30s or 40s would have many different attitudes and preferences about housing, local authority services or banking. Apart from tailoring services for a few high net worth individuals, even income would not be a great differentiator for banking. 

A good profiling suitability test is whether the information can be recorded and updated on your database of real customers. 


The final usability test for a segmentation exercise is “reach”. If you have divided your customers into identifiable groups and you understand their differing needs and preferences you will be able to design segment specific offerings that should improve their satisfaction – but only if customers know about the different product or service offerings and have opted for the most suitable one. To make this happen you need to be able to communicate the right offerings to the right customers. And since customers' satisfaction is often based on perception as much as reality you must continue to tailor ongoing communications by segment. 

For direct, database-driven, communications this is relatively straightforward, but customers absorb messages that influence their attitudes from many other sources. Your segmentation research must therefore gather detailed information about customers’ media habits including traditional and digital media. This is obviously imperative for organisations that don’t have a database of customers but is also essential for companies with databases. First, you can’t be at all certain that customers will notice your direct communications, and second they will absorb messages and information from many other sources. And finally, of course, most companies also want to reach potential as well as existing customers via suitable media. 

Nigel Hill

TLF Research