Key takeaways

  • Customer segmentation was once heralded as the best way for a company to understand their customer base.
  • After falling out of fashion, segmentation is making a comeback with insights that will fuel the whole company.
  • Not all segmentations are equal – there are mistakes to avoid and best practices to follow.

Forget what you think you know about customer segmentation. 

Beyond just better marketing, it has the ability to cut through to the heart of your business strategy. If you don’t believe me, that’s understandable. Customer segmentation has been in vogue, fallen out of fashion, and come back many times. 

Done right, it is a way to decide on business strategy, to pinpoint growth opportunities, to acquire and retain customers, and to build loyalty through better customer service. Moreover, in the age of ‘personalisation’, a structurally sound customer segmentation is key and can accomplish more than one goal, zooming its view from strategic oversight all the way down to the customer of one.

The art
of segmentation

Not all customer segmentations — the division of customers into smaller groups of similar traits — are the same, and a lot of that has to do with the data they are built from. Traditionally, and before data and analytics were as prolific and accessible as they are today, segmentations were built using ‘top down’ market data. This tends to be ‘big picture’, based on broad swaths of people, such as income brackets, ethnicity, or location. Nowadays in business, this high-level data is not that useful. You may be able to understand how many of your customers are likely to be middle-aged female homeowners, but it isn’t really an actionable insight.

To get more nuance, needs-based or attitudinal segmentations were developed for example, by conducting focus-groups or surveys. These can of course be small or large, but they are almost never based on all customers, and sometimes not even on your own company’s customers. The accuracy, therefore, is still not good enough to use for targeting, though it may give a business a ‘general idea’ of who its customers are and what some of them may want.

The more recent way to know who your customers are is to leverage their actual data. Data-driven and behavioural segmentations are based (assuming your data retention and information management is up to scratch) on all of your customers and they are detailed, including behavioural attributes and channel preferences (email, phone, social etc). It should incorporate all the interactions, across all mediums, that you have with your customers. And while it probably sounds like this is the clear winner when it comes to segmentations, in fact it is a combination of all three of these approaches to segmentation that provide a business with the most robust segmentation. 

Is customer segmentation
really useful?

Done well, customer segmentation is more than just a good marketing tool. With the right segmentation, knowledge of the best way to use it and the capabilities to do so, it will reveal tangible insights and prioritise business initiatives.

For customer acquisition, segmentation aids in understanding the people your product will appeal to and how to market to them, and in retention, can identify links between lost customers and your product or experience. For customer satisfaction it will help in prioritising who is ‘at risk’, or your most valuable customers for preferential treatment such as incentives, dynamic pricing, or going to the top of the customer service line. At the same time, customer lifetime value (CLV) can be more accurately projected through better understanding of a consumer’s journey through different segments.

Understanding market trends on the other hand, allows a business to identify untapped opportunities for instance, where to open stores, product needs in five years time as the population ages, and so on. The ability to gain insight into these macro shifts and how they interact with your customer base enables future preparation.

The key, I believe, is that your customer segmentation must have the ability to scale from simple to complex at the same time. It needs to be granular enough to facilitate meaningful prioritisation and targeting for action, but not too detailed to inhibit strategic business decisions. Knowing that every single one of your customers is a unique and complex person who wants different things can be paralysing, but knowing they’re all located on Earth is rather useless.

Common pitfalls
to segmentation

We find that when customer segments ‘aren’t working’ for a business it’s because they haven’t been built right. There may be too many segments or too few, they may not be robust, or simply not meet the needs of the business. Often, different departments have different segmentations that they use to yield different results. It’s not uncommon to find that they were created years ago and never updated. Here are some of the common problems I see that businesses need to avoid:

  • Lack of preparation — When beginning, a business needs to know who their customer is (individuals, account holders, households, business owners) and know what they will do if they fall into more than one category. People are complex, and that needs to be respected. Rather than creating multiple versions of the same customer in different categories or segments, you need to have a single view of the customer that is able to reflect this real-world complexity. Some thought also needs to go into how you are going to use and operationalise your segmentation, such as use cases, action plans, users and the information needed for making better decisions.
  • Data dramas It’s important to know what you have, and what you need, when it comes to data. Generally speaking, most data is good data to include, be it from marketing, accounts, billing, product or customer service. All of it paints the story of a ‘whole’ customer and their experience with your organisation and the resulting segmentation should always be linked back to their customer ID. Customers expect you to know the details about them, including their channel preferences so that when you have the right message you can deliver it to them in the right way. Including this data will also ensure that there isn’t an over-reliance on traditional market data that results in a segmentation that is too generic and unable to deal with personalisation. 
  • Poor implementation  There’s no point having customer segmentation if no one knows how to use it, what to use it for, or even that they should be using it in the first place. Staff need to be educated on what to use which segments for (macro and micro views), and how, so that they don’t misinterpret the data or revert to their trusty spreadsheets or previous segmentations. Segments should be easy to understand and differentiate, and having a segmentation that scales rather than multiple ones for different parts of the business will also reduce confusion.
  • Stale segmentation Markets and customers change. Your customers will need to move from one segment to another, and your system must allow them to do so to understand their true value and needs. This movement through segments should be tracked over time so the trends can also help in future planning and more accurate CLV projections. You should also aim to get customers into a segment as soon as possible, even when you only have a little data on them. For market-level changes, sometimes you’ll need to rebuild your segmentation. This will keep it fresh, and allow for new insights and opportunities to emerge.

The big
picture

Customer segmentation needs to be rethought. The value it can bring to a business is immense, particularly in an era when personalisation is becoming ever more important, but macro trends still need to be planned for. Its use is a real way to pull tangible insights from your data. After all, if you’re collecting it, you may as well make something useful out of it. 

While they may fall in and out of fashion, the value that can be gained from a good customer segmentation will never go away, and indeed, will continue to get more useful as AI finds ways to extract increasingly complex insight. Never has there been a better time to gather your data and divide (to) conquer.


Find out if your organisation has the capability to deliver strong customer analytics solutions by taking our self-diagnostic survey to better understand your data and analytic strengths and weaknesses.

 

Contributor

Phil Bolton

Phil is a director in PwC Australia’s Analytics team

More About Phil Bolton