Search Menu

What’s the future of customer segmentation?

Are we closer to the ‘segment of one’?

Customer segmentation is nothing new, with marketers of old categorising customers purely on who they were – going on age, gender, income and so on – usually resulting in between 5 and 8 categories of customer. Today, the huge amounts of customer data available – from multiple, digital sources –have transformed segmentation into not just who you are, but also what you buy and how you behave.

Segmentation allows marketers to better tailor their efforts to ‘types’ of audience, which can inform promotional, marketing and product development strategies. In an ideal world, of course, companies would treat everyone as an individual, develop projects and services that meet their exact needs, and engage with them in a way they want to communicate.

Historically, however, this has always proved too expensive, impractical and pretty tricky. So, today’s marketers have continued to band together people into similar groupings (again, usually between 5 and 8 segments) – albeit based on much richer information – according to their needs, behaviour, demographics and economic value. There are also different types of segmentation depending on what you are trying to do – for example, improve acquisition, communications, or product usage, etc.

But the world of customer segmentation is changing

Multiple interactions in the digital space mean that data is created from every digital channel that someone touches. Algorithms use this data to build profiles; find customers who are similar to that profile; then target them with a bespoke advertisement or suggestion. Netflix, for example, because of its huge and varied service offering, divides its customers into around 1,300 ‘taste community’ segments, to suggest shows to its customers. A pretty significant, and impressive, step up from the typical 5-8 segments.

At Capita, we are pushing segmentation forward for our clients in two ways:

  1. harnessing more data that is available about customers
  2. as customers move towards engaging with their clients in a more digital way, it allows us to create more granular segments, that we can use algorithms for, to identify and suggest next best actions.

From a wealth of data about our clients’ customers, we’re linking data from the organisations’ marketing, financial, and product databases, to create a single customer view. And, recent advances in text analytics, has meant that we can now start harnessing the vast amounts of unstructured data held on customers. For example, transcripts from when customers phone a call centre or when they email customer services. 

How can conversation data reveal what makes your customers want to leave? 

Andy Moorhouse, director of insight and analytics at Blue Sky, reveals 'the 4 things that drive customers mad'.

At Capita, we’re also beginning to design more sophisticated segmentations, with 100+ customer segments – meaning that the segments are a much tighter fit to the customers’ needs, thus servicing them is much more bespoke.

For example, we have:

  • Built a client segmentation to identify groups of people that don’t pay their bills. By understanding their customers motives and abilities, we can change how our client engages with their customers to increase the likelihood of payment.
  • Worked on an interesting segmentation that focuses on the reasons that people contact a company and what their preferred channel of communication is. By targeting certain contact segments, we can increase customer satisfaction and improve the clients’ bottom line.
  • Developed segmentations to predict people’s future needs. We’ve segmented council residents, to understand what services they engage with, and how. Early intervention on social isolation, for example, helps the citizen and saves on cost.

What does the future hold for segmentation?

The future of segmentation will be driven by two things:

  1. advances in computer speed and machine learning algorithms
  2. vastly more data available to use in the construction of the segments.

Internet of things will add to the amount of data collected about people. Everything from your light switches, your microwave, washing machine, etc, will all send data back to a central source. This data will be aggregated with the data from your smart watch, your phone, your laptop, your TV preferences, your online shopping, your LinkedIn profile, even your online dating profile.

The race is on for who holds the most data to train the next generation of AI algorithms to classify people and predict behaviour. Large companies (Amazon, Google, Microsoft, Oracle, etc) and governments are gathering up data from all the touch points they can. And we’re doing it too, putting segments at the heart of decisions we’re making about customer experience.

All of this points to a future where people are treated as individuals, with algorithms determining the best way to communicate with you, tailoring products/ services to your exact needs. As the algorithms identify taste communities not just for your TV preferences, but for all areas of your life, there is a strong possibility that the AI will know us better than we know ourselves.

Could a segment of one be a reality?

With so much data available, from such a variety of sources, and with the Gen Zs coming of age and running their lives entirely ‘mobile only’, there is the opportunity to build an entire, holistic view of every individual. An exciting proposition but, as some would argue, somewhat scary. Similar to the development of nuclear technology, this could be viewed as a really good or a really bad thing. The future will be ultimately determined by individuals deciding if sharing their data is worth the enhanced, superior service that they would ultimately receive.

Photo of Alan Merlehan

Alan Merlehan

Head of data and customer insight, Capita Transformation

Alan has 16 years’ experience using data and analytics to understand people’s behaviour. Prior to joining Capita, he has worked for several leading marketing and advertising agencies. He has a MSc in Operational Research and an MBA from Lancaster University.

Connect with us

Top