But, I believe the reality is that we are at the very beginning of a transition. We are only just beginning to move away from traditional ways of segmenting and targeting customers (based on generalisations), towards approaches that achieve a fine grain understanding of individual interactions, data sharing and connections.
Good segmentation is fundamental to brands making the most of their market opportunities – from identifying, attracting and retaining the most profitable customers to identifying the most vulnerable customers to the competition.
However, most brands still use segmentation models that are based on averages and assumptions and that focus their products and services at generalised groups of customers. These segmentation models, whilst gradually improving, often fail and for many reasons: lack of alignment across the business, insufficient operational backup and not refining the segmentation model over time.
But probably the main reason for failure is the oversimplification of the customer decision-making process. Because to understand that requires defining the customer’s rational and emotional criteria at each touchpoint in the customer experience. And that is difficult.
Customers have access to near perfect information, from their mobile phones and from peers in their social networks, influencing their purchasing decisions. In this intensely competitive environment, brands are finding it increasingly difficult to differentiate themselves. This makes customer experience an essential area to focus on, along with new approaches to segmentation.
Due to the impact of digital devices on customer buying patterns, segmentation models will inevitably become more data-driven. This brings the opportunity (in theory at least!) to target customers on a real-time basis at an individual level. This could help a brand shift the basis on which they compete to one of value rather than price.
The promise of Big Data is that it provides insight about people’s actual behaviour not just their beliefs or stated intent. Rather than dealing with generalised customer segments, brands are able to find patterns in their data, made up of millions of small transactions between people that may begin to explain specific peaks in sales or high demand of a particular product, for example.
Overlay this with social data and the analysis becomes even more interesting.
Analysing the things an individual posts on a social network like Facebook in isolation isn’t reliable enough because more often than not it just shows what they want others to hear about them. The more intriguing opportunities lie in the integration of unstructured social data with other data sets, such as location data from mobile phones or credit cards. It is this continuous ‘data exhaust’ that tells our real story about who we really are, where we spend time and what we actually buy each day. Where social data can really play a part is the important study of connections between people – something that traditional segmentation methods do not cover adequately.
Developing an understanding of someone’s actual behaviour and then making inferences about their other behaviours (who is most likely to be a safe insurance risk or a frequent ‘switcher’ of your broadband service) becomes possible because of social context.
Comparing you to others in your social circle helps to determine the behaviours that the group thinks are normal and that they have learnt from one another.
This of course means that because Big Data is increasingly about people, the privacy and data ownership issues are only going to become more central to a brand building trust with its customers.
The exciting and more futuristic implications of understanding the way people connect and influence one another are that we can start to see how people form entire markets – financial, government and companies. Once we understand that, we can make predictions and design services that are infinitely better than those we have now – from better transportation systems to dramatically improved public health services.