Traditional geodemographic segmentation today is much the same creature, give or take one or two minor advances, as it was when it first came on the scene in the 1970s. The tools that brands rushed to use, such as Acorn and Mosaic, have barely changed since their inception and the same is true for the concepts that underpin the basic ideas of consumer targeting.
The problem is that the world has changed so much since that those geodemographic products, once the very definition of pioneering advancement, have failed to keep up. While other industries have incorporated new data and methods, allowing themselves to morph into concepts unimaginable a decade ago, traditional geodemographic consumer targeting tools are still using the same basic data sets and ideas that made them revolutionary back in the 1980s.
Today, the value of the insight that those tools can provide is diminishing.
So what is the solution? GDPR, of course, has made finding the answer much more difficult. As a result of the data protection legislation, many of the routes for prospecting and lead generation are locked off to the industry and it is much harder to talk on a one-to-one basis with the customer based on their individual data.
What we need now is something that goes back to the traditional roots of geodemographics – a method that protects individuals, while harnessing the power of big data sets to find the real differences between modern consumers.
Knowing what we know today, the question is: if we could go back to the development of tools such as Acorn and Mosaic, how would we redesign geodemographic consumer targeting?
The key point to understand is that, with traditional geodemographics, we have put the cart before the horse – the cart being demographics and the horse being the actual purchasing behaviour, passions and interests of the consumer.
Until now, we’ve been thinking about what the demographic characteristics of a person can tell us about who that person is and what they might purchase; for instance, if they live in a certain area, are of a certain age and do a certain job, we might be able to deduce that they like foreign holidays. However, this is simply a "propensity", meaning they are twice as likely to engage with foreign holidays, based on results modelled from a survey base, measured in tens of thousands.
Now, with new data sets such as social data, we already know that certain people like foreign holidays before we’ve even tried to figure out where they live or what else they do. By adding that information, we can then complete the picture on who a customer is, how best to communicate with them and what channel to use to do so.
In short, instead of targeting using demographics, we need to target using what actually matters: people’s passions, motivations and mindsets.
It’s important to underline that demographics still have a role to play. Combining them with more dynamic and flexible data sets, such as social data, which can reveal who an individual is and what they are passionate about, results in a far more complete picture of the customer than what has ever been achieved before.
Furthermore, while much of today’s demographic data is static, often based on stale census or survey data, social data presents a picture that is updated in real time, because the information is constantly refreshed and available.
Consumer targeting has always evolved: in the 1950s, people were targeted based on what job they did; in the 1980s, it was all about where they lived; and in the 1990s, it became focused on what they were buying. Now, we need to focus on what people love in order to predict what they will buy. It’s time to usher consumer targeting into a new age.
Clive Humby is chief data scientist at Starcount. He co-founded Dunnhumby and created the Tesco Clubcard