As part of the WARC Guide to segmentation, Alan Mitchell argues that responsible marketers need to be prepared for a new frontier of real marketing segmentation led by data relationships.

GDPR - Europe’s General Data Protection Regulations and now the California Consumer Privacy Act (CCPA)  introduce new rules about how consumer data should be collected and used.

Responsible marketers need to be prepared. While the impact of GDPR is debatable, one straw in the wind is the UK Information Commissioner Office’s current investigation into how personal data is used in real time bidding in programmatic advertising. The ICO has been cautious and collaborative in its approach. But its intent is clear: “we want to see change in how things are done”.

Digital marketers defend and promote their activities, driven as they are by wholesale data gathering (or surveillance as it is now increasingly termed), by saying it helps make advertising messages more ‘relevant’ and therefore, by implication, more valuable to their recipients. But all too often they’re duping themselves (and others) with data.

Never mind widespread click fraud and other questionable practices, all too often response rate data doesn’t tell us what we think it’s telling us.

Even more important, even with the vast amounts of data collected by today’s technologies, the picture advertisers build of peoples’ lives is still extremely limited and often distorted. Peoples’ current moods, modes, life goals and current priorities can only be inferred by digital data - often very inaccurately.

Much of the resulting ‘segmentation’ is deeply flawed if not bogus. Real marketing segmentation is about understanding different groups of consumers’ differing needs and desires - to meet them better. Advertising segmentation, on the other hand, isn’t interested in meeting consumers’ needs. Its only interest is in meeting advertisers’ needs: to reach target audiences more efficiently and effectively. That’s perfectly understandable. But as a consumer product, most of what results fails the first hurdle of good marketing: that of consumer value.

The problem goes even deeper. It’s a dark secret of modern advertising that theoretically speaking, most current practices are based on the stimulus-response theories of the long-discredited behavioural psychologist BF Skinner, who believed that it was possible to control animals’ (including humans’ behaviour) by finding the right stimuli to get the right response. Since then, the advertising holy grail has been to find the right stimuli. Ever finer ‘targeting’ of different consumer ‘segments’ is part of this century old wild goose chase.

But that's not how advertising actually works. And people don’t work like that either. They are not and never were ‘targets’ for ‘messages’ that change their behaviour from outside. They are internally motivated searchers, searching for what has value to them, dismissing ‘stimuli’ that fail to reach this value threshold.

For all these reasons combined - questionable data, questionable interpretations of this data, questionable theoretical foundations, questionable value - online targeting isn’t half as ‘effective’ as its proponents make it out to be. If it works, it’s only at the furthest margins, like wandering in a storm trying to catch a few drops of rain in a cup and then declaring success by making the cup ever so slightly bigger.

By focusing attention on the issue of ‘consent’ regulators are unwittingly opening up this conceptual can of worms. After all, consent wouldn’t be ‘a problem’ if consumers really wanted to be the subject of data surveillance; if they really wanted to receive the resulting ads. But most don’t. This is the elephant in the room of 'consent'.

Yes, the stellar growth of digital marketing is a by-product of a tectonic shift in how consumers relate to marketers. But there is a bubble element to modern ad-tech and it’s going to deflate (if not burst) because, right now, its main beneficiaries are neither consumers nor advertisers but ad-tech middlemen.

Meanwhile, marketers do, indeed, have a data challenge on their hands. Since the year dot, the consumer/marketer relationship has revolved almost exclusively around the product: its value, trustworthiness etc. Now digital is adding a new dimension: what is our data relationship with our customers? Is it one that adds value and builds trust? Or is it one that destroys value and undermines trust?

With this, a new front in real marketing segmentation is opening up. What sort of data relationship do different groups of consumers want with us?  So far however, very very few marketers have got to grips with it.