Reflection on the personal data discussions

In week 3 of the MOOC we covered how personal data is used in digital marketing, and some alternative ways which give consumers themselves some control over the construction of their profiles. A lively and engaging discussion took place around the content.

Some agreed that behavioural targeting could end up being ‘creepy’ if it gets too accurate, much like the ‘uncanny valley’ hypothesis in robotics. The idea of self-created consumer profiles proved controversial. Many of the comments could see the appeal, but were still sceptical about whether it would take off as a viable alternative. In fact, many comparisons were made between the idea of self-created profiles and ‘good old fashioned market research’; the only twist being that it concerns a consumer segment of just one person.

Some comments flagged up the potential risks of self-authored profiles. From the perspective of marketers, they could be bad as they may be inaccurate. Some consumers might present a more ‘aspirational’ version of themselves, or might simply pollute their profiles with entirely fake information in order to disrupt the marketeers’ operations.

From the perspective of individuals, some raised the possibility that they could end up losing even more privacy by giving away even more information. This might include more intimate information about their personality, rather than just their purchases.

But others saw a win-win scenario for consumer privacy and digital marketing. One suggestion was to adopt a ‘hybrid’ model, where “the service tells you what it thinks based on your behavioural data and you have the opportunity to correct it”. Some predicted that despite lacking self-knowledge in many respects, individuals could create profiles that would be, on balance, more accurate than existing ones. However, the question remained about whether accuracy in itself is always good for the individual.

An interesting point raised was that there is always a gap between what people ‘want’ and what they actually do; sometimes our stated preferences differ from those we reveal through our behaviour. It’s not clear whether self-created profiles would strike a good balance between these two.

The fact that self-created profiles are being offered by commercial services raised some interesting questions. Some questioned the motivations and the business models of these services. Are they a fair deal, and how can they be trusted to manage the data responsibly? Another intriguing proposition was that by being offered some money for your personal data, even a small amount, you may end up valuing it even more. Which leads to a strange situation – many consumers happily give away their data for free, but as soon as they are offered some money for it, they may feel less inclined to share it at all.

For those who want to learn more about this topic, at the University of Southampton we researching the topics of transparency and control over personal data. We’re going to be analysing the themes discussed during this part of the MOOC. We’re also running a short study to explore the differences between self-created profiles and those based on prior behaviour. If you’d like to take part in that study, and get early access to the results of the research, we’ll be running it over the next few months. It will take just 5 minutes to complete!