Manufactured vulnerability is an ethical risk that marketing and compliance people in insurance firms need to have on their radars. So what should they do about it? Here are four steps.
The starting point is to set clear ethical principles to show your marketing people what are acceptable, and unacceptable, practices for engaging with consumers. These principles should reflect your business and so should be relevant to your products, target markets, typical customers and regulatory obligations. They should also reflect your business both as it is now and what it wants to achieve, for digital marketing is likely to be a factor in those ambitions. And these principles should be robust enough to guide people through some tricky ethical decisions.
These principles then need to be supported with some tools for your marketing people to use when faced with tricky decisions. They could be the two ‘fairness’ tests I outlined in this post from February 2014, or a variant of the ‘publicity principle’ – don’t use a practice that you wouldn’t be able or willing to defend in public. Then there’s empowering people to challenge rationalisations such as ‘everyone else is doing this’, or ‘no one will really be worse off’. And empowerment of this kind has to be accompanied by techniques for helping people to reach better decisions next time round.
There is a danger of course is that the machine learning that is integral to big data could result in firms never knowing when their systems are automatically manufacturing vulnerability. A set of big data principles setting out the values underpinning those big data algorithms will help of course, but will they be enough? Some people have proposed the introduction of internal algorithms, designed to crawl over your big data and automatically audit projects against key integrity principles. Better to design these for yourself before the FCA’s behavioural scientists introduce their own.
A clear framework of responsibility and accountability also helps bring the ethical risks from this type of marketing practices onto today’s table, even when their impact may not yet be fully apparent. Remember from the first post that a feature of the detriment was that it is likely to be progressive and pervasive.
There may be a temptation for you to think that this is all on the horizon and not really an issue for 2015. I would urge you to reconsider – at a recent conference, I listened to a start up intermediary describe how they were building schemes around online communities of people with special interests. How they talked about doing this signalled to myself and another consultant present that manufactured vulnerability was already nearer to a reality in insurance than most people think.