This is an issue worth exploring because it lies at the heart of the way in which insurers design and implement their digital decision systems. As time goes on, more and more decisions in underwriting, claims and counter fraud could be influenced by this ‘can we trust the consumer’ narrative.
So is the data broker correct? Is the consumer data that the insurer sources through data brokers more accurate and trustworthy than the consumer data sourced directly from the consumer? After all, isn’t it all consumer data? What the data broker is saying is that they are less likely to forgot some pieces of ‘less than favourable’ data than the consumer.
And the data broker goes further, by referencing the cost of living crisis and its impact on those on “the lower end of the financial spectrum that are seeing these pressures”. A clear line is being drawn then between those on lower incomes and those who will be more tempted than others to ‘forget’ that piece of data. What is being implied here is that your credit rating is a proxy for your potential as a fraudster.
Sales and Nonsense
Such narratives are becoming more commonplace in insurance now. They’re also a potent mix of sales and nonsense. Research confirms that it is people on middle incomes who are more likely to commit fraud, not those on lower incomes. Yet if both the seller and the potential buyer are people of middle incomes, that may not be a message either would be comfortable with.
Let’s turn to the claim of accuracy. Is primary data provided by the data subject less accurate than secondary data sourced via a data broker? There are two ways of looking at this : context and completeness.
If I give out data in one context, that does not mean that that same data means the same in another context. Of course, that’s not something a data broker is likely to acknowledge, for their business model is built around the secondary use of consumer data. Every insurer’s due diligence for data should weigh up that secondary use question, for it is associated with a range of privacy and data ethics concerns.
As for completeness, it’s important to remember that completeness is a question that hangs over both the primary data provided by the data subject and the secondary data sourced via a data broker. Neither can say “I have it all”, and so neither can question the other, without acknowledging the incompleteness of their own data. Unless of course, you’re investing in controlling the sector narrative.
Context and completeness can of course be linked. Do you know, goes the narrative, that people who drink bottled water produce lower claims costs than those who drink tap water? So is the consumer who doesn’t disclose their water buying habits to their motor insurer guilt of providing incomplete data? Do I really need to answer that?!
The insurer with an at least half decent data ethics framework should be identifying these questions and using them to challenge narratives like this being circulated around the sector by data brokers. Procurement people in those insurers should be involved so that their critical eye can strip out that ‘we would say that, wouldn’t we’ part of the sales pitch.
Am I making a mountain out of a molehill, some of you may ask. Well, data brokers and software houses played a big part in the success of lifetime value pricing, until that is, the regulator was forced to ban it. Given a trend emerging in regulatory circles for algorithms to be made to unlearn things derived from unacceptable data practices, I think a ‘molehill’ is not the right analogy.
Finally, insurers should weigh up how they engage with that ‘can we trust the consumer’ narrative. How they handle that narrative should be guided by what type of relationship they want their customers to have with them. To put it succinctly, are you a data first insurer or a customer first insurer? If you’re not going to trust the data provided by the consumer, then there’s only one answer, and that answer is a pretty damning one for the sector’s reputation.