So what might those ‘right questions’ be? They can be summed up as the ‘should we’ questions, to supplement the ‘can we’ questions that invariably come first. So while those digital systems are expanding the capabilities of insurers to assess and, if need be, repudiate claims, they also expose the claims function to new forms of attention, new questions, new demands.
So what, some of you may be thinking. What’s new about that!? Well, insurers like to say that they ‘want to pay claims’. It sounds good, and is good, so long as one big (some say the biggest of all) ethical risk is handled properly first. That ethical risk is of course the conflict of interest inherent in insurance, between the need to settle claims in line with the contact of insurance, while at the same time making as much profit as possible. Put simply, if claims go up, profit goes down ; if claims go down, profit goes up. Finding the right balance here is vital.
With a good number of insurance claims, the decision to either settle or repudiate is pretty straightforward. My house was hit by lightening in 2016 and the pile of charred electronics that I showed the adjuster made that a pretty straightforward decision to settle.
Tackling Maintenance
The problem though that insurers have long had, and which these digital systems are revolutionising, relates to policy terms like reasonable maintenance, and wear / tear. These are judgements, and judgements by their very nature rely on evidence and its interpretation.
Insurers have often lacked the scope and depth of evidence upon which to base judgements on a claim for storm or escape of water. Was the roof being looked after, and what does ‘looked after’ mean for this or that type of property? Likewise, has a plumbing system been looked after properly?
These are pertinent examples, for in recent years, insurers have been investing in digital systems to help them answer just such questions. Smart devices on pipes stream information to the insurer about pressure variations in a plumbing system. Aerial imagery and its associated analytics show the claims person what a roof is made of and the condition it is in.
Innovations like these give claims people a lot of information, yet it’s important to understand that they do not in themselves answer those earlier questions (was it maintained, etc). Such digital systems work on data ; the software’s selection and interpretation of that data produces outputs based upon the judgements baked into such systems by a mixture of their designers and their users. I often call these the dials and levers inherent in how such systems are designed and used. They’re designed and set by humans, not systems.
What Systems Output
Am I saying then that nothing has changed, despite all that investment in digital systems? Yes and no. Digital systems still output judgements ; as I said, these are baked into the software’s design and how the user sets the dials and levers. But at the same time, those judgements are now being produced at scale, at speed and wrapped up with a ‘technological sophistication’ wrapper.
This makes it important that insurers, in these evolving times for claims functions, adopt what I call a fair repudiation strategy. In that strategy should be found the judgements and values that the firm wants its people to follow, when setting those dials and levers, when evaluating the suitability of this or that piece of software, when handling tricky situations.
So we’re talking about it addressing the obvious ethical questions associated with repudiations, like fairness, bias, privacy and conflicts of interest. We’re looking to it for guidance on handling those ‘can we’ and ‘should we’ questions, along with the rationalisations (more here) that sometime crop up around them. And there should be clear evidence in that fair repudiation strategy about how it will shape decisions on the ground, and who is responsible for making this happen. Nothing rocket science really.
What about Consumer Support?
All this sounds a lot of bother, some of you may be thinking. After all, goes the narrative, there are surveys telling insurers that consumers don’t want their premiums to be used to pay the claims of policyholders who can’t be bothered to look after their property enough. And that’s a fair judgement to make, but on its own, it’s also a dangerous judgement to make. Rarely do such surveys go on to ask whether those consumers would want other people’s premium to pay for their storm or escape of water claim. The validity of the former relies on asking the latter.
This then turns to the question of ‘have I looked after my roof / plumbing enough to warrant having my claim paid?’ Invariably, consumers will answer that in the positive, but then, they would do, wouldn’t they! What insurers would want to see is evidence, such as a recent buildings survey report and associated invoices for remedial works. And for escape of water, this would some form of testing report for your plumbing system.
Ask your Board
Now, I am very confident that if I stood before the board of directors of an insurer, and asked them if any of them have had a surveyor physically inspect their home’s roof or plumbing system in the last five years, few if any would put up their hands.
What this means then is that while most consumers support insurers taking action on maintenance and wear/tear, on the assumption that it won’t affect them, few consumers are able to actually prove that their roof / plumbing is well maintained. These digital systems allow the insurer to say: this is our evidence ; where’s yours? They’ve moved the insurer from being on the back foot, to being on the front foot.
End of story then? Not quite. When using digital systems, the question of explainability always arises. There’s not enough of it around. A quarter of people whose household claim was repudiated say they didn’t know or understand why their claim was rejected. Another indicator to note is data from the Financial Ombudsman Bureau showing that 41% of all complaints they receive about buildings insurance are upheld.
One reason for this could be that the judgements output by insurers through decision systems on things like roof condition are in danger of becoming too genetic (more here). Statements like ‘not maintained enough’ and ‘this is wear and tear’ are too genetic for a specific aspect of a specific loss situation.
To Sum Up
Insurers adopting digital systems to tackle questions of maintenance and wear / tear in respect of storm and escape of water claims need to adopt a fair repudiations strategy that explicitly addresses obvious issues like fairness and conflicts of interest. Such a strategy needs also to prioritise explainability around the decisions being output.
There’s a lot of similarity at the moment between how price walking evolved, and how these digital systems are evolving in claims settlement. Vague notions that as some people benefit, it’s a good all round thing to do, plus ‘can we’ thinking rather than ‘should we’ thinking, trigger the same sort of warning lights as were obvious for price walking in 2017. A super complaint? I think not. A ban? No, but some form of push back for certain. It’s already taking shape in the US insurance market.