In the going on 10 years that I've been editing ITL, there have been two truisms about claims.
First is that claims are "the moment of truth" for insurers. That's certainly true, and, taking that notion to heart, insurers have made real progress. They've made it easier for insureds to report claims -- via app, sending in their own pictures rather than waiting for an adjuster, and so on. They've used new technologies, such as drones to survey damage after a storm, and have become much better at triage so they respond faster to the situations that are the most important and most urgent. Many have institutionalized compassion, for instance by quickly providing money to people forced out of their homes in a major storm, rather than making them wait for a full inspection and settlement. I could go on.
The second truism, which has taken its full form more recently, is that the industry needs to get to straight-through processing. It's certainly worth heading in that direction. You can already see the benefits that have come from those apps for reporting claims, submitting photos of damage, etc. But it also seems to me that making straight-through processing the Holy Grail has obscured some real opportunities for progress.
I'm thinking, in particular, of all the attention paid to AIs that can look at pictures of damage to cars and, increasingly, homes and render a quick estimate for repairs. The technology is as cool as can be, but having the AI take the place of an adjuster may take two or three minutes out of a process that lasts days or weeks -- and developing the AI is expensive. I believe the AI will eventually prove its worth, but, in the meantime, it may be distracting us from ways that we can shorten the process much more while saving money, rather than spending it.
For instance, as you'll see in detail if you read this month's interview with Andy Cohen, president and chief operating officer at Snapsheet, an insurtech focused on claims management, there are massive savings to be had by implementing a workflow redesign and using AI to play traffic cop on a claim -- acknowledging receipt of documents, sending status updates, responding to routine inquiries from the claimant, etc. None of that is remotely sexy, like the AI that immediately spits out an estimate for repairs, but it takes an awful lot of the burden off adjusters, who can then devote their time and intelligence to more complex (and interesting) issues.
There are plenty of other areas, too, where technology can take time and money out of the claims process without trying to duplicate the straight-through processing attention-grabber that Lemonade gave us a few years ago when it paid a claim in three seconds. AI can respond instantly and help sort through issues with tow trucks that can come up in the first half-hour after a crash and that can sidetrack a car, and thus the claim, for days. At a more mundane level, AI can do a pre-evaluation of a claim file and point out important gaps or discrepancies right as the adjuster starts to dive in, rather than making the person sort through a thick stack of documents and find them all himself or herself. AI can triage files to make sure that the files that need the most immediate attention get the most immediate attention.
So, while claims should certainly be sped up as much as possible and while lots of expense can come out of the process, I think we're better off if we think of straight-through processing just as a useful concept, while focusing on fixing discrete problems that technology can tackle now and deliver significant benefits.
As a New Year's resolution, I could aim to get back into the kind of shape I was in at 25 years old -- but I'd just hurt myself. I've set more realistic exercise and health goals and suspect I'll be a lot better off.