A Caution on a Hot Trend

The buzz about automatic notification of car crashes makes great sense… for insurers… but not always for customers. Some caution is in order.

Sharing Information After Car Crash

Lots of experts see cars’ increasing ability to automatically notify insurers and public authorities about crashes as a breakthrough. And it is. Authorities will quickly be able to dispatch help, if it’s needed, and insurers will be able to take hold of the claims process so early that they’ll be able to knock out lots of cost while speeding the repair process.

But what if customers don’t want to automatically notify insurers and public authorities about a crash?

Many will, of course, appreciate the peace of mind that comes with knowing that help will be sent even if they’re incapacitated in an accident. Many will feel coddled after they have even a minor accident and their insurance company calls or texts to ask if there’s anything the company can do to help. If customers get the sense that they’re being moved more quickly through the repair process, they’ll like that, too, and some PR and advertising could even sell customers on the idea that having costs taken out of the claims process allows for lower rates.

So, what’s not to like?

Well, we’ve already seen that crash detection can lead to many false  positives — Apple Watches, for instance, seem to think that many sudden stops by skiers are actually car crashes and dispatch help. Those annoyances will diminish as the technology improves, but they won’t go away.

And some people simply don’t want to report a crash. They don’t want it on their driving records, because it will raise their insurance rates and perhaps even imperil their license. 

Even if a driver welcomes having a crash reported, they may not want to do it right away. Maybe they’re late for a meeting. Maybe they’re discombobulated and want a chance to clear their heads. 

In rare cases, maybe the driver doesn’t want to use the tow service or repair shop that the insurer is steering them toward as part of cost-cutting efforts. 

Broadly speaking, I’m raising the issue of decision rights. We’ve seen lots of big plans for technology fail over the years because companies failed to account for the fact that consumers value those rights. I worry that the effort to enable automatic crash notification, while totally laudable, will at least be slowed unless insurers find ways to leave those decision rights with consumers. 

Remember “internet refrigerators”? The grand idea was that your refrigerator would know when you were out of, say, milk and would order it for you. But you don’t have an internet refrigerator, and I don’t, either. Why? Because I don’t want a refrigerator making that decision for me. I’d be delighted to have a refrigerator that knew whether I had dill and could tell me while I was at the store, while no one was at home to check, but I don’t want a refrigerator to automatically keep me stocked with it. Sometimes, we need dill. Sometimes, we don’t.

There is a straightforward way of addressing the decision rights issue: Ask the customer.

Do you want your insurer to be notified immediately if you have an accident? How about public authorities? Given that false positives are a possibility, how certain should our algorithms be that you’ve had an accident before reporting it? How serious an accident should we report — one at 5mph, 10mph, 25mph? For someone who’s had an accident, maybe start by asking if they want to report it now, then if they’re okay with using the official towing operator and repair shop. Etc.

Automatic reporting can be a game-changer, for all the reasons our friends Stephen Applebaum and Alan Demers lay out here. They estimate that the technology, now possible because of smartphones, could save an average of $1,000 per claim on the 15 million auto claims filed just in the U.S. each year, for a whopping $15 billion in savings.

I just think everything will go far more smoothly if we develop the technology with the customer, and not just the insurance process, in mind from the very beginning.