Claims has often been called the “moment of truth” for insurers, especially in property and casualty lines. With few opportunities to interact with customers, a bad experience at the time of a claim can be a deal-breaker that sends customers looking for a new provider.
Incumbent insurers have spent nearly a decade focusing on customer experience in claims. However, the rising tide of insurtechs is pushing the envelope even further. Digital technology is being used to reconstruct accidents, find casualties before they are reported and pay out in close to real time.
“We see the claim lifecycle or the velocity through the claim process increasing with a lot of these new tech tools and automation,” says Pat Gee, SVP of claim for Travelers. “There’s been a lot more activity in this space over the past couple years.”
Gee says that geospatial technology is one of the prime digital innovations impacting claims. It is advancing at a rapid rate, giving insurers a closer look at affected property than ever before.
One insurtech working in the geospatial space, Understory, is focusing on hail events. The company is building networks of sensors in major metropolitan areas where hail is likely, including Dallas, Houston, Denver, Kansas City and St. Louis. The sensors give insurers a ground-level view into where hail is most concentrated, increasing the likelihood of a claim and allowing them to deploy adjusters more efficiently, says Understory founder Alex Kubicek.
“Radar and satellite are looking above the ground, but what’s happening in the sky doesn’t translate to what’s happening in the ground,” Kubicek explains.
Understory’s field team enters micro-leases with businesses, schools and other non-residential locations to install its solar-powered sensors. To illustrate the network’s density, Kubicek says there are a total of 150 in the Dallas-Fort Worth area, with about 10,000 people within two miles of one.
“With our data we know who has damage,” Kubicek says. “We are enabling carriers to better understand that and proactively reach out after an event occurs.”
Triggering payments
While Understory is looking to partner with incumbent insurers to improve their claims response, another insurtech plans to use some of the same concepts for a new type of coverage.
Structural engineer Kate Stillwell started Jumpstart with the goal of helping Californians recover faster from earthquakes. She says that after several years working in state, that cities like Oakland had the potential to be “the next [post-Katrina] New Orleans” if a major quake ever hit the area.
Jumpstart’s parametric insurance product, which will launch this year, pays out up to $10,000 nearly immediately to customers whose properties are located in a quake zone. The sum is enough to help customers find a place to stay and replace certain necessities while their home and/or renters insurance takes care of replacing their structure, Stillwell explains.
Jumpstart leans on the established network of seismic instruments around California to trigger events. However, Stillwell says, there are several metrics associated with quakes, and the company ran some simulations and customer surveys to decide which one would trigger a payment.
“One of the big questions was whether we make the trigger based on magnitude, or based on shaking intensity,” she says “People understand magnitude and it’s part of the public vocabulary,” but it’s shaking intensity that tends to cause most damage.
Ultimately, Jumpstart will use shaking intensity as is measurement, which means pulling in more data than just seismic readouts.
“There are all sort of interesting human-based sensors of an earthquake, like Twitter,” she says.
Automation hits the big time
Usage-based auto insurer Metromile has been working intensely over the past year or so to automate as much of the claims process as possible. The company announced several claims enhancements on the back end last December, and this week rolled out a more front-end oriented automation project, the AI claims bot Ava.
When a customer reports a claim through Metromile’s app, Ava accesses the driver’s Metromile Pulse device to reconstruct the accident and prompts the user with relevant questions to adjust the claim quickly. Ava requests details and pictures of damage, which the customer can take themselves. Simpler claims can result in payment within an hour, says Metromile CEO Dan Preston.
“From the customer perspective, it's helping you manage your claim and your insurance experience overall,” Preston says. “On the back end it's like a claims adjuster. It's going through every part of our tech stack and leveraging machine learning on the back end.”
Metromile’s claims department is actively “training” Ava to handle more processes. If the AI hits a snag, it kicks the claim to a human, who takes the correct action, which Ava can use the next time a similar situation comes up. In that way, the company is modeling the software after a human, Preston says.
“Right now she handles a lot of the very high frequency claims that we see a lot of,” he explains. “There are certain edge cases and things Ava hasn't seen yet that pass over to a human. Eventually we will be able to turn those into automated tasks.”
