Racing to im­prove their cus­tomers’ ex­pe­ri­ence, in­sur­ers are turn­ing to AI and pre­dic­tive an­a­lyt­ics.

Digital Insurance - - CLAIMS - By El­liot M. Kass

It’s been a long-time com­ing, but dig­i­tal trans­for­ma­tion is tak­ing the in­sur­ance in­dus­try by storm. And nowhere is this more ap­par­ent than with how claims are get­ting pro­cessed.

As other in­dus­tries have set new stan­dards for cus­tomer ser­vice—com­pa­nies like Ama­zon and Ap­ple quickly leap to mind—to­day’s dig­i­tally-im­mersed con­sumers have grown ac­cus­tomed to do­ing busi­ness any­where, at any time and with any de­vice. When they file an in­sur­ance claim, those ex­pec­ta­tions are in play and they eval­u­ate their in­sur­ers ac­cord­ingly. That leaves car­ri­ers racing to en­sure a pos­i­tive claims ex­pe­ri­ence with a whole new port­fo­lio of dig­i­tal ca­pa­bil­i­ties. “There’s an im­per­a­tive for in­sur­ers to im­ple­ment a dig­i­tal claims strat­egy,” says Martina Con­lon, se­nior vice pres­i­dent for re­search and con­sult­ing at No­var­ica, an ad­vi­sory firm for the in­sur­ance in­dus­try. “They need to do this to demon­strate their value to their cus­tomers.”

The foun­da­tion of this strat­egy is a mod­ern, flex­i­ble claims sys­tem with an ar­chi­tec­ture that al­lows the in­surer to make full use of other emerg­ing tech­nolo­gies, says Con­lon. Chief among these are ar­ti­fi­cial in­tel­li­gence and pre­dic­tive an­a­lyt­ics. Con­lon points to two ma­jor sets of ap­pli­ca­tions for each.

The first is pre­dic­tive model­ing, which is widely be­ing used for fraud de­tec­tion, de­ter­min­ing when sub­ro­ga­tion is ap­pro­pri­ate and whether lit­i­ga­tion should be pur­sued or set­tled. Other com­mon pre­dic­tive model­ing ap­pli­ca­tions in­clude es­ti­mat­ing the sever­ity and fre­quency of dif­fer­ent types of claims and im­prov­ing an in­surer’s op­er­a­tional ef­fi­cien­cies.

The sec­ond wide­spread use of AI and an­a­lyt­ics is straight-through pro­cess­ing, which makes use of rules en­gines, third-party data and pre­dic­tive mod­els to clas­sify and score claims for sever­ity and prob­a­ble out­comes. To­day, a sig­nif­i­cant number of in­sur­ers make use of STP for very sim­ple claims like car tow­ing and glass re­place­ment, but Con­lon pre­dicts its scope will rapidly ex­pand as in­sur­ers’ abil­ity to gather claims data and au­to­mate the claims process con­tin­ues to im­prove.

The fu­ture of claims data gath­er­ing is quite lit­er­ally on the hori­zon. Drones are giv­ing in­sur­ers a po­tent new data col­lec­tion tool that will only get more pow­er­ful over time, as en­hance­ments like ma­chine vi­sion—a vari­ant of ma­chine learn­ing that helps a drone bet­ter in­ter­pret what it sees as it learns from ex­pe­ri­ence—come into their own.

And while fleets of drones serve as the in­sur­ance in­dus­try’s data gath­er­ing air force, on the ground in­sur­ers are mov­ing to de­ploy an army of data gath­er­ing de­vices also known as the In­ter­net of Things.

In terms of auto claims, for ex­am­ple, “We are rapidly mov­ing to a fu­ture where there

“There’s an im­per­a­tive for in­sur­ers to im­ple­ment a dig­i­tal claims strat­egy. They need to do this to demon­strate their value to their cus­tomers.” “If you look at a lot of this claims tech­nol­ogy, it puts peo­ple at the cen­ter of the claims in­ter­ac­tion, while the tech­nol­ogy takes care of a lot of the noise.” Karen Furtado, Part­ner, Strat­egy Meets Ac­tion Martina Con­lon, Se­nior Vice Pres­i­dent for Re­search and Con­sult­ing, No­var­ica

are sen­sors lo­cated through­out the car, and when an ac­ci­dent occurs, be­fore the driver can even get out of the car, the in­sur­ance com­pany will have been called, an email with data about the ac­ci­dent will have been sent to the ad­juster, and emer­gency ser­vices will have been con­tacted with GPS in­for­ma­tion about the lo­ca­tion of the ac­ci­dent,” ex­plains No­var­ica vice pres­i­dent Keith Ray­mond. “En­abled by the In­ter­net of Things, this will all hap­pen in an in­stant.”

The All­state Ex­pe­ri­ence

Many of these tech­nolo­gies are al­ready in play at com­pa­nies like All­state and Farm­ers In­sur­ance, which four years ago in­tro­duced smart­phone apps that al­low pol­i­cy­hold­ers to set­tle sim­ple fen­der-ben­der claims with­out vis­it­ing an ad­juster or get­ting an es­ti­mate from a re­pair shop.

All­state’s app, known as Quick­Foto Claim, al­lows pol­i­cy­hold­ers to use their smart­phone to send pic­tures of their dam­aged ve­hi­cle to the in­surer to have their claim ad­ju­di­cated. Di­rec­tions on their phone guide them through the process, in­clud­ing how to take pic­tures of the dam­age and sub­mit them. Fraud de­ter­rence fea­tures pre­vent the user from doc­tor­ing a photo.

Af­ter the pho­tos are sub­mit­ted, a writ­ten es­ti­mate is sent to the user elec­tron­i­cally. All­state says a pol­i­cy­holder can ex­pect to re­ceive an es­ti­mate and a phone call from the in­surer to dis­cuss it within 24 hours.

The Farm­ers app, called EZ Es­ti­mate, is quite sim­i­lar. Af­ter up­load­ing pho­tos of the dam­aged ve­hi­cle, a re­pair es­ti­mate is sub­mit­ted to the cus­tomer, usu­ally within three busi­ness hours. If the pol­i­cy­holder ac­cepts the es­ti­mate, the claims process con­tin­ues via the app un­til the claim pay­ment is trans­ferred elec­tron­i­cally to the cus­tomer’s bank ac­count.

A “call for help” but­ton is avail­able on ev­ery screen of the app, and if cus­tomers ex­pe­ri­ence prob­lems while us­ing the app, Farm­ers says they can meet with a claims ad­juster to con­clude the process.

At All­state, Quick­Foto now pro­cesses about half of the in­surer’s claims for ve­hi­cles that can still be driven post-ac­ci­dent, which has led to the in­surer shut­ting down many of its drive-in claim cen­ters na­tion­wide.

“We now have our ad­justers look­ing at en­hanced pho­tos, dig­i­tal pho­tos with­out hav­ing to drive to and from the sites,” All­state CEO Tom Wil­son told a group of in­vest­ment an­a­lysts eight months ago.

“In the past, there was a lot of dead time, un­pro­duc­tive time, as ad­justers drove around, driv­ing to cars, driv­ing to body shops,” he con­tin­ued. “And we looked at that and re­al­ized that emerg­ing tech­nolo­gies, data and an­a­lyt­ics could rec­tify that and take some of the in­ef­fi­ciency out of the sys­tem.” Ad­justers aren’t the only ones who are ben­e­fit­ting. The con­ve­nience of claims self-re­port­ing apps is also win­ning high marks from cus­tomers, says Karen Furtado, a part­ner with mar­ket re­searcher Strat­egy Meets Ac­tion.

“Are you kid­ding?” says Furtado. “Who wants to take a half-hour out of their day to get their car looked at, be­cause it was hit by rock?” But the real boon to cus­tomer sat­is­fac­tion, she adds, is how trans­par­ent apps like Quick­Foto make the claims process.

“It’s no longer, ‘I called this per­son and some mag­i­cal mys­tery thing hap­pens in the claims depart­ment. Now I know ex­actly what’s hap­pen­ing with my claim.’”

Other AI-driven apps, she notes, are also help­ing in­sur­ers for­tify their cus­tomers’ claims ex­pe­ri­ence. Chat­bots, for in­stance, are au­tomat­ing cus­tomer no­ti­fi­ca­tions, al­low­ing in­sur­ers to keep claimants more closely in­formed about the sta­tus of their ve­hi­cle re­pair, when they can ex­pect to re­ceive a check for the dam­ages and other mat­ters.

“If you look at a lot of this claims tech­nol­ogy,” says Furtado, “it puts peo­ple at the cen­ter of the claims in­ter­ac­tion, while the tech­nol­ogy takes care of a lot of the noise.”

But for all their progress, the ex­perts agree that in­sur­ers are just at the tip of the ice­berg of what AI and pre­dic­tive an­a­lyt­ics hold in store. And as these tech­nolo­gies ma­ture and move front and cen­ter, “The fu­ture of claims pro­cess­ing will be very dif­fer­ent from what we have to­day,” says No­var­ica’s Con­lon. “In­sur­ers,” she adds, “are al­ready headed very rapidly in this di­rec­tion.”

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.