Data science and AI is driving value across all industry, but its vast potential within motor claims was underlined during a keynote presentation to delegates at the inaugural Motor Claims Showcase event.
Delivered by ARC360 and ILC at the CBS Arena, Coventry on 29 June, the event sought to showcase best practice across the motor claims supply chain through panel discussions, partner showcases and interactive displays from 30-plus exhibitors.
Kicking off the agenda on the main stage, Pierre du Toit, Chief Data Science Officer at Abacai, delivered a keynote address in which he revealed that more data is now being generated every few days than all the data that has been created since the beginning of time.
Every minute of every day:
- 167 million videos are watched on TikTok
- 694 million songs are streamed in America
- 231 million emails are sent
- 6 million people are shopping online
- 2 million Snapchats are sent.
Pierre said, “This is truly the data age, and buried in the data are answers to some of society’s biggest questions. It is already improving things around education, health and transport, and it can do the same for insurance and claims.”
He explained how algorithms are now becoming so sophisticated that they can predict things never previously thought possible, taking real-time data from multiple sources and adding it to reams of historical data to not only understand what is happening, but what is likely to happen next.
“What excites me most is the ability of algorithms and data science to personalise user experience,” Pierre said. “This is when the data we generate is used to develop sophisticated algorithms that are then imbedded it into the technology we use on a daily basis to drive value for the user and the business.”
Insurance
He believes this is an area the insurance industry can exploit, and explained how Abacai, a claims management business and motor insurer selling traditional and temporary policies, is already using its own data plus data from outside sources to drive business decisions and process.
He said that this can be divided into three categories: supervised learning (will someone make a claim and how will that claim progress); unsupervised learning (considering customer behaviour according to geography, age, demographic etc); and reinforcement learning which optimises decision-making.
Pierre said, “Using all three together gives us the most accurate type of outcomes and prediction across distribution, risk management, customer service, claims management and customer retention.”
Claims
In terms of claims, he identified three areas where the application of data science and AI can benefit both the insurer and the policyholder.
The first of these is in first notification of loss. He accepted that claims handlers will still be critical in the process but said their lives and the lives of customers can be made easier by introducing automation where possible. He said it was possible to develop and embed bespoke Q&As and use both current and historical data to provide answers, giving customers 24/7 service.
He said, “Among customers that are able to self-serve like this, we have found a 25% reduction in calls coming through to our call centres. This has already led to a reduction in servicing-related costs and overheads.”
AI can also accelerate and streamline the claims process by predicting next steps.
Pierre said, “We look at the data we’ve captured from previous claims and develop algorithms to enable instant decisions around what action should be taken next. It helps us determine liability, damage severity, whether the vehicle can be repaired and, if so, by whom. Our algorithms are also able to provide an accurate probability of that decision, and we have set rules-based parameters around this.”
Finally, he believes that harvesting and interpreting data can help manage risk and identify fraud. Pierre explained that vast amounts of key data is immediately available as soon as a claim is made: who the claimant is, when did they take their policy out, was the claim made in a fraud hotspot, impact damage from vehicles and third-party information. Sophisticated algorithms can instantly interpret all that data and consider the likelihood of the claim being fraudulent.
Opportunity
However, Pierre said that as advanced as AI and data science already is, it is only just the start and if used correctly then insurers and customers can create a virtuous circle of continuous improvement.
He said, “The insurance industry isn’t renowned for the trust customers have in it. We can change that by using the data shared with us in an ethical and sustainable way to drive value and improve customer experience. That will drive trust and loyalty, encourage customers to share more data and that in turn will help us improve the algorithms again and drive even more value.”
The ARC360/ILC Motor Claims Showcase Event was headline sponsored by Enterprise Rent-A-Car, along with fellow sponsors EDAM Group, Control Expert and Procurato.
ILC would like to thank its motor Corporate Partners: AkzoNobel; Audatex; Autoglass; CAPS; Carpenters Group; Copart; Davies Group; e2e; Entegral; Enterprise Rent-A-Car; Gemini ARC; GT Motive; The Green Parts Specialists; IAA; Innovation Group; S&G Response; Sherwin Williams; and thingco.
ARC360 would like to thank its Corporate Partners: Audatex; BASF; BMS; CAPS; Copart; EMACS; Entegral; Enterprise Rent-A-Car; Innovation Group; Mirka; NWVA; S&G Response; and Sherwin Williams, as well as Partners asTech; The Green Parts Specialists; Indasa; and Prasco UK.
