June 18, 2020 by Greg Meckbach
Brokerages who use artificial intelligence could find opportunities to upsell based on changes in a client’s lifestyle, according to a software vendor executive.
The more data you feed a machine learning model and the more you train it, the better it gets, said Kevin Deveau, managing director of FICO Canada, part of San Jose, Calif.-based Fair Isaac Corp., in a recent interview.
Artificial intelligence (AI) is when technology mimics human cognition such as learning from experience, identifying patterns and deriving insights, said Mark Breading, a partner with Boston-based Strategy Meets Action. Machine learning is a type of AI in which computers act without being explicitly programmed, SAS Institute Inc. notes.
Bigger brokerages with enough money to invest in AI and machine learning could use those technologies to build a “360-degree view” of a customer, said Deveau, in the context of how the COVID-19 pandemic is forcing companies to change the way they operate.
“The pandemic is going to show them they need to be more reactive in the digital space.”
Non-insurance examples of machine learning include receiving recommendations for content to watch on Netflix and e-mail servers that flag messages as spam, Jeffrey Baer, assistant vice president of advanced analytics and business intelligence at Economical Insurance, told Canadian Underwriter earlier.
Using AI, brokers could, for example, track lifestyle changes such as buying a house, getting married and having children, said Deveau.
“The broker could be more fine-tuned to when that event happens to the end customer,” he explained. “‘I see you just bought a house. Do you have mortgage insurance?’ Or, ‘Do you want some additional life insurance?’ With the birth of a child, ‘Do you want to build up an RESP?’’
For most brokerages, this process tends to be manual, said Deveau.
“By incorporating predictive analytics and machine learning, they could be much more reactive and more customer service oriented.”
AI allows insurers to cut cost, reduce risk and generate better customer insight, he said in an interview. They can also use AI to read through policy documents and help detect fraud.
If a first notice of loss comes in, an insurer could assign it a “fraud score” based on the data and how likely it is that the claim is fraudulent. At that point, the insurer could decide to expedite the claim (if there is low fraud risk) or instead send it to a special investigation unit, said Deveau.
Some insurers are using artificial intelligence to conduct policy reviews, Stéphane Lespérance, president of commercial risk and health solutions at Aon Canada, told Canadian Underwriter in an earlier interview.
“Essentially, you compare the wording from a year to another before renewing it. The document would then be submitted to a human to conduct further review. So, it just eliminates maybe one step and makes the process easier. I don’t think we are in the mode right now that the AI takes care of everything,” Lespérance said at the time.
Some insurers are using AI for “virtual assistants,” said Ralph Chapman, vice president, financial services At IBM Canada
“In that scenario, you are taking – at least initially – the personal interaction out of the process,” Chapman told Canadian Underwriter in an earlier interview when asked what major technology trends he foresees in the property and casualty insurance industry in 2020.
“Your clients can ask questions, file a notice of loss, check on payments, claims status, get quotes — all of the things that would normally involve human interaction and taking a lot of time from start to finish. These processes are becoming automated because insurance companies will have the ability to leverage these cognitive platforms.”