Driving Value from AI in Insurance: Delivering Success with Sun Life and Scotiabank
Much discussed, yet often misunderstood, it’s clear that Artificial Intelligence is set to dominate the global Insurance sector for the foreseeable future – and Canada is no different. Despite being a somewhat nascent discipline within insurance, progress in silicon centres across the nation mean carriers recognize the potential of AI to cut costs, improve the customer experience and generate sustainable long-term growth.
Although currently centred around Toronto and Montreal, new government and private sector initiatives are seeking to expand Canada’s AI capabilities from the ground up, but with specialist talent in short supply, a skills gap is likely to exist in the short-term. From tapping into the burgeoning community of start-ups and innovators to incubating their own AI specialists, carriers are finding there is no quick-fix or easy route to success. A clear vision for the future, and an acceptance that not all initiatives will succeed, are required.
Insurance Nexus spoke to two eminent, senior Canadian insurance executives, Inka Bari, AVP Analytics, Group Benefits, Sun Life and Fernando Moreira, SVP, Global Insurance, Scotiabank, to get their views on how Canada’s insurance market should approach AI integration and where it can expect to garner the best results.
The first priority may seem obvious but is worth repeating: do not leap into a technical solution without knowing what problem you’re trying to solve. Admittedly, this sounds simple, but many carriers still struggle with this first aspect, often under competitive pressure to bring a new capability to market; “That’s the big question,” says Inka Bari, AVP Analytics, Group Benefits, Sun Life. “Are people doing it because they have a problem to solve, or because it’s cool and competitors are doing it?” she asks. Fernando Moreira, SVP, Global Insurance, Scotiabank, agrees with these sentiments: “Whether it’s for solving complex business problems or addressing growth opportunities, we need to understand what sort of proposition we should be offering our customers.”
Once a need and solution are identified, implementation is not necessarily straightforward, however. For large, legacy carriers, issues typically arise because internal organization/operations were simply not designed for the demands of modern technology and innovation, with “large organizations like Sun Life…as old as the country itself,” Bari notes. “There is so much legacy that has been adapted to behave in a way that wasn’t originally planned”
Once a need and solution are identified, implementation is not necessarily straightforward, however. For large, legacy carriers, issues typically arise because “there is so much legacy that has been adapted to behave in a way that wasn’t originally planned,” Bari notes. With “large organizations like Sun Life…[that] are as old as the country itself,” internal operations were simply not designed for the demands of modern technology and innovation, and so ensuring these processes are fit to support a sophisticated AI and analytics infrastructure is imperative.
Increased customer-centricity means new propositions must make a difference to customers – competitive pressure and the availability of on-demand, personalized insurance elsewhere means customers will simply go elsewhere if not. But, as competition is high, loyalty is not. “Insurers who want to increase market share, as well as attract and retain customers, need to shift focus and improve in areas that most influence customer satisfaction,” said Tom Super, Director of the Insurance Practice at J.D. Power. “Highly satisfied customers have better policy lifetime expectancy (PLE) scores and are much more likely to recommend their insurer to friends or family.”
Whereas previously, customer-focused initiatives may have focused on lower call-centre waiting times or better online self-service (both valuable components of any customer experience), advances in technology and the capabilities of competitors mean carriers increasingly need to show they are proactive in keeping their customers safe. While it works to lower claim sizes for the carrier, the competitive marketplace also means customers want more for their insurance premium than help in a claim – an aspect of the market that is driving the insurer/customer relationship towards prevention of loss, rather than reaction and restitution.
The perfect proposition is meaningless, however, without the technical knowledge required to put it onto effect. Element AI noted in its 2018 talent report that there were only 20,000 AI experts on the planet and even if they all turned up in Canada, they would need to use all of them, plus every student graduating in an AI discipline from University. “AI is not built by technology; it’s built by people who know how to find the right technology. It’s not cheap. You have to invest to pay for the right people. They are very in demand,” Moreira admits.
To address the shortfall, a number of large organizations have been looking for ways to incubate their own AI specialists; Manulife’s four-year, $400,000 CAD partnership with the Waterloo Artificial Intelligence Institute at the University of Waterloo, focusing on research into disability claim prediction and fraud prevention, for example. Intact Financial is taking a similar route, announcing a partnership with Université Laval in January 20197 that will see the company sponsor two research chairs focusing on AI and actuarial science.
Despite these efforts, there are many potential pitfalls; AI is a data-hungry discipline and without proper checks and balances in place, a breach could deal a fatal blow to consumer trust – vital above all else in the insurance industry. Equally, there are many opportunities to choose from, but a focus on customer need and business prerogative is essential, as is a fail-fast, experimental attitude towards innovation. AI is never going to be a quick fix. As a technology, it has the potential to reinvent insurance as we currently understand it; products can become more relevant to the individual, the claims process can be condensed from weeks to a matter of seconds in some cases, while some risks can even be avoided altogether. But it is also an insurance event horizon. Customer expectation has been set, potential has been revealed and carriers who fail to adapt must inevitably find themselves obsolete.