ai concept insurance
Image: Shutterstock

The Covid-19 pandemic spurred on a shift in digital innovation in the insurance industry. In a 2020 KMPG study, 85% of insurance CEOs said that Covid had accelerated the digitisation of their operations, while 78% said it had turbo-charged progress on the creation of a seamless digital customer experience.

One significant area of development within this has been automation, with machine learning helping insurance companies to streamline CX.

To learn more about the trends in this industry, I spoke with Christophe Bourguignat, the CEO and Founder of Zelros, an AI platform that helps insurers offer personalised product recommendations.

Zelros focuses on the distribution side of insurance, i.e. matching the policy holder with the right cover, depending on their specific lifestyle or needs.

“We do that by using data…” says Bourguignat. “We leverage the data that insurers already own, of course with the consent of the policy holders, and with the security that is required to manipulate that data. We also enrich this with external data – more and more governments are giving access to what they call open data, which is statistical information about weather or natural catastrophes.”

Transformation at incumbents accelerated during the pandemic

I asked Bourguignat how the Covid-19 pandemic impacted both Zelros, and the wider insurance industry.

“I see two effects,” he said. “First, people started realising that society is more and more shocked by crisis – economic crisis, climate crisis – and I think insurance will play a huge role in absorbing the shocks of those crises. People are now understanding why insurance in any domain is crucial for the stability of our planet tomorrow.”

“The second effect is of course the purchase habits of consumers – the fact that, during the pandemic, agencies were closed and face-to-face meetings were impossible. So it has deployed new ways of purchasing insurance, through virtual meetings, or direct-to-consumer and digital channels. We have seen an acceleration of what we call incumbents speed up their transformation.”

Indeed, automation is one major part of this transformation, with artificial intelligence playing an increasing role in the industry, “with policies being priced, purchased, and bound in near real time,” states McKinsey’s ‘Insurance 2030’ report. “Enough information is known about individual behaviour, with AI algorithms creating risk profiles, so that cycle times for completing the purchase of an auto, commercial, or life policy will be reduced to minutes or even seconds.”

AI, Machine Learning and Predictive Analytics Webinar

Amazon sets the bar for CX

So, what are the over-arching benefits of this technology? Bourguignat refers to speed and ease.

“The new standard is purchasing on Amazon, so insurers are also trying to make the experience as fluid as this. So, at any moment of the policy holder’s life cycle – maybe during subscription or management of a claim – it radically accelerates [the speed of interactions] by 10 or 20 or maybe even 50 times.”

“Regarding the insurer, [artificial intelligence] enables them to better assess the risk (helping them be more profitable in the long term). Also, it helps them gain productivity and efficiency, which again is also for the benefit of the policy holder, because it avoids extra costs that are put on the premium.”

AI meets a slow-moving industry with legacy IT systems

While the pandemic has sped up digitisation, Bourguignat says that there is still a long way to go.

“We are speaking a lot about technology, but [transformation] is also about change management in organisations, so it’s also about human beings,” he states.

“Insurance is a very traditional and I would say very slow-moving industry, so one of the first challenges is to raise awareness about what artificial intelligence is, about how it works, and what the limits are.”

“The second challenge is to do with the legacy IT systems. Again, traditional leading insurers have systems that have maybe existed for decades and evolving them can be a very long and painful process. But we have seen clearly, since maybe three or four years ago, an acceleration in the maturity of using technologies in the evolution of these IT systems.”

Humans + machines

In EY’s 2022 Global Insurance Outlook , the authors state that “The consensus among forward-looking executives is that human talent is every bit as important to future success as AI, machine learning and modernised processing platforms.”

Indeed, this seems to be the consensus among insurance customers themselves, who seemingly still want a human (as well as increasingly hybrid) experience. A recent study by Zelros – involving a survey of over 1000 people in the US, Canada, UK, France and Germany – found that 46.9% of people would prefer a completely digital experience when purchasing insurance, yet 30.1% still would like a physical experience with an agent to provide personalised policy recommendations. This means that 23% would prefer a hybrid experience, combining the digital experience with the option to speak to an agent.

Bourguignat, too, agrees that human interaction still plays a vital role in the industry – and not only for the benefit of the consumer.

“I think the error would be to create insurers that are only 100% digital,” he said. “We see some examples that show their limitations, not only because it costs a lot in terms of digital acquisition, but it’s also difficult to address particular cases or to up-sell or make complimentary sales. So, I believe that human agents will keep on existing, and the winning players will be those who focus on both digital and physical channels.”

Human agents also play a role in ensuring inclusivity, tackling complex or speciality cases. “Thinking about, for example, life insurance or accident insurance, which are topics that, as a consumer, you tend to want real advice from a real person. That’s another thing we are working on a lot with the machine learning, is how we ensure the algorithms we put in place on the automated channels are fair and are treating consumers equally.”

Bionic’s third age of CX: hybridising digital and world-class human service

The goal of customer-centricity

“I believe it is about placing the policy holder at the centre of the conversations, instead of the product itself. Part of this journey has already been made but there is still some progress to go,” declares Bourguignat.

He adds that silos remain an issue. “There are some specialists in home insurance or car insurance, but as a policy holder you want to have a 360 offer, and one that covers all your needs. Becoming more customer-centric and avoiding product silos is definitely something to improve on in the coming years.”

In terms of the evolution of insurance overall – and one of the biggest outcomes of increased machine learning – is the growing shift towards a more proactive industry, moving towards prevention rather than pay-out. Bourguignat points out how this aligns with what is likely to be the biggest risk for policy holders in years to come.

“I think everything regarding climate incidents – fire, flooding, hailstorms and so on – will be a huge topic for insurers in terms of prevention and bridging the increasing gap between the real risk that is rising and the lack of coverage of the population.”

As well as consumer uptake, Bourguignat says that transformation within insurance has only really just begun. “Digital acceleration will not stop – it will only amplify.”

Quick Guide to AI, Machine Learning and Predictive Analytics