Blog

How to Implement Intelligent Automation So It Truly Works In Underwriting

At the end of September we hosted a webinar in partnership with Insider Engage on the topic of implementing intelligent automation in underwriting so that it truly works and provides value. Matt Scott of Insider Engage moderated the discussion and has also written a write-up of the webinar for us.

You can find the original version of this article published here on the Insider Engage website.

Want to tune into the webinar? You can view it on-demand here.

Automate to Deliberate: How to Spend More Time on Your Underwriting Decisions

By Matt Scott of Insider Engage

The importance of underwriting in the insurance value chain cannot be overstated, but despite this importance, there are several pain points that continue to dog the industry.

Underwriting is the very foundation of the insurance industry, meaning that optimizing the underwriting process is crucial to success in a highly competitive market, and intelligent automation, with its power to create efficiencies and remove process bottlenecks, is seen by many as the solution to this optimization problem – but maybe not in the way you think.

The importance of underwriting in the insurance value chain cannot be overstated, but despite this importance, there are several pain points that continue to dog the industry.

Many of these pain points revolve around data – whether it be compilation, analysis, input, or manual rekeying – and without the help of modern technology, these processes can be slow and painful for everyone involved.

The good news, however, is that intelligent automation has the ability to automate these repetitive tasks, eliminate rekeying, alleviate bottlenecks, and free up time for higher-priority work.

Slow and Steady Wins the Underwriting Race

But the key to realizing these benefits, according to Eigen Technologies principal product manager Chris Mullan, is not to simply speed up processes and replace the role of an underwriter with automation. It is to enable underwriters to do their job better, even if that means slowing them down.

“Intelligent automation can only really work if it makes the underwriters themselves slower, smarter and better at their jobs,” he said on a recent Insider Engage webinar, hosted in partnership with Eigen Technologies. “It can’t help if it is simply viewed as something that can do the underwriter’s job for them.”

Now, it may sound odd that automation is most effective when it makes people slower, but if you listen to Mullan’s argument it all begins to make sense.

Mullan argued that the traditional way insurers have tried to achieve scale has been by asking everyone to move faster and “cram more stuff down the pipe”. The problem with this is that it can lead to people relying on “cognitive biases, heuristics and judgment calls that are just reactionary.”

Instead, Mullan said that intelligent automation should be used to remove a lot of the grunt work associated with underwriting, giving the humans the option to slow down and engage this more analytical side of their brain.

“If we use intelligent automation to make all the surrounding work go away, we allow our underwriters to take more time on the actual decision and do what they are actually employed to do,” he said. “Then they can use some of that Kahneman Thinking Fast and Slow so they can get into that 'system two' brain and be more thoughtful, more analytical and more considered.

“That's how you're going to ultimately get the real commercial impact of using intelligent automation.”

Meanwhile Mark Patterson, a partner at Deloitte, said intelligent automation should be all about forging a competitive advantage in a marketplace where it is notoriously difficult to stay ahead of the game.

“It’s about finding out how to out-underwrite the competition,” he said. “We operate in a very competitive market and if you are able to leverage that technology to pull out unique insights [into a risk] in order to change your underwriting decision, while your competitor is still on page 12 of a report reading line-by-line, that is how you are going to get a really differentiated result.”

When it comes to technology, however, the proof is in the pudding, and that is why it was great to hear from Nanditha Nandy, Head Analytics Impact+ L&H Global at Swiss Re.

The (re)insurer has been using intelligent automation for several years now, and in a variety of different ways.

Nandy said that this includes scanning emails to pre-populate and rank submission documents – including auto-declining risks outside of appetite – and automating underwriting decisions through a highly sophisticated rules engine.

Nandy added that this had created a number of efficiencies for the business.

“Intelligent automation can make underwriters much more efficient, giving them more time to take up increased volumes of business and be more competitive in the market,” she said.

Sometimes, however, the route to these benefits does not always run smooth. So what do you do when teams or individuals are putting roadblocks in the way of incorporating intelligent automation into the underwriting process?

Guenter Kryszon, Markel’s global executive underwriting officer for the property markets, knows this problem all too well.

“You sometimes get the view from underwriters or regional management that they know their portfolio better, because maybe they had drinks with XYZ Brokers the night before and they promised them a great opportunity,” he said. “Meanwhile, we know that broker doesn’t produce the right opportunities.

“We might know the profitability of that broker, for example. We also know where some of the buying signals are coming from a hit ratio when we get these opportunities.”

To combat this, Kryszon said it is all about communicating the information in a clear and concise manner, to make the benefits easily accessible and as obvious as possible.

“Being able to filter that information and provide a curated view to the underwriters is what we’ve been really focused on at Markel,” he said. “My vision for the underwriter desk is that they should be able to log in and already have the top ten individual submissions that I want them to work on, on any given day.

“And those submissions will be the ones that have the highest yield with the highest hit ratios that are going to produce the best return for Markel.”

Watch the webinar now

If you enjoyed Matt Scott's webinar synopsis, you can check out the on-demand version in its entirety here.