Get a demo
Select the type of demo you'd like to experience with Indico Data.
Select the type of demo you'd like to experience with Indico Data.
Watch Indico Data CEO Tom Wilde step in as co-host alongside Michelle Gouveia, VP at Sandbox Insurtech Ventures, in season 2 episode 11 of Unstructured Unlocked with Rick Russell, retired VP of underwriting for Chubb and underwriting director at Nationwide.
Michelle Gouveia:
Hey everybody. Welcome to a new episode of Unstructured Unlocked. I’m co-host Michelle Gouveia.
Tom Wilde:
And I’m Tom Wilde,
MG:
And we are joined today by special guest Rick Russell, a retired vice president of underwriting from Chubb. Hey Rick, thanks for joining us.
Rick Russell:
Hey, thank you very much. I’m glad to be here.
MG:
We’re excited to have you. I’m very much looking forward to the conversation today. Maybe to kick things off, can you share a little bit about your background and your vast experience in the insurance industry?
RR:
Absolutely. I’m one of those unique few that actually chose insurance as a career unicorn.
MG:
We don’t get a lot of those.
RR:
I spent 37 years in the industry, two in life and health and then 35 in property and casualty all on the company side, commercial underwriting and product. I joined Allied Insurance who was purchased by Nationwide Insurance in 1998. Spent 30 years with that organization and then spent another five with Chubb Insurance. The last two years I’ve been retired and performing consultant work.
TW:
Excellent. What are some of the things that remain true today that were true 25 years ago, and then what are a couple things you would highlight that are dramatically different today than they were 25 years ago? That kind of compare and contrast is interesting, especially because you’ve had sort of a long lens on this.
RR:
The one thing that I think has remained true over the years is underwriting discipline, underwriting integrity, and that’s always been important. Maintaining that balance between coverage and pricing. Insurance is a cyclical industry and so you have areas or you have times when you have a soft market where coverages are expanding and pricing is reducing and you have hard markets where the industry tightens up and there’s less coverage, maybe less markets for customers to choose as well as increased pricing. I think that will continue. That’s just the nature of the beast and the industry that we work in. I think the biggest changes have been in automation and really the efficiencies that you can gain throughout that. 30 years ago we shuffled a lot of paper over the years we moved to document handling systems, but maybe they weren’t, in some ways they weren’t as efficient as paper and created other challenges, but I think as we moved into the future and the automation that exists, embracing that is going to be the big game changer for everybody involved in the industry.
TW:
I think I remember, it may have been a conversation you and I had, I can’t remember if it was or not, but there was a description someone gave me of a couple decades ago, the underwriter would print out all of the documents related to a piece of risk and put them on the desk and make sense of them synthesize to come up with a view on the risk and ultimately the pricing. So the sort of activity of synthesizing that was true then and is probably true now, right? It’s just a question of the tools you can use now to do that versus back then,
RR:
Right? Absolutely. In the day when I began in the industry, yeah, we received mail and you would sort those applications maybe in the order of coverage that you wanted to underwrite those today, I think underwriters still do that to a certain extent. It’s just in a different environment, in a different format.
MG:
I think probably too, one of the big differences is the speed of access to additional information. Going back to Tom’s example, if you had that 50 page stack of a submission and there was information missing, you probably have to mail something back to the applicant and then wait one week or two weeks or three weeks to get that back to complete that. Whereas today there’s obviously data vendors and partnerships and API connectivity that allows you to fill in those gaps kind of on the spot. Sometimes it enhances, sometimes it might make it more challenging, but back to that automation thing, it’s just an amount of information at your fingertips that’s probably makes it a little bit faster at least to get it to the next phase.
RR:
It does, and I think the other benefit that exists today is that if you have one database for underwriting guidelines, for other underwriting criteria, underwriting reference material, it’s all in one place for the underwriters to use. There’s no more updating a paper manual, so that is also a benefit that exists today that didn’t exist then. You never knew if the underwriter had correct information that they were using, but they do today.
TW:
One thing that strikes me is, and this kind of builds on Michelle’s question, if the cycle time for understanding the insured’s desired coverage because you can basically receive ask information very quickly in a very tight loop because it’s almost all electronic, number one and number two, as an industry, if the carriers are closer to sort of the nirvana perfect data to understand the risk, do you think that will ultimately attenuate the cyclicality of the market between hard market and soft market? It’s always, when I’ve talked to people, it’s not clear why the market inflex from hard market to soft market and back again, even though it is a cycle, but why is a little bit unknown. Do you think that as we get closer to real time and robust information, does that attenuate that or is it still going to exist anyways?
RR:
It probably will exist to some extent, but I think having the data of your existing book, all your renewal, everything that you’re currently writing, having that data and being able to analyze that and take action as far as speed to market, I think that will help level out some of the valleys and the mountains in the cycle. It’s just being able to take that underwriting action as quickly as you can on your portfolio business and level out your return on investment as you achieve your goals. Doing that.
MG:
Rick, between your experience at Nationwide and then at Chubb, I’m curious to get your thoughts on maybe the evolution of changes or what differs between what you have to consider in the underwriting of personal lines insurance versus commercial insurance or specialized commercial insurance even deeper than that?
RR:
That’s a great question. There is such a world difference between personal lines and small commercial and then large commercial personal lines is a homogenous product just like insurance is designed and it is a fairly simple product and I think with that you can automate that and manage that book of business a lot easier than you can commercial business. Now, your commercial business owners policies, a lot of those are going to be simple accounts, simple risks, maybe just one location or two and two or three locations. Those can be managed much in the way that personal lines is managed and companies are doing that today. I think the real challenge that exists today are your larger commercial accounts where you have so much data coming in on Accord applications and you have to process that and massage it and manage it in some way that you can make an informed decision on that account. I think that’s where the challenge really lies today is in that piece of it and having the means to intake that application and gather that data and summarize it, I think will become much more important in the future as it is today.
TW:
Yeah, I think one thing that we’ve heard, and this is something we’re actively working on building into our product, is more real time ability to manage your portfolio of risk. Again, to sort of cut that cycle time down in terms of addressing new submissions, whether it’s, hey, we have too much concentration in this geo or in this category. That was a real eyeopener for me to think about the dynamic aspect of the portfolio management versus, hey, it’s the beginning of the year. Here’s our goals for the year. Here’s our underwriting goals for the year. Here’s our GW P goals for the year. It’s actually possible now to have a quite dynamic portfolio where you can be more aggressive or less aggressive given what’s either the kind of, and I think the ability to close the loop with the losses that you’ve seen versus the policies you’re writing historically, my understanding is that’s a sort of pretty good gap between those two, and that’s shrinking as well.
RR:
And with that gap, assuming you’re writing annual policies, you’re always going to have a one year process if you discontinue a class of business or take some sort of action on a class of business, you’re going to go through that one year cycle to complete that process. But I think having better tools to analyze upfront and then identify those accounts and act on those accounts in an extremely timely manner is of utmost importance. Absolutely.
TW:
Rick, it’s kind of like a strange question maybe, but what types, so maybe on those large commercial size or on the specialty insurance, even in the small commercial space, some of those more complex risks during your tenure, what information or process do you wish existed that didn’t in terms of data or automation or data exchange even
RR:
Most recently in my career I dealt with a lot agriculture and food accounts, and when you look at the Pacific Northwest and you look at some of the fruit accounts that exist up there, they have extremely large property values, hundreds of millions of dollars, and it could all be at one location and trying to manage the accumulation of property exposure at those locations and address that in the appropriate underwriting manner or action, and then getting the right price on that and being able to identify some of the unique exposures of that because if there is a property loss, the downtime and the specialization of some of that machinery, you could have a big business income loss. So being able to identify all those exposures and manage that in an automated process, which I believe can be done fairly effectively because you’re just dealing with property and the values at specific addresses, that would’ve been very valuable and very helpful at that point in time. Yes,
TW:
There’s some real curve balls there. One thing I didn’t know was that a wildfire 10 miles away can ruin a vineyard that the smoke can ruin a vineyard crop. That’s another set of variables to try to figure out when you mentioned crop crop insurance as an example there. So those are some of the variables that I think I sat through a panel with Google where they’re trying to create a digital twin of the earth, which I thought was a really talk about sort of a big audacious goal, but kind of a fascinating one when you think about things like wildfire risks and other type of cat risks
RR:
And whether it’s wildfire or hurricanes, but specifically wildfire, having better tools to map those properties. And there are tools that exist out there to do that
TW:
For sure,
RR:
But being able to interface the information that I receive in as an underwriter on accord applications and then interface that with a wild file tool would be much more effective in the underwriting process and making decisions of which locations would be acceptable or which locations would not be.
MG:
How do you think about the role in AI in some of those use cases that you just talked about and maybe even expand on that and how do you think about where does AI best fit in? Again, contemplating the differences between personal lines products and underwriting versus commercial lines, products and underwriting? Are there different spaces where AI is more impactful or more meaningful to have in the workflow?
RR:
Just from a perspective of, and I’m thinking new business here, the Accord application comes through the door, being able to take that information and put it in a condensed, easy to use format for the underwriter. I think that that is probably one of the biggest changes that could benefit your desk underwriters that are sitting there looking in accounts day in and day out, they could make a decision quicker, and I know we really haven’t talked about this yet, but when we look at what the expectation that brokers have of companies today, we live in a world of immediate gratification. You get immediate responses to everything, and when you think of larger commercial accounts, when those come through the door or a broker will email that to an underwriter and the underwriter gets the call five minutes later, did you receive this and is it something that you will,
TW:
When will I have my quote? Yeah,
RR:
When will I have my quote? Exactly. Instant responses, instant answers, and sometimes the underwriter hasn’t had a chance to look at that account. It will take time to sort through that large account and decipher and get a feel for whether it’s something that they are interested in or not interested in, and I think having a tool that could assist the underwriter in summarizing all that information would definitely be a value.
TW:
Yep. What’s the underwriter’s anxiety about ai?
RR:
I think underwriters, you look to the future and what AI is going to bring to the table and it’s going to look like, and underwriters are generally a stubborn bunch. They have not only their guidelines, their underwriting criteria, they probably are set in their ways a little bit as far as how they like to process things and see things done. And I think the biggest challenge dealing with underwriters is going to to convince them that AI is the way to go, that there is a better tool, and if we can show them that it will make their job easier, I think they’re more apt to embrace that. The one obstacle I think that a lot of underwriters are going to have, especially your seasoned underwriters, they’re so used to looking through the information themselves and determining whether it’s an account they want and then the underwriting process where they want to dot all their i’s and cross all their T’s because they don’t want to have a loss on that account. I mean, that’s ultimately the goal, but your accounts are going to have losses. That’s why we exist, but convincing the underwriter that the information that they’re receiving from whatever AI tool is being used is accurate, that’s going to be a tough sell. You’re going to have to show them the accuracy in order for them to embrace that, and I think that’s big.
TW:
I mean, they’re the ultimate human in the loop, right? At the end of the day, kind of capital H human in the loop compared to the way we think about it in machine learning. Do you think, how does AI change the competitive posture of carriers who are competing for the same risks? Does it fundamentally change at the corporate level how to think about where to build the competitive advantage?
RR:
I believe it does. I mean, it is going to be a competitive advantage in a sense that brokers are going to do business with companies that have the least resistance
To a certain extent, whatever least resistance means. But so many times underwriters will have questions for agents or the company has additional information that they need regardless of what level it is and being able to identify what is needed and get that in through the door to the underwriter and through the process is in a timely manner is going to be a benefit to that organization. And I say that in a sense that if you are a company that still has a processing department or some sort of support function that’s actually keying that information in, that’s going to take a lot longer. I mean, there’s a cost involved with that. It’s going to take a lot longer to complete that quote or that piece of new business for that broker when the companies that embrace AI and can get that through the process in an extremely fast manner is going to have the advantage. And that’s the same as the reverse of looking at a book of business and being able to take something to market or take a change to market quicker.
MG:
Is it similar, you think? So we’ve talked a lot about the dynamics between the carrier and the broker and how AI impacts that. Do you think it’s similar between the carrier and the end insured in that sense? So does the end insured care if their insurance carrier is using AI to automate manual heavy processes or streamlining certain elements of the process?
RR:
From an underwriting perspective, I don’t necessarily believe that the policy holder will recognize that or maybe not even care how it gets done. It’s really just that it gets done. And a great simple example is the addition of an auto to an auto policy. If we can provide that documentation that the policy holder needs on an immediate basis rather than waiting even a couple of weeks to get that thing processed and that’s, that’s going to be a benefit to the policy holder and something that the policy holder will notice, again, it’s just that immediate response.
TW:
Do you think maybe in reverse staying with that, are there opportunities to help the insured and perhaps the broker as well acting as their proxy to better understand the insurance they’ve purchased? Because in commercial lines it can be quite complex, right? And there’s all sorts of exceptions and prohibitions and restrictions and these are often if you’re insuring an airline or a fleet of marine freighters or something, there’s many, many aspects to that policy. Where do you think the frustration lies today and what are some opportunities there to make that better? That’s a problem if you perceive it as a challenge.
RR:
That’s a tough one, Tom, and I’m not sure if this specifically answers your question, but what I think of when it comes to service and maybe on larger commercial accounts, the agent or broker is going to serve as that frontline and it’s going to help them understand their policy. But getting on those types of accounts, you’re also going to have the company loss control folks involved or the risk management. I think that having the ability to interface with loss control and risk management with the customer, and the sooner that can get done, the more valuable to the customer. I hope that answers the question
TW:
Well, maybe a follow on to that. Are you saying that should the carriers be thinking about having a more active role in that particular piece of the policy lifecycle or is it really about enabling the broker to be more proficient at that?
RR:
Right. Providing the tools for the broker to be more proficient, I believe is valuable. Having the loss control and risk management folks involved on the larger accounts I think is that’s where the, other than when you have a claim and have to deal with the claims team, having that loss control person out there might be the only company representatives that they may have interaction with, so I think that is important.
TW:
Directly with a carrier, you’re saying insured for the carrier?
RR:
Yes. I think some of the challenges that companies have with that are the cost associated with that service, that function, and to which accounts do I provide that service. If you’re looking at some small business owners policies that are paying a couple thousand dollars in insurance premium a year, it may not make sense, may not be economically feasible for a carrier to provide those types of services as opposed to an account that may be paying a carrier a million or $2 million a year annually in premium. There’s a big difference between those two types of consumers.
TW:
Sure. Aside from gen ai, what are other big catalysts that you see on the horizon that carriers are looking at right now on the commercial side that you think are big game changers? It can be flavors of AI or something entirely different, but where do you think the next big set of opportunities there are? For real inflection points?
RR:
Two areas that I believe are going, or I shouldn’t say two areas. An area that I believe will continue that exists today and will continue to exist is really the climate change and natural and that’s going to remain a challenge for the carriers. I think that is probably one of the biggest challenges that there are and being able to manage that risk and do it in a effective manner using AI, I think will be a hurdle to overcome. I know carriers are working on it right now, but it will continue in the future also
TW:
And talk about something that is stochastic, right? We talk about that in gen AI terms, but climate risk is the, as stochastic as they come, it’s very challenging to predict and what used to be a hundred year storms, well, it doesn’t matter if it’s a hundred year storm, if it happens next year, right. Yeah. Well, great. Michelle, last questions for Rick.
MG:
Yeah, I was going to ask probably to kind of wrap on the AI and what’s the next big thing that underwriters need to prepare for is Rick, how do you think about knowledge transfer or knowledge management in this landscape as it is today, right when it was just paper, there was a challenge there. Now that you’ve kind of documents and digitize something that comes with other challenges in terms of training and really having the discipline of underwriting be the skill sets that’s developed outside of all the technology and other capabilities available. So how do you think about the next wave of underwriters and how do they develop and gain experience and expertise
RR:
Today? There’s a challenge of finding quality underwriting professionals and whether it’s out in the marketplace or companies training their own underwriters, and I think that will continue. There are fewer folks entering the insurance industry and so down a year from now, five years from now, that’s going to continue to be an area that carriers need to address. Now how that whole knowledge transfer within the underwriting team or the underwriting industry? That’s a tough one because having those experienced folks that have been in the industry for some time, they have a lot of knowledge that isn’t easily transferred. Now, whether it’s through training new underwriters and which can take a number of years in my experience to train an underwriter and have them be operational at an underwriting desk is going to take 18 months to two years and then gain knowledge from that. Now, if we can adopt AI tools to help the process and help the underwriting decision making, of course that’s going to help.
But it makes me think of one thing, and it may be a little bit different than what your question is Michelle, but there are going to be those books of business that underwriters will get to know very well, whether it’s a territory working with the same agencies and brokers for a period of time. They’re going to know those accounts and they’re going to have history with those accounts, whether they write them or not, they may be accounts that they’ve seen every year. There may be issues with some of those customers that the underwriters that have handled those territories are going to know, and it may not be in the file, it may not be on the new submission. It’s just because of their experience that they have this knowledge and as underwriters leave the industry that knowledge is lost and finding a way to address that is going to be difficult. And I don’t know if I have the answer or whether there is an answer, but insurance is a relationship business and it always has been. And having those relationships with the policy holders and with the agents and the brokers, we’re going to lose that as the underwriting staff turns over. So anything that we can do from an automated standpoint to help alleviate that or help enhance or improve that I think will be beneficial to the companies that embrace that.
TW:
Yeah, there’s this notion of individual expertise becomes enterprise expertise and that’s I think AI has a lot of possibilities and data, obviously data strategies and AI strategies have a lot of possibilities there to make it an industry that is more accessible, I think. Would you say historically underwriting has been very much an apprenticeship. It feels like one of those kinds of roles.
RR:
It really has, and that’s pretty true, Tom. It is. We team up new underwriters with seasoned underwriters and they learn the business, but truly, and this is something that I’ve believed throughout my underwriting career, underwriting truly is an art. There are times when everything can, all the objective criteria that we analyze something on, checks all the boxes, but there’s just this innate feeling about the situation that just isn’t right. And I don’t know how we put that into an AI format. That’s more the human element that would exist, and that’s why we’d need humans involved in some of those ultimate decisions is because of that.
TW:
Fascinating. Well, great. So we’ve been talking to Rick Russell, a former underwriting leader from Chubb and really enjoyed the conversation. Fascinating insights on what’s ahead for underwriting. So for Unstructured Unlocked, I’m Tom Wild again,
MG:
I’m co-host Michelle Gouveia.
TW:
Well, thanks for listening and thanks again Rick for joining us.
RR:
Oh, thank you very much. I appreciate it. You folks have a great day.
Check out the full Unstructured Unlocked podcast on your favorite platform, including:
