AI Wisdom Ep. 12: The Future of Commercial Insurance with Eric Joost

Digital Transformation - May 19 2020

On this episode of the “AI Wisdom – Talking Innovation in Insurance” podcast, host Ron Glozman speaks with Eric Joost, former Global Head of Property & Casualty Insurance, Willis Towers Watson. Eric shares his insights and opinions on how commercial insurance companies will re-invent themselves to meet the demands of an ever-changing marketplace, how they will embrace emerging technologies like AI, RPA, chatbots, drones, etc., and what areas of the business will require the most critical changes. Click the link to listen or read the full transcript below to hear what the future holds for insurance.


Ron: Hello, and welcome to “AI Wisdom Talking Innovation in Insurance”. On this podcast, we talk to business and insurtech leaders about how artificial intelligence is transforming the way we buy and sell insurance. I'm your host, Ron Glozman, founder and CEO of Chisel AI, and a strong believer in the power of AI to help people work smart and enrich their lives. So, let's get into it.

I'm very pleased to have Eric Joost, former Global Head of Property & Casualty Insurance at Willis Towers Watson join me today as we discuss the future of commercial insurance. Welcome, Eric. Before we jump into the main discussion for the day, do you mind introducing yourself?

Eric: No. Thanks. And thanks for having me, Ron. One thing, and again, and it's just in the sort of scheduling, I have left Willis Towers Watson very recently. But my background is 30 years in the industry, insurance industry I should say. It's a combination of broker Willis Towers Watson, insurer Allianz, a German-based company, and then AON previous to that, my background has, through luck versus strategy, been diverse. So, I've done everything from direct client sales, broking. I've also led technology and innovation particularly inside the risk business, and in particular, at least as it might relate to some of these topics today, I repositioned a small-medium enterprise portfolio at Willis and actually partnered with a firm called Insurian to sort of reset how we serve that space. And again, there's a lot of value in both the good in that process and also what was challenging.

Ron: I love that! And I love those are some of the great companies that we work with and companies that we want to work with. So, let's touch base on that after. But I would love to get your thoughts obviously right now there's a couple of things on everybody's mind and everybody's right now talking about coronavirus and COVID. So would love to hear a few things that's disrupting the market and then I would also love to hear your thoughts on what technology you think is really disrupting the commercial insurance space.

Eric: Sure. I mean, I think from I take the first question really at a general business level and it's not exhaustive. And first, and this is pre-COVID, right, the business results for the risk-takers insurers were lagging and whether you want to say that's the last two years or the last four or five years, whatever dimension, there was clearly an imbalance between the price that the insurers were taking risk and then also, obviously, the other side of that is a loss as they were paying out.

And more recently, I would say in the last year there's been a lot of what I would call strong change and repositioning in the market by the market leaders. So, you have firms like AIG or Liberty, or Lloyd's or Allianz reestablishing kind of an underwriting ethos in their businesses and starting to bring some change to the pricing risk dynamic that I mentioned before. You also have smaller players in the market that are perhaps leaving some of the segments that were unprofitable over an extended period of time. And there's examples of that just in terms of client segments or in the construction space, the energy space, the marine cargo space, where the business results have been very, very difficult for a while.

And again, that took a period of time to develop. Now you add on COVID or COVID-19 that does probably two things. It's probably a little early to speculate and once the insurance impact, there's lots of kind of arguments out there about what it should and shouldn't be. I'll leave that to the side because at this moment in the cycle or in the story, it's just an argument or a rhetorical debate.

The one thing though that COVID is impacting is the investment side of these insurers. Many of them are heavily invested in sort of bonds in corporate bonds and municipal bonds and things that are very conservative. But there are some insurers that are certainly a little more aggressive and so this can wreak a little bit of havoc in the investment portfolio, which if you just combine it with bad business results or underperforming business results from a risk point of view, then you compound it with an investment result that's probably not terrific.

The one thing I would say is while important, it isn't in the drivers in terms of what's having a big impact is technology. You mentioned personal lines and that, I think, has a different trajectory, but yes, while technology is everywhere, I don't think you would look at the major drivers of the commercial industry and suggest that it's having a large impact. It has great potential and, obviously, it's in use throughout the industry. So that's sort of how I look at it.

Ron: So, I wonder if you see different types of technologies. There's a couple that I can think of that often people talk about. So, there's things that are like AI, which is a very broad branch and we won't break out all the different types, IoT - Internet of Things, RPA, do you see any differences? Do you see one technology is more promising than the other? Are there technologies that you have seen already implemented? And I'm curious to see also if you see any differences in the way the brokers are approaching this like Aon and Willis versus carriers like Allianz?

Eric: Well, I mean, at first on the tech, I think they're all potentially valuable and I think it's early. And I think there are good examples of technology that work. We certainly were using them at Willis and externally whether it's AI, IoT, RPA. Scale, in many cases scale, achieving impact at scale is still difficult. I might be wrong about this, but I look at the transition...the most advanced insurance market digitally in my opinion in the U.S. at least is auto, and that took almost, say, 15 to 20 years.

And I'm not suggesting everything has to go at that pace, but I think sometimes as we get in this conversation, we look for that the good, the early example and then we assume it's just gonna exponentially scale. What I think is missing these days is it is not a criticism of the technology; it is there's a larger ecosystem all of these things plug into.

And we suffer a little bit in the industry of not really addressing that issue and that's when I think some of the larger companies, they get something that's neat and cool and it makes a process that...(I'm making this up a little bit), but it makes a process that was once 24 hours, 10 seconds. But they haven't changed the processes around it that either feed it or are or downstream from it.

And so, while that piece of the ecosystem is much, much, much better, the rest of it's pretty much the same. And some examples, I mean there's some great AI, some interesting AI. I was in India last year and India allows an insurer to look at each of their brokers or agents and go, you know, what part of the's like just a four-box square, what's profitable, what needs to be managed away and unprofitable. And it's just kind of a classic sort of consultant visual.

Now two things about that, they used to take this insurer months to produce this for each of their brokers. And then that would guide a conversation about, "Hey, can you avoid this? Can you try and help avoid this kind of business and find more of this kind of business?" One of the things is, so now this can be produced in a day or maybe really in hours. But so that was great. Now these people are producing this really great insight except the issue is the insurer can't on its own go change that two by two chart because that broker also has a portfolio they're trying to run. And as soon as you force, let's say a broker, an agent to say, "Hey, we don't want this business anymore, it's too risky," that will put in play your good business as well because the broker has to go to market with both of those things to sort of achieve some portfolio results.

So, it's just another example of I'm seeing pockets or pieces that are really effective and insurers or brokers or even some of the third parties that are around this. What they're not doing, and it is very difficult is to how do I improve the overall result, or the client experience, or the result for the insurer, the result from the broker. Everyone is sort of making a step 10 to step 14 really great but steps 1 through 13 and sort of everything after step 14 are the same, that's really not a different environment, it's an incremental improvement.

_Eric Joost Blog Quote #3 (1)

And the last thing, I mean some of the IoT in auto and cargo I think is really interesting. I've had a lot of experience trying to use RPA, I think it can be really, really effective. I think there's still a challenge of getting some of the variation out of a process before it can be helpful. I know when we looked at it, we basically could achieve a cleaner process but more expensively than just using people. And that didn't seem to make a lot of sense.

Ron: And it's always about the people, I think that's most important, especially at a time like this, that's what always comes down to. And it's also to something that you said a little bit earlier on, which is like, it's about the processes around it and those have to go with that. I talk a lot about this when I think about a project and one of the biggest not necessarily risks but one of the areas to focus on for sure because if you don't pay attention to it, it's probably the one that's the most critical is the change management. And change management is difficult because it's a soft skill, it's not like a hard skill that you can necessarily teach or transfer and so it comes down to the processes and the people. And when you think about it at a time like this, is this a time for more automation or the same amount or less, I guess?

Eric: I think it depends on your horizon. I think what I'm noticing now just in this around COVID last sort of five or six weeks is one of the sorts of things we all may not have anticipated is process becomes more important when you're operating remotely. Because sometimes what's great about the people and if they tend to be around each other, they tend to mask the imperfections in the process because they all kind of understand where they are. There is a great conversation I had with Brian Duperreault who leads AIG now, and he talks about this a lot in terms of their technology transformation. He said one of the challenges about the transformation is we've been masking a lot of the problems for a very long time. So, as we make all these changes, some things may not look like they're improving because we're basically building the plumbing after the fact. And I think there's a bit of that.

I think in the longer-term, whether it's what I've seen, the very, very, long-term, I think, more automation obviously. I think in the shorter-term it's just smarter combination of a little bit of automation to then help sort of supplement what the humans are good at.

And that's where I see things kind of moving ahead whether it's today, I've seen it in the last couple of years, and I certainly see in the future is that for people like us helping firms find the right balance of tech automation and then where precisely the humans need to be involved. And it isn't always obvious, it isn't always a purely logical sort of set of things but when you get it right, suddenly it goes from sort of grinding and clunky to which just makes sense. It's like I trust these things to give me...I bring data to it, it gives me either an output of an algorithm or some insight and I then start to trust that instead of trying to get 10 different inputs on the same question and then that makes either me more efficient, faster, a whole host of things become available. I think it feels like that.

Eric Joost Blog Quote #2 (1)

Ron: That's right. A lot of people I have been coming to realize that the processes that existed may not be the same ones that go forward. So, I love that. What would you say is the biggest challenge today facing commercial line insurers?

Eric: I think it's complexity. And I think its complexity and then as they continue to basically approach the world in a more traditional format. So, most of my 30 years, and I say this to my peers that are of a similar vintage, what's the biggest change in the industry? Almost to a person, we say email. It doesn't sound like a lot, but that to us...and other than that, there's not a lot that's all that different. Yes, it's all delivered digitally now but for behind the scenes, it's still more or less the same and it's a very messy process.

And it's not messy because the brokers decided to make it messy or the insurers wanted to make it messy, it's messy because we're dealing with complex customers that want custom things and they tend to drive a lot of decisions, meaning they're paying for things and they want things in certain ways. And it doesn't always mean that you can approach them with like demographics of between 40 and 50 and drive this kind of a car. It's risk and risk is a funny business because it's kind of a second, third-order thing, so people just approach it differently.

And, yes, insurers operate differently, brokers operate differently but that complexity, I mean it's just, if you think about the big brokers, the permutations that exist, if you think about how operate across products, hundreds of countries, many different size clients, and then inside all that are different client preferences. And as you get into smaller brokers, some of those permutations go away but not all of them. And even someone operating here in the U.S. and maybe just in the kind of more traditional lines of business, they're still gonna have a wide range of permutations of things they're trying to bring to clients.

And so I think that's not a problem in the sense of complexity I said before, it wasn't created by anybody, but the way we've been trying to deal with it both as a business and to some degree from a technology point of view is we try and force simplicity onto it as opposed to just saying the world is complex and it moves around a lot. We can shape that and then why don't we think about processes and how we run the business and the technology in a way that needs to address kind of more of a moldable kind of shape of clay type of environment versus what is at least stereotypically a technology approach, which is basically, "Hey, if you organize perfectly, my technology will work for you great." And that's just not realistic.

So it's really just managing through it, I mean, I think acknowledging that there's complexity and acknowledging that the traditional approaches are going to need to evolve and then slowly managing through it, that's a path that's got some legs.

I think constantly I notice this in some the InsureTech world, there's a lot of people that get up on stage and just go, "This industry is awful." And the moment they start doing that, they start simplifying, sort of, their approach, they usually end up rolling right over all this industry context, it's very important. So that's the challenge. It's different than it was 30 years ago, and we'll see how people embrace it.

There is an industry investor and analyst, VJ Dowling, Dowling & Partners, and he talks a lot about technology, and he is got a very pragmatic, or his firm has a very pragmatic view. And what is the biggest technological shift in the industry or piece of tech that's worked the best, Excel, a spreadsheet, that's the most...He sees the potential, but he's not wrong either.

Ron: I agree completely. I think without Excel we would be like economically trillions of dollars probably deficient.

Eric: Yeah, that's a good point.

Ron: So, what trends do you see that you're excited about in the short-term? Like let's say, what do you see coming 2020, 2021 that's really gonna have an impact that you're excited about?

Eric: I mean, I think there's, I guess, a couple of things. One is I think some of the tech of all the.. there's a lot of my old firm did a quarterly review on InsureTechs and things like that. And tracking, we would track a thousand-plus of these firms but inside there's a lot of stuff there. And what I generally find is most of the tech works, it's just how good are the people leading those firms to integrate it into environments and ecosystems. And I think there's some pretty good stuff out there. I've been really surprised just in terms of whether it's AI, all the ones you mentioned, and others where people are really kind of putting together really good stuff. And then the challenge is...which is great. Not everyone survives, you gotta run the business that's building this stuff. And then it's who can adapt to the industry and make it truly valuable.

I've seen great stuff in like auto claims payments. Like you said, AI is a broad church, but I've seen some really interesting combinations of using machine learning and AI – which maybe you might argue are not two different things – to do some interesting things to sort of help manage or lessen the inbound data variation that comes in to insurers or comes in to brokers, for that matter.

I'm certainly excited about that thing. I think there's a bunch of firms both big and medium-size and they're putting out interesting risk algorithms that are kind of approachable if clients do want to look at their situation without always having to bring a human out to their office or get on the phone, it's just like there's really good math out there that you can provide some inputs to and get a reasonable model of what you're dealing with and then obviously kind of decide later whether you want to work on it.

I also think there's just a healthy...I certainly believe in this, it's not always a popular opinion but you've got to respect how hard it is to get some of this stuff into the industry. And there's a graveyard of people that have ignored that and that's why there's headstones there. But the people that really respect that it's complicated, go find work, go attack the complexity, deal with it, and bring some industry context to it, I think are actually making a lot of progress. It hasn't really, really shown itself yet but it's certainly much, much better than it was two, three, four years ago where everyone is still in industry's bad and my new product is great and this is gonna solve the solve everything at once.

I think people are starting to realize how it's gonna be an iterative build to basically sort of modernize most of this ecosystem. And frankly, the other thing I'm excited about is no one's shown me where the human being has to come out of this stuff. The very best and most talented people in this industry are just gonna be made better by what happens in the next 5 or 10 years.

We need as an industry to stop sort of suggesting that the automation eliminates the people, I think the automation eliminates administration or adds speed so that the really good people can make better decisions about either what they sell, what they buy, or what kind of risks they take on.

Eric Joost Blog Quote #1 (2)

Ron: I think it's a very exciting time and I think it's gonna be a time to really, really compete and really differentiate yourself and as you said, some people will crack under the pressure and some people will see this as an opportunity to really push forward. Let's take a 20-second break to tell everyone where you can find more information and insights about insurance innovation. We'll be right back.

[If you liked this episode of AI Wisdom, subscribe to our blog, Writing the Future: AI in Commercial Insurance at for feature articles, interviews, opinions, and more.]

We're back with our featured guest, Eric Joost. Let's jump right into the next question. For those that understand transformation is a necessity but are stuck in analysis paralysis, there's so much information out there nowadays. There's so many startups that at least on the surface seem to do the same thing and it can be something, even something very specific like claims processing or, analysis of loss runs or other things. On the surface seems like so many companies that could serve the need and it's impossible to choose or differentiate between them. What recommendations can you offer to those people?

Eric: I think first, and I know where you're coming from, but I don't see analysis paralysis, what I see is there it's sort of I've noticed it in my career over three decades, when I first started you didn't advance, you may have become specialized in something, but you were forced to understand like how your business worked, how the industry generally worked. And I think to some degree everything is become so hyper-functionalized that there aren't many people inside whether it's a broker or an insurer really understand how it all kind of knits together.

And I think, I talked to a lot of CEOs about this, they understand that the world is gonna become more digital, they understand that these things out there they have to analyze loss runs, they have to do these things and they can do it digitally. But what they generally aren't getting is whether it's themselves or their teams, like a comprehensive plan of how you sequence all those moves. So, like as soon as I start doing the loss run stuff, what are the other implications for that? And then what are the implications downstream or in two or three years?

And the way I like to talk about it, I said insurance is an industry generally at the leadership level still has a very big technology maturity gap that just seems to be true to me. And then the flip side is technology as an industry has a big insurance maturity gap. So, unfortunately, you've got a lot of, I think, people driving that change in the macro idea and then when you really got to get into is the granularity of it and how you approach things. Blockchain is a really interesting technology. I don't know that it applies to insurance very easily, not because it doesn't work, but because it's an expensive way to deal with the problem or the opportunity, number one.

Number two, I would try to be positive. I see some really interesting examples, Chubb in the U.S. and then Hiscox in the U.K., I know their leaders, they talk sensibly about this. They know it's iterative, they are paying serious attention, their leaders actually are in the details. And their leaders force the team around them to become more technically mature. They don't just act as business unit leaders and what's your P&L, what's your result? It's part of your responsibility in those firms just to get better at technology because it's coming.

So, you see some of that stuff really work. The other part that I think is difficult that everyone misses is, sooner or later, some of these digital businesses it takes five, six, seven years to transition from say manual to digital, if you're standing up two businesses, you're basically then standing up two businesses temporarily, that's expensive. That's two infrastructures and in some cases, you have to coordinate a manual business and a digital...kind of constantly synchronize it. And all of that is technically doable but I think people aren't used to running things that way. So I think that's why everyone gets stuck and they kind of get into kind of the hedgerows of this as opposed to take it and then think ahead about what's next rather than buy the first cool thing and then get frustrated when it doesn't really fit into your bigger engine.

Ron: Are you a car guy?

Eric: Not really, although today where I'm at, I've got a Mini Cooper sitting in the sun because for some reason or other it has tended to take on water for the last two days during a rainstorm. So now I'm a car guy.

Ron: Well, hopefully, you have insurance on it if you need it. So, thinking about a bigger timescale because it's always interesting to live in the next sort of two to five-year horizon. I want to go far out, I want to go 10 to 15 years big picture type of change, what do you see?

Eric: What I see in terms of that time timescale, I think you're gonna see a lot of these things. I go back to the auto example, some of these technologies that around as an industry and individually people are gonna get better and better at using the AI machine learning big data. And then over a period of time, I think that the tech providers or the people that kind of come with that bias are also gonna get some industry context that's really powerful. And so, I think that's actually...what that starts to do, I think, is really change the game because I think in the beginning...the industry, fundamentally, spends too much of the premium dollars on expenses. So, there's a force, there's a movement to take that expense ratio down. And in the shorter-term, that's always difficult cause you kind of got to run what you have, invest in new things.

In the medium to long-term, I think you're going to start to see the application of a lot of the technologies you've mentioned. And it's not just making an administrative process better, that's easy. It's also taking really big amounts of data and then getting it to a human being who can do something that's actionable that before maybe would have taken months to understand, and then the opportunity to make a decision has gone, where you can actually do some things with the best minds in the business and at speed. And I don't think that's there today, but I think it's coming. I'm certain in the personal lines space some of that exists already. I've seen it. But in the commercial lines space, it could be very interesting.

I mean, just take one example is the energy industry is littered with IoT devices, they run their industry on it. What's the possible impact from an insurance and risk point of view to start using that data more strategically, first as a client, maybe second as a broker in terms of how you go out and procure insurance for your clients or as an insurer, how you take on risk. Go to the places in the world where you've got IoT devices, big boats, planes, energy, industry, cargo, there's a lot to do there where suddenly you can really get some insight into risk but you got to wade into it. We didn't just immediately figure out by throwing kind of GPS monitors and commercial trucks, okay, how we can monitor the behavior of all the drivers. But over a period of time, we started to look for what was more important and less important because you cannot manage a thousand things, but you can manage five or six. So, I think those come. And that's the shorter-term.

The long-long-term is I think you fundamentally start...there is a possibility that you can really begin to manage risk differently. There is probably just's easier to make different decisions based on your balance sheet, your cash flows, whatever you're sensitive to. And I also think you can start to do some things where you can take non-correlated risks and kind of wash them against each other and overall lower the risk profile of a firm. And some of that might bridge out of insurance into other spaces. But you can see the math getting there, and then you really still need the people to believe that that's the right thing to do, that it works, it's gotta be tested and that will take a lot of time.

You mentioned it earlier in the discussion. It's sort of, there's a lot of economic value if you can better optimize what you spend money on in terms of risk and also just, but overall, maybe lowering the risk curve, that's important. That's money that could be going into building bridges, driving profit, and invest it back in the company or growing a business and all those types of things. So, I think that's...I'm really optimistic, but that will be a very patient journey.

Ron: Patience is very, very important, time heals all wounds. So, as we wrap up, is there one piece of wisdom that you would give to your peers in the industry?

Eric: I mentioned this earlier, you have got to respect the facts and the history. For an industry that likes data and even the technology people that come into it, they start with, you know, the industry is messy and I'm smarter than the industry. You have got to go back and look at what worked and what failed and why. And you got to realize that things we are talking about are incredibly valuable potentially and they're very hard to get at. And there is a large graveyard of ideas that kind of recirculate in the industry because people jump through's just easy, it's right there in front of us.

And I think the other part is don't ignore...we get all worked up about, kind of, the user experience, the human experience of all these new things we're putting on people and I think sometimes it gets wrapped up in people don't like change. I would say for the insurance industry, most of the change that's been brought to the colleagues in the industry has been poorly executed. So, the reason I guess would say two things. One is most of the smart people in the insurance industry like change, most of the change they've seen has been terrible.

And so understanding that human interaction and getting it right and learning how you would if you were a consumer-facing technology firm like Facebook or Google and things like that, it is the essence of getting some of these things to work. The technology is there but I find the leadership in the industry tends to hide behind its change management and people do not like it. And what I think that they don't do is go at the human factors on this. Because what I've seen some firms kind of really breakthrough, it was just this admission of I can't reset this whole digital platform with all the changes I want to make in one swoop, I actually have to split it up into 15 chunks.

And suddenly they realized everyone's kind of with little change all the time, everyone is moving along and it's fine. If they drop the stuff in all at once, everyone freezes. And it was just one example but you got to respect how hard this is and you've got to get after the human stuff instead of assaulting the humans with, I know you don't like change, but you know, we're just gonna tough it out.

Ron: Eric, I think people are gonna really connect with that. Thank you so much. Is there somewhere where people can connect with you, Twitter, LinkedIn if they want to find out more?

Eric: I'm on both, but LinkedIn is a lot easier. So, it's pretty straight forward, there's not many Eric Joosts hanging around, so you'll find me pretty well. Then I'm on Twitter –  @ericjoost. But either one works. And, Ron, it's been fun and thanks for having me.

Ron: Awesome. Eric, thank you for joining us. And As always, you can find us on Twitter, LinkedIn, Facebook, and you can find me on Twitter @RonGlozman. Thank you, everybody, and stay safe.

That’s a wrap for this episode of “AI Wisdom” hosted by Chisel AI and me, Ron Glozman. Thanks for listening.

If you like our podcast and want to hear more, check us out at or tune in and subscribe wherever you get your podcasts: SoundCloud, Spotify, iTunes, Google Podcast, or Stitcher

Join us next time for more expert insights and straight talk on how AI and insurtech innovations are transforming the insurance value chain. See you on the next episode!

Browse different topics

Recent Posts