Talking Tech special: Dynamic Products

CHARLIE SAMOLCZYK: Hello, and welcome to Talking Technology. I'm your host Charlie Samolczyk. And in Talking Technology, we explore the wide range of technology issues facing insurers, from AI and data science through to open source solutions and cybersecurity, and we look at how we are helping our clients to tackle these issues.

So today's topic is dynamic products and their impact on insurance. I know, Magda, this is something that you speak about quite often from your strategy role in the UK&I and I think this is something that affects you. Maybe as a bit of an intro question, so we heard today people leading off with their love and their passion for insurance. Would love to hear from you what got you into insurance and what do you love about it?

MAGDA: Thanks, Charlie. I always joke that I actually got it into insurance by chance. It seems like a lot of people get into insurance by chance and then we never leave. And today I walk into Lloyd's and I really get emotional. I think this is an amazing industry. This is an industry that's the backbone of society and of innovation and has been for 300 years. So that's why I love insurance because we enable everything else. And without risk transfer, we wouldn't have grown. And humanity really owes everything of the last 300 years to the insurance industry. Although we're very invisible, we are that foundational and inherently social industry.

CHARLIE SAMOLCZYK: Yeah. Excellent. I know I ended up here kind of by mistake. I was in telco and then healthcare, then I made my way into insurance. Iain, over to you.

IAIN: Hey, look, I'm going to give a really flippant answer, which is it's an industry where you can still have a good lunch. Yeah. It's great. Yeah, you can-- and, yeah, there's complexity, intellect, good fun, good people. And I've never seen-- in my time in the industry, I've never seen people making decisions. I've seen people make bad decisions but not the wrong decisions for the wrong reasons. Yeah, that's-- so the ethics of the industry are good. I'll stick with a good long lunch, makes me very happy lunch.

CHARLIE SAMOLCZYK: OK. Had I known about the lunches, I might have switched earlier actually. Perfect. Well, maybe kind of jumping into the topic itself, so I think for the audience it'd be interesting to just give a bit of context, Magda, around what our dynamic products and maybe what need are they addressing in the industry?

MAGDA: OK. So dynamic products are-- we use the term opposed to what we used to have as products in the insurance industry, which were very static monolithic products that wouldn't change over time and could not react to data. So what are dynamic insurance products? They're products that generate the data at the right places and can adapt to data input. And therefore, the risk assessment and the needs of a customer can be assessed and the product reconfigures to maximize a certain type of objective function. It could be client satisfaction. It could be client value. It could be profitability. It could be underwriting footprint. And that's essentially what dynamic products are.

CHARLIE SAMOLCZYK: OK. And why do they matter?

IAIN: I think when people think about dynamic products, they often think it's personalized differentiation, adding a component, adding an extra cover, adding x or y. And I think that's an oversimplification. If you think of commercial insurance, I'm a commercial insurance guy, commercial insurance was everything was handcrafted for a very long time, everything was bespoke. The risks were underwritten in a bespoke way.

Then the industry spent 20, 30, 40 years industrializing that stuff and standardized everything. So you were left with the specialty markets, like Lloyd's, and other specialty markets that did the handcrafted stuff and everything else was machine-led. Now I think, we'll come on to talk about technology, technology innovation, the things that tech can do now mean that for elements of the commercial value chain, we can move back to some level of differentiation for customers. And that, yeah, at the right price point everything's not handcrafted but can be worked through. So that's an opportunity and kind of exciting innovation that's coming through.

MAGDA: You're asking about the benefits of why we should do this as an industry, right? And essentially, it's because risk is heterogeneous. And we are in the business of neutralizing risk and we know that risk is heterogeneous, but to be able to assess that risk and then to transfer it effectively, we need to be able to decompose it and digest it in a better way. And essentially, both in commercial lines as in impersonal lines, what dynamic products do is they have a more granular view at risk and are able to match it to customer needs and to underwriting appetite and to everything that makes our industry sustainable.

They can also be used for evil. So we could also say, oh, in this way, we're just going to get out the pieces that we don't-- I don't want to underwrite certain risk pools. It could lead to under-insurability. But essentially, the reason we are trying to look at products as the backbone of generating data is because if they can reconfigure and are not a commodity that one size fits all, then we can access and make insurance more relevant to larger segments and close the protection gap.

CHARLIE SAMOLCZYK: OK. So let's assume that they are going to be used for good, not evil. And Iain, you let in or open this up a little bit from a technology perspective. What are the recent advancements in technology that were needed to make this reality and underpin dynamic products?

IAIN: Well, let me have the first go on that one. Look, I think the tech stack here is interesting to fascinating. So the first thing is, monolithic isn't good. There's no monolithic answer to this stuff. It's a collective specialism. It's connected tech components that allow this to happen. And there are, I think, five of them that are really important. So the first-- and all need to work together in unison.

So the first is, you need rating engine. So you've got rate. We know of firms that have rating engines. So that's component one. The second is-- and in personal lines, the rating engine is, if it's not everything, it's a very, very important part of the process. Commercial is different in that at least a big percentage of underwriting risk is about the rules and what rules you apply, and also what judgment you apply, how you apply your distribution lens, how you apply your commercial adjustments.

So component one, rating engine. Component two, the rules engine, which could be the same thing, but those two components are kind of super, super important in that picture. But the next is you need an ability to look at the insurance contract and add and take away and deal with that in a modular way. So that's component three, which is kind of super, super important. And yeah, and work is underway to digitize contracts and we've done work on that. I think as an industry, we're in the foothills with that, but I think that'll be a big piece of work for the next few years.

Right. So all of that data's got to flow. So you've got to have some architecture to-- architecture for that to connect together. And then you've got to have the point at which the rubber hits the road. The UI that the underwriter typically is interacting with, where those changes are reflected and visible to the underwriter so they can craft a policy which has got dynamic components to it that meet the needs of a particular client.

Now, all of that hasn't been possible for a very long time. All of those components are available, possible - They're in the real world of where we are now. So that's my take on the tech stack of how that stuff's emerging. Magda, what do you think? What would you add? What have I missed?

MAGDA: No, you didn't miss anything! There's like-- we talk about technological convergence. And I think right now we're at a time where AI and advanced analytics and high performing algorithms, as well as the ability to generate data where we didn't have data before, as well as ways of handling distributed infrastructure for data, including blockchain plus generative AI and other things that enable us to interact with risk and assess it at a more granular way.

And that's new. We have each of these technologies interacting with each other able to deliver a level of digitization and risk tokenization that we didn't have in the past. And that's very much true for commercial lines, but also in the context of personal lines. You have the same components coming together. And so if we're going to look at this data to understand and predict future behavior for risk and for consumers, generating it at the right places, we can actually transform products so that they adapt to this and they maximize some things that are critical for us to survive as an industry.

So I think one of the main reasons why right now technology enables risk assessment and risk computability is because we are able to translate data into risk assessment in a much more granular way than in the past. And once you have that and risk becomes more computable, then it can actually connect to capital and be transferred in lots of ways. Some will need humans. Others will be automated. Others will be algorithmic. And that's why we're at a point where technology has enabled something that we couldn't do 10 years ago.

CHARLIE SAMOLCZYK: And Magda, you were telling me about some of the research that's happening in that area. So that's probably worth sharing and talking to people about. That in box three, digitizing the contracts, allowing that to happen, that, from what we've done, felt really hard and quite difficult to do on an industrial scale. But I think some of the things that you're seeing now perhaps open up new opportunities.

MAGDA: Indeed. So we have been looking at something called computable contracts for quite a while. And essentially, the idea of computable contracts is that the contract is a code and the code is the contract. And it's a North Star, but there's a lot of things you can do before getting there. And for a very long time, we were seeing computable contracts as something that was very difficult to industrialize and reach at scale. And also, it was very difficult to apply it to existing books of risk. And so that conforming to the past was also very different.

Today what we're seeing based on what Stanford is doing, part of their Codex initiative together with insurance companies and some of the things we've done internally, is that by using additional tools, including generative AI, you can go from a process where generating a hierarchy and a logical model to transform a contract into something that is computable in terms of its parts go from six months to a week. So that's the game-changing because that means that the economics of transforming and digitizing contracts now work. And so that's very interesting and it's a very recent development.

IAIN: I like your reference to North Star. So I'm curious, so to what extent is this still a direction of travel that we're aspiring to or are people really doing this in practice today?

MAGDA: The first prototypes-- Stanford has been working on prototypes and those prototypes were already tested and they exist. Is this being industrialized at scale right now? No, but we know how to do it, which is a big difference. I mean, a year ago, we were all discussing, how do we make this a reality? How far do we need to go before this becomes something impossible to do? And now we're at a very different moment from a technology perspective.

CHARLIE SAMOLCZYK: But these trends are happening, aren't they? I mean, things which felt quite distant a couple of years ago are very on trend now. Yeah, so there are multiple vendors out there providing really good solutions around ingesting data into the insurance ecosystem. That felt impossible at a scale a couple of years ago. There are people out there doing it in a real world now that have got solutions that allow an underwriter to go onto a screen and to interrogate a policy and identify features of the risk which are in the submission.

I mean, all of that is hybrid technology because that's all the brokers are going to digitize everything coming in a five-year window anyway. So you take that, you take the digitization of contracts, all of that moves you to a world where risk is-- or a risk coming into an insurance ecosystem is going to be priced by the machine before it goes to the underwriter. And then the underwriter will finesse and will make decisions from that point.

And at the back end of it, the process risk will be tailored to the needs of the insured and the contract that will be produced at the back end of that will be detailed and specific to that need. And, the insurer will be able to manage that risk and understand it in a way that it couldn't, i.e., the insurer won't have to go back to looking through all of their documentation to understand what risks have got on the books. So that's all here now and, yeah, that's-- yeah, as well as a good lunches, that's part of the reason I'm in this industry too, to bring that change about.

MAGDA: And not only can you start now having a better view of all of those risks in your portfolio individually, but you can also see how they are aggregating and accumulating in ways you couldn't at a much more granular level. And then there also very operational and practical implications. So if you can have an algorithm read from a PDF and determine what exposure looks like and what loss looks like, you can automate claims. And that's already happening. There's two players out there that already are able to do that using a combination of these technologies.

IAIN: OK. So it's important and it's real and it's happening.

MAGDA: Yeah.

IAIN: I mean, you guys are both kind of at the coalface of it. So how can we help people? How can WTW help people on this journey, and what can we offer them?

MAGDA: So I always say digitization is not linear as a journey and you need to understand what type of benefits you want to unlock to see if more modular or more dynamic is enough for whatever you're trying to do. But the foundations of what we're building and the way in which we digitize do have an impact into how far into the future you can go and how many of these benefits you can unlock.

So I think that's essentially where we're helping the industry, it's in how do you digitize what you already have, how you generate products that generate data at the right places so that you can iterate them quickly so that you can monitor their performance so that you can scale up products that you have in one country into another country, that you can have learnings going at a larger scale?

And all of that is a complex transformation program. And therefore, what we do is to embark on that journey with our clients and then look at what they have in place and how they can prove that, how can they build a business case for internal stakeholders, where the benefits are. So it starts from education all the way down to implementation actually.

CHARLIE SAMOLCZYK: Iain. You look like you're itching to get in there?

IAIN: No, just do what Magda says. It's got to be the answer, isn't it? And of course, we can push our products. There's very few problems that Radar can't solve would always be my answer to everything. Yeah, very good.

CHARLIE SAMOLCZYK: OK. Well, on that note, let's end it there. Thank you very much for your time today. It's been a great conversation, and thanks for joining.

MAGDA: Thank you. It was super fun.

IAIN: Thank you, Charlie.

MAGDA: Cheers.

SPEAKER: Thank you for joining us for this WTW podcast featuring the latest perspectives on the intersection of people, capital, and risk. For more information, visit the Insights section of wtwco.com. This podcast is for general discussion and/or information only, is not intended to be relied upon, and action based on or in connection with anything contained herein should not be taken without first obtaining specific advice from a suitably qualified professional.