Subscribe
Share
Share
Embed
Discover how insurers are entering the decision era, utilizing artificial intelligence to refine their decision-making processes, boost underwriting profitability, and achieve premium growth.
Tom Wilde (00:05):
Welcome to another episode of Unstructured Unlocked. I'm your cohost, Tom Wilde,
Michelle Gouveia (00:08):
And I'm your cohost, Michelle Gouveia.
Tom Wilde (00:12):
Well, great Michelle. We have a really exciting guest today, Thierry Daucourt, who is going to talk about the future of AI specifically as it relates to the more human element of it and how we use AI as an advantage and an accelerator, but very much human-centric in terms of the benefits that we can derive from it. It's a leadership mindset change. We have to think about in some ways with H Gen ai, as are these also workers in a sense that are augmenting the workforce. So Tiri is going to share some insights based on his decades of experience in the industry. So Tiri, welcome. It'd be great to get a little bit more of your background as we dive into this super topic.
Thierry Daucourt (00:54):
Hey, thank you. Thank you for having me. So given my age, I could go back four years into my career and explain all the details. It's too long. More details are the LinkedIn profile, but I think what is interesting and what I really enjoyed in my career, I'm a Swiss national. I spent about 16 years abroad. So in other countries outside of Switzerland, I felt that that market is kind of small. And then also I worked in almost every function in insurance except it. So I've never been a programmer, but I also had an agency. I was a Tide agent, that's how I started in the insurance business. And then I worked in claims at bookkeeping, accounting, underwriting, et cetera, et cetera. Generally in commercial line property and casualty insurance business. That's where really my experience is coming from. And as a good Swiss, you can identify me because I wear two watches. One is a real Swiss watch, the other one is an Apple watch, which is a gadget. So I like both of them. That's why I wear two watches.
Tom Wilde (01:58):
I'm a big Swiss watch fan as well. I wear one every day. Despite all of the smart devices out there, it's just very comfortable to have that on the wrist. Let's start with the bigger picture. You've been on the front lines of digital transformation for a while with your experience at AXA and other places and commercial. In my experience, it's a unique animal, right? It's a complex, complex business that sort of inherently resists digital transformation not because of an unwillingness, just because of the complexity of that business. So where are we today in terms of the progress towards digital transformation in commercial? You and I offline have had some really interesting conversations about today and what the future holds.
Thierry Daucourt (02:48):
I used to, I still do, describe commercial line to be a little bit Jurassic Park of the insurance business in the sense that we try to avoid to be disrupted as much as we can. I think. So some of it is resistance because it's very hard to automate completely its commercial lines. So it's about ensuring large industries or small businesses worldwide or in a given country. And I think that there were a couple of events, I would probably name the pandemic as one first wake up call for the management to realize that actually it was not that easy to understand what our exposures were in that situation. And to collect that information as actually was tough because you had to ask people to go into the archives to get the paper files and try to find out what was insured under the pandemic. And then came ai.
(03:48)
And with AI, suddenly management, senior management, everybody wanted to do ai, although we didn't really understand what the hell do you do with ai. But that has given another attention. And now I think that you see a lot of investments in transformation projects or there are almost always IT related projects going into commercial lines. And there's one advantage also and one change that I've seen in the last years. We are moving away from just insurance product to add services either to the product or besides that. So adding services. And then the fact that commercial line is easier as a customer segment to collect data. There's less data privacy implications in it. So gathering data about your potential insureds about your potential customers is a bit easier. And that has opened up I think a huge possibility combining insurance product with insurance services and leveraging technology and data.
Michelle Gouveia (04:53):
Excellent. A lot to, in fact, I agree with you on a lot of those points about the complexities, but also the maybe plethora of more opportunities to capture the information you may want, even if it's more complex to organize it or analyze it. One thing that you said, Tara, that I wanted to maybe have you elaborate on is it sounded like a little bit of maybe the slowing of implementation or slower progression is not so much technology holding us back, but I'll say the limitations of people in the process. Can you talk a little bit more about your thoughts on that?
Thierry Daucourt (05:42):
Good question. And the answer could go for hours, but because it's quite complex, I think that there, there's an element of a few years back, I was able to dream about stuff that it could maybe help me to do, which it wasn't able to. Today a lot of things can be solved by it, but now it has changed because now it requires management to be able to embark the people to that transformation. And it's more than a transformation is a disruption almost. And so that's one point to keep in mind that you are now asking people to really change the way they've been doing their business. And for an underwriter who sees the client meets the client negotiates the wording, the contracts and all of that to say, well, actually we're going to change a few things in your process, not easy and particularly not if you don't allow them to digest a little bit what's happening with them.
(06:37)
I think that now we are also confronted at a pace of technology. The rollout of new solutions, of new potential capabilities that you can embed, implement in your underwriting process in the insurance industry is so fast that you really have to make sure that you as a leader who hopefully understands the topic, that's also a point, are able to really make sure that your people are able to follow what you're trying to do there. Because otherwise you end up to be alone as a leader in your new world, but nobody is following you. And I think that their implementation requires very much an attention from the top management to make sure that you can bring those people along that journey so that they adopt those changes. In some cases, you have believers, you have those who are technology savvy, who are used to are interested to disrupt a bit their business, they will help you.
(07:36)
And there's another point to consider. I often found that it's hard to ask a team of business people to disrupt themselves because with today's capabilities, you can do much more than just to digitize your process. What you have to do is to reinvent how you do insurance, and then you have to add all those capabilities to it. And that's something very hard, I think, to ask from your own people who've been doing that for the last 30 years and don't really want to change the way you do that. And I found that for large organizations, it's a good way to do that is basically to create a greenfield kind of operation, almost like a startup in the group and to say, look, guys, you are going to attack the business of that team over there. That's your job. Because then you separate the two units. You have those who keep doing their job and you have those who really are operating on a greenfield or then that's the reason why you have so many startups in insurance suddenly with the rise of technology and AI and what it can do because they are able to deliver solutions in a much faster turnaround time. And they have the people who are there to disrupt what the insurance industry is doing.
Tom Wilde (08:57):
It's a great point. I want to poke at that one some more. One of the challenges I think with commercial insurance and probably the whole insurance category is these are complex processes, very large dollars at stake. And so the buyer of these technologies are the institution, right? The institution has a goal to improve efficiency, to improve combined ratios, et cetera.
(09:24)
But inside are the individuals who do this work. And I think kind of billing on the point you make, I think if you ask the average underwriter in commercial insurance or specialty, is there a burning house here? Do we have to solve this? They would say, no, I hit my numbers. The company's profitable. There's not a burning house here. So to your point, why do I need to disrupt myself and what I'm doing? But on the other hand, they've almost been conditioned to accept when I'm inside these companies and see, these underwriters might have to use 2025 systems to underwrite a risk and do tons of cutting and pasting and re-keying sort of they're used to it. And to your point, it's almost like they're not sure what's possible on the other side of that. So the ability for the institution to both bring them along, but also focus on them almost as the customer. Where do you think we are with that?
Thierry Daucourt (10:27):
Well, I think probably some people do it really well and others struggle a little bit more. And there's another element in that one, which is the more you are confronted with this AI at home at work, the more you can be either concerned or excited. And when you roll out a new solution, a new technological solution in general in the organization, it means expense saving or productivity gains. It can have a negative connotation immediately because people then think, okay, that's what they want to do. They want to roll it out. Basically they want to eliminate my job. Whereas actually there's a huge opportunity I think, to do it now, which is to manage the workload that you have and to increase the quality of your work because you're suddenly supported by technology, but in commercial, and you still keep that control. You are that human on the loop who will make the calls, you will make certain decision and decide where your digital assistant basically is going to bring you.
(11:33)
And I think that starts to happen also in geographies, like in most countries in Europe with the demographics, we have a situation where it's hard to recruit underwriters. We have a whole generation of senior underwriters who are going to retire soon. I'm one of them maybe already as a first example. And so you have now the options to say, okay, I can keep a pool of talented underwriter and hopefully also hire some and train them, but I can now equip them with technology that enables them to manage a much larger chunk of business or a portfolio or number of customers compared to what they've been doing in the past. And I think that there is an opportunity to show, to demonstrate to employees, to the underwriters or whoever is impacted by a technological solution that is actually a win in there for them to improve the quality of their negotiations with the clients, to improve the decision making.
(12:38)
They look smarter in front of a client. If you're well informed, well sort of educated by your technology, which is able to bring you the right information at the right moment in time in front of your screen or your desk. So that should make your job actually more interesting, more qualified, and it allows you to manage still well your portfolio. That's the other big component. I think that with all that data that you have, that smarter, faster, better underwriting decision making has really improved, dramatically improved, which can also be viewed negatively because you could argue well that the system will tell you how far you can go off in certain decisions because you will be guided, you will get indications, and you will have to have good arguments why you have decided to overrule a rules engine, for example. So it's still, I think very interesting, fascinating because of the potential, but I can also see those who are a little bit concerned and maybe don't know yet enough or don't know what to really think of this new age if you want. So, and I'm not sure that all of them are ready to just jump over and say, Hey, let's do it in a new way.
Michelle Gouveia (13:58):
I think there's an interesting dynamic or what I've witnessed or heard in conversations is very much what you're saying, Terry, but I think people generally fall into one of two camps. The resistance to ai, which is a little bit of what you just talked about, where it's the fear for what is this due to my job? Do I need to learn a new technology at this point in my career? And then there's the folks that are all for it because they do recognize the value of the taking out the reductive work, and maybe it's generational. You come from your day-to-day is evolution of technology. And so it feels that if you don't have AI support, that's a miss. I'm curious to get your deeper, is it generational? Is it by function or job? Do you think that some of those need to shift or will there always be people in two camps as part of this?
Thierry Daucourt (14:56):
I think, well, if you're generalized generational, maybe I think it's individual. You as an individual decide whether you embrace it or not, whether you are against it or just on the sideline and watching it. And you decide once you know a bit more, whether you jump on that wagon and want to be part of it or whether you are enthusiastic and you say yes to every new innovation and basically want to be the first testing it. I think it's individual. Some of it can be with age, can be with experience and all of that, but I would not say that every 25-year-old is totally excited about the new potential that AI is offering. I'm not sure about that to be honest. I think that some of it is really knowledge the problem. So the longer I've worked in my job as an underwriter, and the more you throw stuff at me that we could do differently and you show me things that we could be doing, the more I will hopefully reflect on it and look at it.
(16:01)
And then either I take it as a negative thing or I take it as a positive thing, which allows me to learn something new to I see the potential, I see the improvements that it could bring to my daily work life, and therefore I'm jumping on the wagon. But I'm with you. I think that when you start a project in particular, you have to make sure who are the people that you want to be part of that project. You need believers, I call them believers or fans, but you also want to have some critical people with critical voices, not destructive people, but constructive, critical people. And if you're able to convince those people that this is really the right thing that we're doing here, they will be able to promote it internally and bring all the others along in that journey, is my view.
Michelle Gouveia (16:52):
You said one, oh, sorry, Tom.
Thierry Daucourt (16:54):
No, go for it.
Michelle Gouveia (16:55):
You said one thing there, Terry, of continuing to test it. And so the question I asked was very much focused on maybe the individual adoption, but maybe zooming out a little bit, generally adoption across insurance, wanting to get your thoughts on that. I think there's similar thing, two different camps, the desire to not fall behind to leverage AI to be competitive in the market, but this hesitancy on fully implementing and deploying an AI solution in certain either customer facing or key decision making functions versus others. And so generally, where do you think based on the regulatory and the compliance environment where carriers are maybe testing or deploying ai, how do you think about or how do you think the industry, where is the industry at this moment in terms of AI adoption and success?
Thierry Daucourt (17:56):
It's a good point. So sharing a story with you, many, many years back, over 10 years now, I bought an electric car, which was claimed to be self-driving. It is still not self-driving, which gives you all the arguments for those who are not fan of new technology that actually sometimes we promise things, but it doesn't really deliver in the day-to-day. And individuals who go to work into the insurance industry, well, they have a life at home and they have their experiences at home with technology, with new technology. It's either their car, it's their smartphone, it's whatever their search engine that they're using every day, their chatbot experience with the telecom provider. And some of it, I have to say are disappointing to be honest. Now on the contrast then I've seen very well performing machine learning solution, but these are well-trained for one specific need tools that do a terrific job and that are well accepted with no objections by nobody.
(19:05)
I think that we have a little bit to be, well, not careful, but I'd rather talk about technology and AI as part of that. It's the big boost right now. But I think that whenever we deliver something, it has really to deliver and to solve the problem it has been built for because otherwise you will just create negativity which you don't want to have. And I think there we go back a little bit. In today's world, so many things are going so fast, but some of it is out there maybe too early. And so it disappoints people. And if you're critical, it's just a confirmation for you. I always told you it doesn't work. If you're a fan, you will find a good answer to say, well, actually no, no, it's really terrific because look, this and that. And I think in insurance, if I look at pricing, for example, if I look at rules engine, they work because they are very clear, they have a clear job that they need to do and it works really well in that area.
Tom Wilde (20:10):
On this topic, I've had this sort of question in my mind for a bit about the sort of basis of competition. If you look at what McKinsey or BCG has written on these topics as a company, you sort of, in any industry, you have to have a primary basis of competition that you're going to use. It could be low cost provider, it could be customer service, it could be quality of the products. There's all different basis of competition. I've wondered in commercial and specialty, if anyone has tried or if it's viable to think about the basis of competition as being obsessed with the employees doing the work. That's your basis of competition, meaning that everything you do is based on making the underwriters, the claims adjusters, making their job as slick and powerful as possible, their experience doing their job as amazing as possible. What do you think of that idea? Have you seen people try this? Is it a dumb idea? I think so often what happens is they're at the tail end of these initiatives and it's sort of like, oh, here's another thing that we have to learn how to use, and there's an eye roll, right? And so what do you think about that idea?
Thierry Daucourt (21:25):
Well, I think it works in commercial line because it's a business that I don't see, at least for the next couple of hours, to be disrupted completely by technology and managed by technology. So to answer your question, I think that commercial line, the beauty of it, and we spoke about it in the beginning, is it can be really complex. It's not only the big companies that are complex research lab can be a very complex, although they are SME character and large organization are complex, international exposure of that. So the human negotiating with the broker, with the customer, the human being involved in that digital supported process is still key. That's why I think a good strategy is to make sure that you really have kind of a customer-centric approach when you deploy your things. And in this case, the customer is your underwriter, your claims manager, your accounting people, your finance people, your group risk management people to make sure that you address their needs because they will stay around for a while. Whereas I think that in motor, in personal lines, maybe a little bit a different conversation.
Tom Wilde (22:43):
I think that's right. Yeah. Yeah. I think there, just to build on that point, I think in personal lines, I think the ship has sailed, right? I think the injection of mobile and now AI into the relationship in personal lines is we're never going to go backwards on that. I think that sets the stage for an increasingly automated experience with a customer. I don't see in the near term anything like that in commercial lines. I think there's, it's just too distributed a market. There's too many players with too much power throughout the ecosystem. The brokers, the carriers, the wholesalers, the MGAs, nobody can dominate the other, you know what I mean? When they can't dominate the customer either. The way in personal lines, what I mean by dominate is set the rules of engagement, right?
Thierry Daucourt (23:37):
Absolutely. And I think that what we really have added in commercial line over the last years is services. And so every insurance company now ideally would like to add services to insurance product just for the stickiness of the clients,
(23:50)
But adding real value. So I know of examples where satellite images, earth observation, all of that vessel tracking, all IOT, all of those things are really adding value sometimes for the customer, many times for the underwriter to better understand the risk. Again, we're back to smarter, faster, better underwriting decisions. And so all those services have become a key component enabled, again because of technology. And so I think that commercial line is really the place where you can almost say is a big lab environment where you can test and try a few things. The challenge with services though is how many customers are really willing to pay for a service. Still an open question, but if the services adds value to yourself, to your underwriting, then it's a different discussion or conversation. And a point maybe that I would like to raise is what we underestimate. Sometimes I see solutions being delivered to address a pain point in underwriting, but actually there are collateral benefits as I call them for group risk management as well. So we tend to underestimate what it means if we're adding value in an underwriting process or a pricing process or a risk selection process, because that could actually be very interesting for claims or for group risk management or compliance or whomever. And we tend to forget a little bit those people that are not part of that core underwriting process or claims process.
Michelle Gouveia (25:33):
So Terry, we've spent this last little bit of time talking about the challenges or the benefits for deployment across the different types of insurance, but given your experience and your background, I'd be curious your thoughts on how big of a factor, if any, does culture and national versus an international or a federated versus not federated company structure impacts technology, deployments and appetite to do so?
Thierry Daucourt (26:10):
Huge. So I think huge. So you have the culture, the country culture, language, all of those things. You have the company culture and the way they are organized, I think they play a crucial role. It'll always, for someone like me who's native language is French or German, I should figure that out one day, but it's one of the two, not English. I will always prefer to negotiate a contract or to work with someone who, my language,
(26:45)
I don't like it obviously to work in English, but it will be easier also. And particularly if the impact is on different levels of the staff in the organization, it's always better if you have a team that is able to work in German, in Germany and deliver that solution in Germany. So I think that will always be an advantage if you're able to bring together teams that are able to communicate in their language with their cultural background, that will make it easier. It'll increase the complexity, a level of true higher up on a group level because their English is the language probably in almost every company that I know of, at least in this part of the world. And then the other point is the organizational model of the company. You have totally decentralized models where decision-making happens in that country or even in that business unit.
(27:38)
Increasingly, I think the trend is that the group wants to know or has rules, you should not implement something before you have talked to the group and we need to validate what you're doing. Or the group has a view of what they would like to be deployed and ideally leveraged across all entities, all businesses, to maximize basically the results or the benefits of such a rollout. So company culture matters a lot. Size of the company obviously matters then a lot. And then I think culture and language and how you're putting your teams together matters also a lot, definitely from my point of view.
Tom Wilde (28:22):
Great. Well maybe bringing the conversation to a set of good conclusions for the listeners, what would be your top three recommendations on navigating the pace of change here and integrating AI into these processes for today's commercial insurers based on your learnings and your observations?
Thierry Daucourt (28:44):
So maybe the first one would be for the management, unlike in previous times, you need to understand the topic and not just tell your people, you have to use ai. So it's important that you don't stand in front of a team and say, okay, we're going to do an AI project. Goodbye. I'll see you tomorrow. That's not going to be good enough. You have really to be savvy on the topic and understand also the ethical challenges of it. People have questions, regulation, governance of all of that. So I think that's one aspect that I would highlight. The other one is probably are two words. One is balance, and one other one is time balance, because I think I use always the analogy with sports athletes. They don't run a marathon every day because otherwise their resources will be killed after a month or two and they would never do a best time ever anymore.
(29:45)
So part of their career is to be able to train at low intensity, to do a race every other month or so, to go to massage, to sleep well, to do whatever, to be balanced basically. So don't be extreme also with your solutions because Extreme is not going to be able to align the majority of your people to be part of that and time because I think that the machine is much faster than the human. And deployment at a human pace requires you to be able to communicate properly, to really gain the attention and the willingness of your staff to embark on that transformation. I have never seen a project that happened in two days. It's just not possible because from I need to understand what we're trying to do, then I need to understand what we want to do because we don't want to just change the status quo.
(30:47)
We want to leverage that technology and disrupt the way we do business today. So that takes time. And allowing people to have time means they can talk to each other, they can exchange opinion, and that's a way to create them a base basically for something that will be accepted by everyone and therefore you are winning all that time that you invested before because the implementation will be a no brainer. If everybody felt that they are part of that journey, then they will not say no the day you say, okay, now we are ready to switch and let's move on to the new model on Monday. I think they've been part, so they will join you and they will do it. I think that would save a lot of projects from my point of view.
Tom Wilde (31:36):
Great. Great insights. Well, we've been talking to hort, sharing his wisdom from decades of experience in the insurance industry. A great conversations here. Thanks so much.
Michelle Gouveia (31:46):
Thanks for joining
Thierry Daucourt (31:47):
Us. I thank you. I thank you both. Thank you very much.
Tom Wilde (31:50):
Very good. Well wrapping up again, I'm Tom Wilde, your co-host.
Michelle Gouveia (31:53):
And I'm your co-host Michelle Gouveia,
Tom Wilde (31:56):
And you've been listening to another episode of Unstructured Unlocked.