17 mins read
01
.
11
.
2022

The Role of Culture, Digitisation, and Embedded Insurance in the Evolution of Underwriting

by Juan de Castro, COO, Cytora

This is a shortened version of the Making Risk Flow podcast, episode “The Role of Culture, Digitisation, and Embedded Insurance in the Evolution of Underwriting” with Thierry Daucourt, Head of P&C Commercial Lines European Markets at AXA.

You can listen to the full episode here

Juan de Castro: In previous episodes, we talked about how to drive successful digitisation in commercial insurance. Today, I’ve got the pleasure to talk to what I would call an industry lighthouse. Somebody who’s, for 30 years, had executive roles in three of the largest global insurers and led underwriting teams in different geographies. In his current role leads commercial lines for Europe at AXA.

Thierry, it’s really great to have you join us today for an episode of Making Risk Flow. As I said, those 30 years of track record are quite impressive. Can you start by giving us an overview of your background?

Thierry Daucourt: Of course. Thank you for having me. You mentioned 30 years, obviously, it’s a long time, so I will try to make it really short. Commercial lines from the beginning, well, not exactly, but most of the time in commercial line P&C, and I had also the advantage of going through almost all the functions in the insurance. I worked from bookkeeping to accounting to HR to claims, and underwriting.

Then with time I moved into the international business space, so the larger mid-sized corporate business and held roles and jobs in that shop for a long time, I call it the old shop. I also, worked for XL Catlin before it got acquired by AXA, and was at AXA when the acquisition of XL Catlin was done. I started the integration team and was happy to contribute there to the acquisition that we did at the time.

I spent about 13 or so years in markets outside of Switzerland. I’m Swiss national. You recognise a Swiss by the fact they wear two watches, a combination of, not everything which is old is necessarily bad, and not everything which is new is necessarily the solution. I love old watches because it’s continuity, it’s craftsmanship, it’s solid, it’s standing across time, but I like the smartwatch. It’s innovative, it’s the new world, and it’s technology-driven. The combination of the two, I think, is an ideal mix, because you also need to be able to move the people along any transformational project that you want to do. I found it important actually more than relying on technologies, to have a good plan on how you deal with the people and the cultural change that you’re introducing.

Juan: You’re totally right, often in insurance we think “okay, it’s the future, everything is digital,” which is probably not, right? That balance of the two, I think, is an area we should definitely explore in this chat. Additionally, culture is one of those core components of that evolution. What’s the role of culture in all the contributions you’ve spoken about and how are you dealing with that?

Thierry: Culture is almost everything. Because there is a culture by country, by where you come from. It’s so multi-faceted, the culture really will determine whether you’re able to go through transformational projects with the team. It’s not helpful if you have to look in the mirror and nobody is following you, you’re not going to get things done. No chance whatsoever.

If I think about the projects that we have going on, pilots, it’s important that people can talk their language, local language, local environment, local setting. It just helps a lot to make sure that they really feel comfortable. English is the second language for most of us, and it’s fine to do business, but once you go really into the detail, it’s so much more effective to have teams working in their local language.

That’s why also I try to step away from those places where I do not understand the language. It’s fine to supervise it, but it’s not good to be involved deeply locally. Culture is important because human beings say they like change, but they don’t like change, they like routines. Most of us, after the pandemic, go back to the routine we had before, which is a funny thing, because we stay in the same traffic jam that we were in before, although we said we were going to change it, we did not.

Like I said in the beginning, not all the old stuff is bad, and not all the new stuff is good. You need to find compromises. Then it’s about the speed at which you want to do the transformation, which is important. If you are able to transform continuously at a pace where people follow, wonderful. If you are five years ahead and nobody is following you, well, it’s frustrating.

Juan: You mentioned people don’t like change, I think what they don’t like is change where they don’t understand the value. I think the solution to that often is, “Okay, how can you prove this is valuable for them?” This is driving change with the underwriters in mind and making their lives easier and better.

Thierry: That’s true.

Juan: I think everybody loves that type of change. I guess it’s about how you drive it.

Thierry: Absolutely. You have to show it, do good storytelling and then deliver. That’s the challenge. If you promise too much and you only give them 5%, they will be obviously deeply frustrated. You should find the right sales messages and delivery.

Juan: If you can give us also an overview of how do you think that balance of the old and the new informs your vision for commercial lines in AXA Europe?

Thierry: Sharing the vision statement I use when I work with the commercial line community in Europe is really to be a market maker. I like that because in one sentence it captures the fact that being a market maker, I like it more than to be the biggest or the fastest or whatever. A market maker is someone who knows what he or she is doing, is confident in his own capabilities, has a view of where the business needs to go, and is not waiting for market trends to happen.

That’s why I actually like this market maker attitude or behavior. It’s being humble, but self-confident in what you’re doing, and not necessarily all the time waiting for another big player to move because it’s up to us. I also don’t like when people tell me, “Well, the market is doing this, the market is doing that.” I say, “Well, we are the market. We are part of the market, so don’t blame the market for anything, just try to find the right attitude and behaviour in the market.”

Juan: When you say being confident, is that about what type of products to build? Is it about what type of clients to underwrite? How are you confident?

Thierry: It’s a good question. Actually, it’s almost everything. It’s to have a strategy, to be clear about the environment, the economic environment, the societal changes, the technological evolution, and the pace at which technology is transforming the business. It’s not only internally focused on understanding what’s going on in your company, it’s actually much more. Do you understand what is going on in this world today? Then what is the role of insurance, particularly in the 21st century, what is the role of insurance and what is the role that we want to play as a market maker?

I think the product is one aspect of it, I would say, but it’s not the only one. It’s how you embed and integrate capabilities that are coming from third parties InsurTech, FinTech in particular, but not only. That’s one aspect of it, how can you develop the new generation of products going towards computable contract solution at some point? It’s all of that. Self-confidence is really, knowing your book, understanding your capabilities, making sure you have the know-how and expertise that you want and need to have for the strategy that you’re driving, and most importantly, I think in today’s time where we switch quite fast from one strategy to the other, my philosophy is giving the strategy a chance to prove it was wrong because it could be right. Don’t switch after three months, the first quarter results are not good, and then you move over to the next strategy. Strategies are about longer-term initiatives and not short-term ones.

Juan: Which interestingly enough, goes back to your point about confidence. As you said, make a decision, be confident about the strategy, execute, and then analyse what did work and what didn’t, right? How do you see underwriting at the AXA five years from now? How is it going to feel different than it is now?

Thierry: In many dimensions, and some of it is visible already today, I think if I look at it, we will play an even more important role in ecosystems, which we orchestrate or are part of. Ecosystems, like mobility or smart home, the cyber, these are all partnership approach where actually it’s only with partners that you can really tackle the opportunity. It’s not a challenge, it’s an opportunity.

If you look at cyber, it’s hard to imagine that you can assess the risks properly and offer after a breach and pre-breach services on your own, it’s about working together. With the auto industry, it’s about collecting data from the cars to be able to provide better services. It’s the ecosystem approach, I think, is something that we will see more of. We will generally also, I think, generate more income from services, from fees, not only insurance premium, which as you know is less risky than the insurance premium side of it, and I think we will continue to be a trendsetter around the ESG topics, and environmental ones in particular.

AXA climate today is the lighthouse, I would call it, for what we are doing around the environmental topic. It’s not only with insurance products but actually, we’re training, building up resilience, consulting smaller and larger companies to be able to navigate through those difficult times or challenging times that we have today. In general, I think for the underwriter, I would almost say we will have a fully digitised process where the human in the loop will remain but with tighter controls, with more discipline in underwriting. Because increasingly, it’s not good enough to look back to project the future and say, “That’s how much it should cost.” It’s about actually understanding what’s going on today and what are the scenarios for the future. That requires really a good oversight of your underwriting teams who have that pen in their hands and basically can engage the company to very high limits and assist them and help them with technology.

Having the right information at the right moment at their fingertips, I think is crucial so that they can make smarter, better, faster underwriting decisions. Further down the road, we will increasingly see a bit of computable contracts coming up. Computable contracts with a human in the loop. I don’t see the machine doing everything but I clearly see a need to implement computable contracts.

Also, because there is a wake-up call with the pandemic that has actually demonstrated that it’s not easy to find out what we have insured in our portfolios if you want to dig into the really detailed coverages, endorsements, and clauses that you have. Particularly when you work a lot with the brokers as intermediary is not your own products, it’s products from the market, and there to have the technology to help us to first of all, better understand and run through scenarios about our own portfolio and what we have in our books but also down the road to be able to have modular products that are easier to assemble, to develop solutions rather than to just sell a product. That’s what computable contracts or with contract lifecycle management as the philosophy that will be able to achieve it.

I see the next years, like the last two years, an accelerated transformation compared to what I used to live in the past years.

Juan: That is very interesting. I think you’ve mentioned ecosystems, income from services, ESG, the future role of the underwriter and computable contracts. On the ecosystems, do you see it as two components? One is embedded insurance and the other one is about partnering with third parties around data services. Is that how you’re thinking about it?

Thierry: Yes, absolutely. Embedded is another big topic, if not a hype, and embedded is different from the ecosystem. Embedded is an idea and maybe has even more potential in a retail space where I buy an iPhone, the insurance is packaged with it. I have a credit card, all of us have a credit card, and there’s insurance provided by it. We mostly don’t know and forget it but it is there.

Embedded, we’ll see how far it can go. Because embedded will actually automatically add the insurance for something that the customer is buying which is for them emotionally much more interesting and is willing to spend time. You don’t sleep in front of an AXA office when there is a new product. That’s not going to happen. Therefore, embedded would have to see how far that can go. With ecosystems, really accepting, understanding that there are third-party providers that are actually much better than you and maybe they also have the control or leading the hand of the customer where you are providing part of a solution and the other part comes from third parties, from partners basically.

We have in health a big initiative. The health platform in a partnership with Microsoft where we do exactly that. We are also working on different ecosystems like cyber. We’re working with a third party that is really good at assessing a company before we insure the company is something that makes total sense rather than us building up that capability.

Juan: To pick on that cyber example, is that a third party, more a technology data provider that helps you inform the underwriting or is it more a services upfront?

Thierry: It’s data and it also services, particularly after the breach. If it’s about reinstalling the capability of the insured to go back to business, it’s servicing capabilities, making sure they can get up and back to their business as soon as possible. On the data side, it’s more data-driven in the assessment of the risk profile. After that, actually if we go beyond cyber, we have three major streams right now. One is enriched data, the other one is unstructured data, and the third one is contract lifecycle management computable contracts. We think that we need to play in the three of them in order to improve what we’re doing and what we’re delivering.

Juan: Yes, which I guess is also particularly relevant to the line like cyber where there’s a high volume of submissions and risks to analyse, where typically appetite is strict, right? You want to be quite data-driven to understand what percentage of those submissions you really want to be quoting, which ones might sit outside of your appetite. I assume you’re planning on doing that through the use of external data and the right process.

Thierry: That comes almost a side benefit to actually some of those external data providers, which is the democratisation from IT. If you’re able to have a low code or no code solution, which allows you to react more dynamically on impacts on your portfolio, whatever they are, that gives you that freedom to be much faster in adjusting your underwriting appetite.

If you see things developing in the desired way, fine. If it’s a not desired way, well then you act traditionally, today, you raise your hand, you go through the planning process and every year you have to line up with all the other businesses and basically make sure you have enough IT capacity to do whatever changes you want to do on your system. Some of those capabilities that we see emerging or giving us that freedom to actually act quite rapidly and manage the funnel, open it up or close it more depending on the line of business, the situation, and whatever is happening.

Juan: Yes, and I think we see that with a number of our cyber clients, which is this capability of analysing quickly any changes in the market or any changes in profitability of certain segments. Being able to then almost in real-time change your appetite or your triage rules based on that. I think there’s a third component which I’m starting to see more is also using those types of data enrichment, so things like port vulnerability scanning not just at the point of underwriting but using that external data almost as a monitoring exercise?

Thierry: As we’re going down a road where actually you have an almost continuous overview and control where you constantly are able to compare your own data with external data and you’re constantly monitoring potential risk profile changes of your insured, let’s say, at least once or twice a year because you have renewal piece, you have your annual renewals going up.

I think before the renewal, what we’re doing with some of that enrich data work is obviously to compare our own data or information that we have received from our distribution partner with available external data and see if there are gaps, things that we need to clarify, which allows us to be actually more proactive and either tell our own driving teams, “Hey, you should check this and validate whether this is right or wrong,” and accordingly make changes to the contract or not, to also avoid for the insured, under insurance or wrong insurance in some cases, so that that dynamic portfolio measurement, monitoring almost, is something that we see coming.

Juan: Actually that’s a full circle. Not only you can change your underwriting reaction to that, but also you can even go all the way to the clients and say, “Hey, we’ve identified this issue. If you fix this, we would be happy to renew on these terms.” It’s also you’re going all the way to providing a being helpful to the insured too.

Thierry: It’s also a way to automatically quote business where you maybe have only the address and the name of the insured. You almost don’t need much more. You have traditional submissions. Sometimes they are complete, sometimes they’re old. It’s a way to complete submission information that you have with the available data and actually be sometimes more accurate. Again, helping you to do smarter, faster, better underwriting decisions.

Juan: That’s very exciting. One of the other area you mentioned was income from services or risk-free income, and as I say, line of growth. Can you give us a couple of examples of what have you got in mind?

Thierry: We already have a big team in risk consulting, an entity doing risk consulting services and that generates a fee. Some clients pay for that and they are not always insured by us. Over a year ago we launched the AXA Climate School as an example and we have lots of clients subscribing to those classes, short videos, very professionally done where actually employees of a company are attending those classes to do knowledge transfer, to educate people, to help us or each of us also to understand what is actually climate change, how does it manifest itself? What can you do, what should you be doing, where can we inform yourself? It’s a very nice class that we’re offering and that we’re selling to our customers is another example which is live already.

Juan: Definitely, it’s a risk-free income but as importantly, it provides a great service to your clients, right?

Thierry: Yes.

Juan: You moved from being a provider of insurance to being an educator to those clients.

Thierry: Absolutely. If you look at insurance, some people think insurance is not emotional. I think it’s hugely emotional. We are like providers of services of a hospital. Ideally, we try to avoid our clients having to go to the emergency services but if they have to because there’s an incident that happened, we want them to be able to go back to work as soon as possible. Ideally, we work on prevention, resilience, help them and a part of that is education, particularly in today’s time with environmental and societal changes.

Part of it, obviously, is to pay claims but make sure that they will go back to business as soon as possible. Sometimes we tend to forget in commercial minds, although we insure a brand, what we ultimately insure are people doing or not doing certain things every day. We insure people and their doing or not doing and the consequence of it. It’s very emotional because the more you can help them to avoid going to a hospital, the better. If they are in the hospital, well, the faster can get out and go back to business will the better days for them as well, which is I think a great way for me to visualise a little bit the importance of insurance in today’s world and the economy.

Juan: You see that often in the claim, right? Paying your indemnity, for sure, it’s almost a hygienic factor, right? It’s the service. It’s how do you help them get up and running. If it’s in a property line of business, you might be giving them accommodation, but then in a commercial line or in a cyber event, might be just enabling them to keep on doing business after a cyber attack.

Thierry: Exactly.

Juan: You mentioned the future of the underwriter role. One of the analogies that I found really interesting is the difference between human-in-the-loop and human-on-the-loop. This human-in-the-loop is where a human needs to do all the activity to process a risk and we are all familiar with the underwriters having to double key information and doing a lot of manual activities, and that’s a human-in-the-loop. To do every step in the process, you need a human versus human-on-the-loop where technology processes the risk but you’ve got a human making those types of judgments that only humans can do. Is that how you think about it?

Thierry: Absolutely, yes. Actually, it’s a good explanation. First time I hear it and I will keep it in mind. I like that. You have both, you have in-the-loop, that’s more the administrative side. It’s improving the MLs, the algorithm. Algorithms, the better we train them the better they will be, like an athlete. You need that human-in-the-loop to sometimes say, “This is wrong. You’ve got to go into this box,” and with that it learns.

The on-the-loop is really supporting an underwriter to make a decision, to provide the advice, to make up their mind on the risk to ideally keep them away from all the administrative tasks, filling in spreadsheets or filling in an IT tool. All that can actually be done by the machine guided by human-in-the-loop, but then the underwriter can start to really work.

Juan: I think that that approach also allows an underwriter to be spending less time on individual risk and almost zoom out and look at the flow of submissions. To some extent, your flow of submissions represents an industry. That’s where you can understand trends and where are you off market,

Thierry: With one caveat, we differentiate between mass market and single-risk underwriting. An SME, the lower part really is a mass market where we talk about exceptional underwriting. You manage your sales funnel, you say what I like, what I don’t like, and then you look that you have a straight throug processing process for 80% of the business, but the remainder will be sort of a human-on-the-loop validating the risk.

Single-risk underwriting is really the Fortune 500 business, they’re the large corp. Every risk is you look at on its own merit. Those are two different fields within commercial.

Juan: It’s really helpful to get to this level of detail about how you’re thinking about each of these areas. I think it summarises really well. Thierry, thank you so much for joining me. I truly enjoyed the episode and get your views on the industry.

Thierry: My pleasure. I enjoyed it as well. Thank you very much for having me.