Live Edition: Digital first operating models with Markel and Beazley

by Juan de Castro, COO, Cytora

This is a shortened version of Making Risk Flow podcast, episode 10. You can listen to the full episode here

This episode was recorded live in London, Juan had a fantastic panel discussion bringing Nic Brown, Divisional Director – Broker, from Markel International, and a star of a recent episode James Wright, Head of Technology at Beazley Digital. They’ve chatted about the challenges that major insurers face and the solutions they adopt.

Juan: Thank you so much for joining and thank you, everybody, for coming. Let’s start with brief introductions.

Nic: Nic Brown, I’m the Divisional Director for the broker channel in the UK at Markel International. I’ve been in Markel for five years, I was brought in to consolidate various different acquired brands, create a single brand identity, and get that brand identity to be recognised in the channels that we distribute and ultimately drive profitable growth.

James: Hi, everyone. James Wright from Beazley. I’ve been at Beazley for 18 years, which is a long time, but more recently I’ve been a part of the leadership team at Beazley Digital. We’re focused on the global SME commercial business. My journey to this role has been one from a technology background, originally in the London market, then I went to the US for eight years, got very close to products and distribution in the US. I have a real passion for how we really make money and service brokers. 

Juan: Perhaps we can start with an overview of your vision of what you’re doing at Markel, then do something similar with Beazley, and perhaps we’ll find kind of points of commonality across the two. 

You’ve got quite a strong vision for how you’re transforming the broker channel at Markel which is quite exciting to see. Could you share some of the highlights?

Nic: In Markel, in terms of our business in the broker channel, we’re very focused on managing the specialisms that we work with. So looking at industry sectors and trying to go as deep as we possibly can to understand the customer that’s in there, the risk and ultimately trying to deliver sustainable solutions.

Sustainability, managing risk and genuinely adding value to customers is key for us. We’re very lucky, we’ve got our own law firm, we’ve got a specialist tax division where we look after and support businesses with their day-to-day challenges, with the Federation of Small Business we are helping 200,000 SMEs. We try and pull all that together to really be able to identify the customers in each sector and bring solutions to them. That isn’t just about insurance. I mean that’s obviously the important piece, but it is about the ancillary services that we can offer that either help manage risk or provide support in different kinds of challenges that the traditional SME has in the UK. Our strategy is all about driving growth and driving profitable growth out of that, but ultimately getting into new sectors in order to facilitate that growth. We want to be a $1,000,000,000 business in the UK.

Juan: Often when you talk about growth and how to enable that growth, there are two themes that come to mind. One is about capturing efficiencies that drive growth and the second is about how a better broker service can accelerate that growth. Perhaps you want to touch on those two things.

Nic: I think when we started we didn’t want to force brokers into a different way of working because that’s what is best for us. Various insurers in the UK are trying to save money, so digital solutions are a great way to ultimately get brokers to self-service and allow that work transfer. And that’s OK if they want to do that. 

We’ve definitely tried to support the broker on that journey, but we’ve grown and we are now writing 500% more new business a year than we did four years ago. The reality is that about 30% of our monthly revenue is coming from new business and the challenge that gives us is delivering 40% year on year. You can’t recruit 40% more people to be able to do all of the support whether it’s just getting a new business quote out, managing claims, managing the operations, managing the renewal activity and providing the same level of service that the broker requires.

So we’ve had to look at things differently, not through gaining efficiency as an operational requirement to save money, but we have to make efficient decisions to facilitate growth. Otherwise, we wouldn’t be able to service the business that we’ve brought in. And that’s why the relationship with Cytora has been incredibly helpful for us, specifically on the new business.

Juan: You’ve also touched on a slightly different operating model and flow, being something that you’ve driven in the last year, exactly how are you enabling underwriters to be able to write 40% more business without having 40% more underwriters. But also if you could tell us what the flow looks like on the high level. 

Nic: For me, there are two different types of business that we operate in. In our environment, there are transactional risks that are not that complicated and don’t need to go through a traditional underwriting process.

And there are more complicated risks that absolutely need to go through the lens of the underwriting view. We’ve tried to do is get our underwriters to spend more time underwriting those complex risks, and create a process that provides a better service for the broker on the more transactional activity.

I’m happy to go through that in terms of where we’ve worked with Cytora. First thing – auto declining risks that come through the door, rather than having someone look at them and spending half an hour on something that’s never going to be a pound value to this organisation.

But to get to that point, we’ve had to look at all of the ways in which business comes to us.

Some of that is structured in terms of data, and some is unstructured emails from brokers. We use Cytora to turn that into an automated process to reduce data entry, which in turn gives us a system that enables us to work and create our workflow for underwriters off the back of that. 

It also enables us to put algorithms around prioritisation, as well as the auto declines so that we can spend more time on the things that are going to ultimately result in a better outcome for us.

That goes against every bone in my body as a salesperson, I want to treat every single opportunity equally because I want to win them. The reality of the growth story is that we’ve got to be able to be more structured in the way that we do that. Cytora and the use of CRM together have been transformational for us in terms of having a slick process.

Juan: I think the auto decline specifically, summarises really well how you’re not looking at pure efficiency because you could say to the clients that auto decline is an efficient way of making sure underwriters are not wasting time on out-of-appetite submissions. I think the perspective you take is that auto decline is even more important to get back to the broker right away to let them know that you’re not going to write it. That’s the difference between pure efficiency and looking at it from a broker service perspective.

Nic: Without a doubt. We had our US speciality team over about a month or two ago and they write billions in terms of the kind of markets they’re in, not too dissimilar to us. The question I was asked was, “why aren’t you managing the broker to stop giving you those businesses in the first place?” – Well, because we might change our mind.

The areas that we’re in today, by definition won’t be the areas that we’re in tomorrow. And every auto decline that I record is an opportunity to pipeline it for when we are ready to do it or we do change our appetite or we have a solution to that sector. I’m building a process, but also storing that for future benefit is huge for us in terms of our long-term development.

Juan: Yeah, exactly. Let’s switch to Beazley Digital. I have been fascinated also by your journey in the last year or two setting up a whole new business unit. You’ve changed everything from processes to technology, you’ve got a different management board and board of directors – it’s quite incredible.

Perhaps you could give us a sense of why you set up Beazley Digital, what were you trying to achieve?

James: This might resonate with some of you that are in the incumbent or legacy organisations. We started the digital journey maybe seven years ago. We had pockets of digital trading going on across our organisation, some of it was working well and some of it wasn’t working so well.

But more importantly, the way we were showing up to our brokers was really inconsistent. We stepped back and thought we could do much better than this. That, on top of the fact that we had a fragmentation internally on efficiency points and we didn’t have a crisp enough focus on what SME means to Beazley in terms of taking those specialist products and making them work really well for small clients. So we took a bold decision to create a new division, carve out that SME business and place it there.

The bolder piece of that was making it completely cross-functional. A lot of us probably work in organisations that are broker or underwriter-led. We decided to have technology operations and underwriting in one team, they have shared goals and they work together on achieving outcomes. That’s probably one of the biggest challenges that we faced in the first six to eight months.

I realised quite quickly that whilst underwriting colleagues might have one idea of what success looks like, it’s quite different to how operations look at the same thing. Trying to get people to use the same language, and to be aligned to the same outcome was a massive challenge. And that was really the set-up phase.

We are now trading as a separate trading team at Beazley and we’re pursuing building up a $500 million business in the next three years.

Juan: Which I’m sure you will do. Every time we talk about you having aligned all your teams against very specific goals and you do actually by using OKRs. I think one of those is 80% straight-through-processing within Beazley Digital which is the goal that goes back to efficient processing of those more homogeneous, simpler risks.

Perhaps you could share how does a new business workflow look like for products like cyber which is not easy to underwrite but it’s quite a high volume.

James: Yeah, sure. If you think about the workflow as a set of capabilities, it’s not dissimilar to how an underwriter thinks. You’ve got that first stage, getting the information about the risk. It could come from an application form, but increasingly it’s coming from other sources as well. We use a service called KYND, many of you may be familiar with it, a classic vulnerability scanning service. So bringing that data in, getting confidence in that data and then moving it further. 

The next stage is around triaging that risk, and we’d love to win all business as well, but also we’re primarily targeted at writing profitable books of business. So that means carefully selecting risks. For example, if a client does not have an MFA solution, we would classify that as not an insurable risk. Now, we also incentivise to try and remediate that, so a big part of what we’re doing in that flow is providing back some degree of service. So we are not straight declining but we go back to them and let them know they’ve got a complication in their security hygiene. Once they resolve it and they come back to us. So that’s ingest, enrich and triage. 

The next phase of that is much more speaking to the transactional part. Many of us here have been probably been building a very automated policy system or policy lifecycles platform. We call that a product engine. And I really would classify three things as a product engine. That’s the rater, underwriting rules, then the forms that go with that and that sit under a component.  The flow for us as a component perspective is modular, but it does work in a harmonious way. It has to start with high-quality data. And we did some sort of analysis on common application forms, and although underwriters and brokers would say that’s the gospel, we think we think the data accuracy of the forms received today is around 60%, which means 40% of it is wrong.

When we come into talking about portfolio analytics or micro-segmentation, that’s fundamentally flawed because we do not understand what that client looks like. We are using Cytora here to extract the data from documents, bridging that with data that we trust, not just to drive efficiency and broker service, but to drive better analytics is ultimately where we’re trying to get to.

Juan: And I think that that combination of data that only exists in a submission, with third party data – in your case is KYND for port vulnerability, but it could be flood scores – with your internal data is one of the things we’re doing with you guys. And it’s also about automating the clearance process. All those three sources of data are needed to be integrated to be able to start making automatic decisions on those risks.

James: Correct. The other challenge that many of you will probably have is that distribution is increasingly becoming omnichannel. So whilst we might think about portals, we’re also moving into market hubs. We’ve got some direct API connections with brokers, and then we’ve got our email channel as well. Meeting our brokers where they want to be met is really important because just putting out a portal to make the broker do the work for you is not going to work. It’s not a long-term strategy.

And there are pockets of it where it will be successful, where the broker really wants a specific product. But we’ve got to think beyond that and think across all those different distribution channels.

Juan: Absolutely. But both of you touched on that service to brokers being at the core of profitable growth. So where do you see the opportunities?

Nic: Highly digitised brokers would want a very specific, clean, easy to use, oriented user experience, minimising the amount of information that is required. They are going to take full advantage of that. But a large chunk of brokers is still quite traditional in the way that they want to operate. Many insurers are pushing them down a route in terms of that self-service as a piece of work transfer are not doing that for efficiency. They are doing that because it’s cheaper. 

I think that for me at the heart of what we do, is figuring out how we are going to add value to the broker, how is this going to improve the way that we manage risk, how is this going to enable us to transact more effectively? It sounds like it’s also something that Beazley is focused on. And ultimately the better that is, the more we are going to win which is what it’s all about. So for me, the service experience has to recognise that you’ve got two different audiences here. 

And then the second bit is that constant engagement. Now, this market is notorious for quoting, binding and not speaking for 12 months and then renewing (unless there’s an MTA), that’s insane. The digital communication capability and using data, segmenting opportunities especially for marketing purposes, and creating a conversation becomes fundamental. It ultimately comes down to not making it a price discussion. We as a market have got to change that commoditised approach because it’s not good for any of us, we’re not going to make money out of this.

James: I would agree. I think service is going to become table stakes pretty quickly. The availability of cloud-based services, third party products, and the amount of money that’s going to InsurTechs is massive. If you go back ten years, you had about two choices of decent core systems. We now have a lot of choices around core systems, they’re quite economical and they’re really quite good.

We’ve got to start thinking about those service elements of the products or provide services in times of need beyond the payout. We’ve thought very carefully about that for cyber. We’re thinking about how we can help our clients with ESG or ESG concerns for next year. So I agree with you completely that’s where we need to get to. We are still focused on the basics though.

Juan: I think the other topic I’ve heard both of you mentioned is the opportunity that opens up once you have access to the data of your full submission flow. I think the insurance industry is notorious for often only capturing data for the risks bound. Some insurers capture the data they’ve quoted and very few capture data on the 80% they don’t quote.

Perhaps you have some thoughts on what could you do if you had access to all that data, the full submission flow.

Nic: I think there are two routes that the question takes us down to, one is purely around data. It amazes me that we’re still tied to the proposal form as the source of the truth to be able to get a quote. Almost all of that data can be augmented in some way, shape or form. And that’s what we do with Cytora. Let’s stop asking for some of this data, let’s go and find it ourselves and make the whole process slicker. Now, I’m not going to sit here and say that the whole of the underwriting process can be automated because it can’t be. 

James: This is it’s still not easy, right? I’ll give you a simple example. Part of the triage process for underwriting a risk is understanding what does the client do. In the SME space, there are certain classes that we certain activities we can’t write. So if you look like you’re a bar, that’s completely fine. But you may also be selling marijuana or doing other activities that we definitely do not want to write. And clearly, from a regulatory perspective and from a risk perspective, we don’t want to take those on. So the art of underwriting often is in the assessment of that. Another good example are casinos in Vegas – what are they? Are they restaurants, are they hotels, gambling facilities or entertainment facilities? Well, the answer is – all of the above, and that could be quite a difficult thing to assess and price.

I think there are technologies coming along that can start to help. 

Juan: Traditionally many insurers focus on mapping a client to a SIC code or a single activity. You might need that as an input to the rating system. But in reality, clients are involved in plenty of activities. It’s more about understanding all the activities because then you can decide whether it’s in appetite or not. 

It’s been fascinating chatting with you both, thank you!