15 mins read
11
.
04
.
2023

Risk Selection Consistency: Empowering Underwriters with the Help of Data

by Juan de Castro, COO, Cytora

This a shortened version of Making Risk Flow podcast, episode: “Risk Selection Consistency: Empowering Underwriters with the Help of Data”. In this episode, Juan is joined by Mandy Hunt, the Chief Underwriting Officer of Commercial Lines at RSA, to discuss the crucial role of data in the underwriting process. During their conversation, Juan and Mandy consider how automated and structured data can aid risk selection consistency, help underwriters understand the data at their fingertips, and how to identify new sources of data. and an increasingly connected world.

Listen to the full episode here

Juan de Castro: Hello. My name is Juan de Castro, and you’re listening to Making Risk Flow. Every episode, I sit down with my industry-leading guests to demystify digital risk flows, share practical knowledge, and help you use them to unlock scalability in commercial insurance. Welcome to another episode of Making Risk Flow. Today we’re going to discuss data and more specifically data in underwriting. And I cannot think of anybody better than Mandy Hunt to join me today. Mandy is the Chief Underwriting Officer of Commercial Lines at RSA and is very passionate about this topic and many others that I’m sure we’ll touch on. So, first of all, Mandy, thank you so much for joining me today.

Mandy Hunt: It’s a real pleasure to be here. Thanks for inviting me in the first place, to be fair.

Juan de Castro: So perhaps you can start with a brief overview of your current role and your background and the journey that took you there.

Mandy Hunt: I’ve been with RSA actually 25 years this year and during that time I have predominantly had underwriting-led roles. So I started as a junior underwriter in 1998 and I’m pretty proud of the fact that I’m now the CUO of the business I started in, to be honest. But in the interim, I have taken steps to make sure my skills are balanced in different places. So leading a trading site was the first big one of those in 2012. And then I went and became the MD of our Channel Island business called Insurance Corporation of the Channel Islands. So that’s a fully operational mini insurance company operating in Guernsey and was a brilliant area to learn the whole business end to end. They operate like they are their own entity, even though they’re under our umbrella. And from regulation to HR to how people walk in and pay their insurance premiums in person, we did all of that in the Channel Island. So a brilliant way of learning about our business. So, yeah, my job today, Chief Underwriting Officer of Commercial. So that covers our regional or mid-market business, the SME business of our digital online opportunity, and our delegated business, i.e. where we offer the capacity to third parties, be that through an MGA or a scheme. And my role is to look after the underwriting performance and the portfolio management elements of those areas, as well as look after risk consulting. So the people that visit customers, the eyes, the ears for us as underwriters and help them shape their strategy. Now, the difference for me with the risk consulting business is it’s a team that also serves our specialty lines business in the UK and our European business. So we get a real breadth of interest and things that we’re trying to do in that space and they have their own unique needs with data as well. So I feel very fortunate to have the job I have. I feel really lucky to be able to lead a team of people that are trying to make life easier for our underwriters and make RSA successful.

Juan de Castro: Definitely. And I think that’s one of the reasons I really wanted to talk to you. I think you’ll bring that perspective from every angle of insurance. As I mentioned in the introduction, you are quite passionate about data and specifically, data and underwriting. So where do you see the main benefits?

Mandy Hunt: So there are a couple of areas, really, I think so at the heart of it, it will make underwriters’ life easier. Because we’d have consistency in information and therefore they will be able to make better decisions more in line with the underwriting strategy. I think when you don’t have data, you’re a little bit in the hands of the underwriter, picking and choosing whether things fit. But where you have a data strategy that sits around it and rules and structure, then you can be more confident in the risk selection. But I think it will give us insight which will help us make better decisions. But what I really don’t ever want anyone to think is I don’t want it to replace an underwriter. I want underwriters to have tools that help them be the best underwriters in the market. And I want them to be able to do that because the data is doing the hard yards for them and they can spend the time really reflecting on risk.

Juan de Castro: Let’s deep dive into each of those two areas and perhaps bring both to life. You mentioned making life easier. And the goal of data is about driving insights, not replacing the underwriter. Perhaps can you share with us some examples of how you’ve seen data used to make an underwriter day to day easier?

Mandy Hunt: Yeah, so I think there are a couple of areas, really. If you look at things like motor, we obviously can have access to lots of information about vehicles and how they work and what you can extract out of them. Underwriters are looking for ways that help them understand the risk quality that they are trying to seek that meets any insurer’s risk appetite, and every insurer will have a different view of it. And the hardest part of doing that is trawling through things. Let me search the internet, right? Let me try and find all of these things. Whereas when we talk about data, what we really are looking for is a consolidated view of some consistent things that help us assess those aspects of risk that we think are important. We might historically have asked for something like, do you have a health and safety policy? That might be a question you would have historically asked. As an underwriter, it would be an important thing to do. It’s also a legal requirement when you’ve got a certain number of employees, but having one doesn’t mean to say that somebody’s using it and applying it. Whereas if you looked at data from things like accident reporting, you would see whether the health and safety policy is actually helping a risk get better or worse. And so you can see those things through the data that you can get from information. Now that might be information you have. Simply looking at a claims experience, are the claims volume is going up or coming down and then you get a sense of whether the data is actually influencing the underwriter’s decision to underwrite that risk. We see the same in motor. Does the same driver’s name appear every time there’s an accident? Is it the same driver? Does the same vehicle always need a new windscreen at really basic level? That’s the simple way we can use data. What we obviously want to do is add more and more complexity into that, which is maybe more external data that’s available, that helps us understand, or it might be building on the data we have and creating new insight from some things. So I was talking to somebody this morning, and this might sound almost basic, but just which trader, which office, which brokers are really helping us win business? And why is data that can actually really change the way we trade? So from an underwriting point of view, there’s loads of things we can go at. And I think as an industry at the tip of the iceberg, right? We are starting to think about it. And actually a conversation I was having, in fact today, not that long ago, which is what data do we have? All of the things we consider are important from a risk point of view. But what data would we love to have that we don’t even know if it exists? And our job is really to find those data sources that we already have, look at the ones that are external and then work out. Do they actually make an underwriter’s life easier and make those decisions more consistent? Because that’s ultimately what I want. I don’t want odd risks being put on. When I have a strategy, I want to know I’m going to get what I want to have.

Juan de Castro: Definitely. So based on what you’ve said, sounds like there are almost three themes I’ve captured. I think that the first one is about how can you provide your underwriters with readily available data so that they don’t need to go to 100 different places to capture the data. The second one is you’re not going to just throw all the data to the underwriter. It’s how you summarise that data? How do you create insights from that data? So the data needs to go through all the data pull on a given risk, but you’re highlighting giving risk characteristics to the underwriter. And I guess the third one is how you identify new sources of data.

Mandy Hunt: Absolutely. We talked the other day, it made me think about, nothing to do with underwriting example, right? So when I was first pregnant, I went into Mothercare to buy a pram. And I was in there for half an hour and I walked out. I could not understand all of the data that was being thrown at me about why that pushchair was better than that one and what this one did. It was too much. So I was like, let me just take all the booklets, I’ll go and read. That’s what I needed. My husband, on the other hand, wanted to buy a new telly, right? And he had a spreadsheet of all the features and he had all of this information. So he had decided by the time he got to the end that it was this telly. Because of these three things, we both came at purchasing expensive points at the time in completely different ways. And we have to recognise that’s what underwriters will do. They’ll take the information and they’ll receive it in different ways and want to use it. So our job is to make sure that it’s readily available, that it’s consistent on all of the things you’ve described, but they also understand it and they’re not overwhelmed like I was buying a pushchair or they can use it in the way Neil did, although he didn’t use it quick enough. Right. By the time he went to buy the telly, they’d stopped making the thing he wanted so he had to start again. Right. In a funny way, it’s a good example of overwhelm and speed and they’re the things we want to bring out of these pieces of data. And I also think, and this is interesting, talking to a company like yours, data is just a word. It is just a word. And I think when we talk about it and people, I want more data, tell me what you want. And then that becomes harder to answer. I think the specifics and that’s when I would really encourage people to think about the exact question they’re trying to answer with the data, not just say I want data because I could give you data, doesn’t mean it’s going to help you.

Juan de Castro: Perhaps just picking an example. Right? Just an obvious one is the claims history of a given risk. But that can be 15 pages of detailed claims activity and that is probably not what an underwriter wants. What they really want to know is what’s the frequency of claims, or whether has there been any claim with a certain severity. It is perhaps changing the question.

Mandy Hunt: I would agree. So as a liability underwriter by trade, I’m not worried necessarily by the size of the claim, but the volume of claims in one sense. So size is an indicator. But we had a chat on a risk the other day and somebody said, well, I think I should divert this. The claim is bigger than the premium. I said the claim was always going to be bigger than the premium, it’s tiny. What was the claim? Tell me about the specifics of the claim and are you comfortable that that is unlikely to happen again or if it does happen, it’s as likely as it was the first time? Or have they had ten losses that are repeatable that actually we could have avoided and is there a way of avoiding those which you could do with a deductible or something like that? So that’s why it’s important to me that we educate people about the quality of the data and what they need to look for, because numbers are just numbers, you’ve got to understand the background to them and the questions you want to ask. So, yeah, why have the claims happened? Are they repeatable, have they repeated a lot, what they’ve done about it? Those things are important.

Juan de Castro: So I guess the two enablers you were discussing there, one is how do you make sure the right insights are provided to the underwriter almost before they start looking at the given risk?

Mandy Hunt: That really is important to me, actually.

Juan de Castro: Where is the industry in that regard? At the moment?

Mandy Hunt: I think it’s probably a bit inconsistent. We’re in a marketplace, certainly where I work in commercial, where brokers are equally trying to have a role and win business. So they will use their own insight and their own ways and their own information they believe will make a difference to the insurance purchase. So we don’t have a consistent approach to how information is received, so our job is to make it consistent by the time it lands on our laps, which is to be able to extract the data out of those quotes and figure it out. But I think the industry is learning and you see brilliant examples of it when you look at things like SME and our SME business and its digital journey, and then you might go to the other end and people will still walk into an insurer’s office with a slip in paper format occasionally. So we have got a whole plethora of versions of how we use data and how we’re going to move, but we’ve got a long way to go. We’re definitely not like a really high tech company, although there’ll be versions of that. Of course, ensure tech is doing some great stuff as well.

Juan de Castro: But at least it sounds like the direction is the right one.

Mandy Hunt: Without a doubt. Yeah, I think we’d be naive to ignore it, but I also think we can see the value of it. And I think that’s probably the difference between now and maybe ten years ago. When we know how to underwrite, we’ve got a book, it tells us how to do it and we’ve done our processing, whereas now we are starting to see what is coming and what is available and the possibility. And that’s the joy, isn’t it? What is the possibility of the data that we could do something with that makes our business better and whether that’s underwriters’ life easier, our financial results better, risk selection, whatever those things are, there are endless possibilities with what we can do with data.

Juan de Castro: Something that you’ve mentioned a couple of times I would like to potentially dig a bit deeper into is consistency in risk selection.

Mandy Hunt: Yes.

Juan de Castro: What do you mean by that?

Mandy Hunt: So I guess in the worst case scenario, you’ve got somebody that says you can write offices. I don’t know, this is a really basic one and I mean concrete or brick built buildings that are very purpose built versus, I don’t know, writing an office above a chip shop. And of course they present very different risks because not necessarily because the office work is any more hazardous, but the processes that are going on around it. And that is going to be true across every trade that you can think of. There will be some risks that are acceptable and there’ll be some that are not for any insurer. And what we want to avoid are the cases that could hurt us, that we’ve taken on by mistake, or we didn’t understand, or there was a bit of extra information that if we’d known we wouldn’t have done, or we’d have charged a bit more for, or whatever it would be for us. I would like our underwriters to not have so much responsibility to make sure they’ve covered every point. Are there some data points that go, if these things are true, then this is a risk we can quote, and it not be so reliant on trawling through Google, trying to find pictures, sending surveyors out. Because ultimately when you define as an insurer your risk appetite, you want the business to sit within it. And my guess is occasionally some of the losses that impact insurers are those that are not quite within the risk appetite or they’ve moved outside of risk appetite because something that’s changed in the risk and we need ways of being able to spot those things too. And there’s lots of opportunity. I think there are lots of companies doing lots of interesting things with company SouthData and things like that. And I think sometimes, certainly with the smaller end of our business, sometimes they don’t realise the impact. If RSA suddenly tomorrow decided to sell insurance but also make cars, we would understand that would be a different risk profile. However, a small SME who currently makes, I don’t know, curtains, who then realises they’re quite good at making furniture suddenly does the two together changes the risk dynamic from something we’re comfortable with maybe to something we’re not because of the process and the fire hazard, but probably isn’t thinking an insurer is going to be worried by those things. Just thinks that’s just how businesses are. So where we can have tools that enable us to see and spot those things, we bring far more consistency to the decisions we make. And then that means that on a personal level, an underwriter is not struggling with some of those things where maybe there’s a loss and everyone’s saying, well, why did we write that? How come we ended up with that habits, the rules would help us avoid those situations.

Juan de Castro: So if I understand correctly, it is a combination of rules to help underwriters identify what sits outside of the underwriting appetite. So you should not be quoting it. It’s either a small component or even those that are within the appetite. I’m sure there will be some segments where you want to prioritise higher because it’s a segment where you want to grow or you’re rebalancing the book and it’s a target segment. Is that risk selection consistency also have a component of guiding your underwriters to the type of risk you really want to have in your book,

Mandy Hunt: Where you should spend your time. Ultimately, an underwriter is sitting there trying to choose, do I do broker A’s or broker B’s or C’s risk? And if we’ve got tools that help them go, B is the one to do, C will give it to you, but it’ll take a long time. A, you never win from and actually that trade is a little bit on the edge of appetite, then you will make the right choices and as long as we manage all of those brokers in the right way from a relationship point of view, I’ve not come across a broker who objects to a no. They object to a no months after they sent you the inquiry, but if you tell them early on, they’ve got plenty of time to go and look. So if we can get to those decisions quicker, then we make the broker’s life easier and then they can trade better with their customers, so everybody becomes a winner.

Juan de Castro: And also you’ve mentioned there, I think there are two dimensions to that risk prioritisation. I think one is how does a risk align with your underwriting strategy and the other one is how does it align with your distribution strategy? Because it’s like a two by two, right? It might be highly attractive from an underwriting perspective, but from a low converting broker or from a high converting so depending almost where it sits in the matrix, you might take a different approach.

Mandy Hunt: Right, completely. And I think in a world where brokers have, there’s a lot of consolidation. So you can deal with a national broker in the South, the North, the Midlands, the West, overseas, they also want to know that they’re going to get a consistent answer. They don’t want to come to London and get a no and go to Manchester and get a yes because two underwriters are taking a slightly different view. They want to know that if we say no, we say no and we mean it. And if we say yes, we absolutely mean it. So that’s where consistency really helps us because you avoid some of those things. And probably early in my career, I can think of one or two of those examples where that became you were managing off, why did you decline it and why did you quote it and things like that. Now, obviously we have data and tools today that make that less likely to happen, but still, it’s not impossible and that’s what I would like to avoid so that we’re really clear and brokers when they think about RSA, but I guess any insurer know that they’re getting a really consistent answer based on the best information possible.

Juan de Castro: Almost like this consistency in risk selection also goes back to the point we were discussing earlier about summarising risk information for an underwriter and making their life easier because I would assume you also want your underwriters when they are looking at the risk, rather than having to almost like analyse the full risk. If technology or data could highlight specific characteristics that require specific attention, that also goes a long way in terms of underwriting consistency, right?

Mandy Hunt: Yeah, completely. So I was listening this morning to people talking about some of the opportunity for that insight out of the marine world today. And I’m definitely not a marine expert, so I’m not going to sit here and pretend I am. But when you listen to the opportunity that exists from the level of sensors on boats and things, imports and on fenders and all of those things, you start to realise that there is so much opportunity. But again, right, it comes back to me. You can give loads of things to people if they don’t know how to use it, they don’t know what they’re looking for. Then we just have to think differently in my team, which is how do you train people to use the data in the insight in a way that talks to them? And we have a massively different age range of people now in our industry and we will have people that are my age and generation who will have been used to a calculator, a pad and a pen and flicking through. And then there’ll be people when I think about my daughter and my sons, they learn by watching YouTube videos. So we got different spectrums. I jokingly tell people I learned to do cooking off of TikTok, right? And there are people that will learn in different ways. So I think when we get data, we’ve got to make ways of being clear about why it works, what you’re looking for, and then where you move forward from.

Juan de Castro: So this helping underwriters understand the data, I guess you could also take two different approach, right? The more traditional one, which is you deliver training courses to your underwriters and keep them up to date, et cetera. But is there almost like a more modern way of almost can technology be that simple? That almost you don’t need to train your underwriters. That should be what we should aspire to, right?

Mandy Hunt: The decisions they’ve got are much smaller. At the minute the choice starts from the first presentation that comes in and from a broker to the insured title, right the way through the journey, the more we can make easy for people in chunks of that the time they spend on the technical decisions and really the pricing decisions, right? Does the premium that you are asking for cover the risk you’re being presented with and do the terms and conditions of that policy? So basically the policy wording do the things you intended it to do and sometimes in our current world we spend loads of time trying to get the information into systems and tools and we have less time to do those technical bits of the decision. And ultimately underwriters are technicians, right? I listen to people talking and in fact I was just in with our risk control team and we were talking about sprinklers and one of them said life saving sprinklers or property protection sprinklers. Very different way of thinking than giving me the types of sprinklers by some code that exists, if you get my meaning. Now, when I think about my children, they’ll probably get life saving and property protecting sprinklers. They won’t understand the code. So when you ask how do we do it? We make the rules as easy as we can at the front end and then the stuff that comes that’s a bit more technical. Let’s talk in a different language. We don’t have to talk in the 1920s, insurance speak to make it real for people today. And I think the other part is we have to recognise people learn in different ways because we’ve got a generation that have go to school and learn differently to the way I was taught at school. I need to make sure that my team are building technical capability and training and insight in a way that talks to the whole community of underwriters we’ll have and there’ll be loads more data savvy than me at their age. My son is doing an economics degree and he is just doing data, data coding, coding, coding. That is one whole term of work. That is all he is talking about. We didn’t talk about that when I was there. Right? So we are a different generation who will have different insights. So when they arrive, if I tell them they’ve got to fill in three different forms to get a quote they’ll be looking at me thinking what? That’s not right. We should have that all automated and then they’ll be looking for the data that helps them make the decision. So we have a great opportunity with the generations that are coming behind us because they’ll think so differently to help.

Juan de Castro: Make it better and often when we talk about making the underwriters life easier. Often people just think about, okay, how do we drive efficiency? How do we how can we get them to quote more? I think it’s much more valuable than that, right? It is how do you create an environment where underwriters just love working at. And especially nowadays where it’s like it’s so hard to find underwriting talent to retain them. Unless you create that environment you’re not going to have underwriters in ten years time. Are they going to go to your competitor that offers them a much better underwriting experience?

Mandy Hunt: I would agree. So it is a skill. Okay. So in some parts of our business, it is absolutely about rules. Let’s be honest. Personal lines is about putting rules on systems, trading it on a volume basis, optimizing your channels, working out when you’re pushed enough rate, and then you’re suddenly dropping off your retention or your renewals or whatever. When you get to commercial, we will still need the expertise that can work. Is that really far away from a river? So we were presented with some data once that said basically the whole of the Channel Islands was a flood risk. Well, it wasn’t probably difficult to work out. That would be what a map would tell you. Because they’re small islands surrounded by lots of water. If you live in the island, you know which ones are truly at risk of a flooding and which ones aren’t. And you have to use that personal insight to make the decisions and make it much more relevant. And I think that’s what we’ll see more of people using data and then bringing that personal insight and giving them the space to really think about those things and understand what processes mean. A former colleague used to talk about the probability of a loss and the possibility of a loss. Every claim is possible.

Juan de Castro: Is it probable?

Mandy Hunt: Is it probable? I think we sometimes have underwriters that go, well, it’ll have a loss tomorrow because we’ve just written it. It probably won’t, but we get a little bit anxious. So when you can make some of those things easier for underwriters to make those decisions on, then it’s about the quality of the risk and the quality of the risk management, the quality of the building there. Does it fit your appetite? Those things are much easier to manage, definitely.

Juan de Castro: So perhaps bringing all these together. What would be your vision of the end to end process? Let’s take the new business process where a broker is sending you a submission to quoting your risk. Where do you see the role of technology and data and then what’s the role of the underwriter?

Mandy Hunt: Yeah, that’s exactly a question I’ve asked this year, which is, if you looked at our whole process, where does automation play its part? Where does an underwriter add value and how do you make the process. Easy for somebody to manage. So ingestion is critical, right? So we want underwriters to receive something in their lap that is easy for them to pick up and run with. That requires us to have ingestion into the systems. Then there’s simple things like avoiding dual keying into a number of different systems. How do you automate some of that stuff? The front end of that process you described, probably up to the point that you start to think do I want to write this risk or not? Because you’ve gone through all your triage and you’re at a point you’ve decided and the system has decided yes, then that’s when an underwriter comes in. So the first 25% of the process, all automated, data driven, insight driven in a lap underwriter now looking, I don’t know, 35% of the time in that negotiation with a broker having the conversations, what are you looking for? What does the customer want? What’s driving the marketing? How can I help?

Juan de Castro: In that transition from technology to underwriter? What would be, what the underwriter would see to start that negotiation?

Mandy Hunt: Because there are so many different types of insurance, we probably want some basic company information. So I don’t know, director’s information. Can you look at the CCJs? Can that just all land? Then there’s some specific risk information, like the type of building, is it near a river, or is it in a high arson area? Data we actually have today. But can you see all of that key underwriting information, things that are not in a high hazard area that you’re happy to write? Some of it will be the broker’s information that they’ve listed in their conversations that they want to share around the journey. Sometimes it’s in conversation though. So the broker sends us their standard presentation and they send us, you reference claims experience earlier so we can put all of that in the system and the system could say no. The conversation with the broker, which is where the underwriter really comes in, is to say they’ve had quite a lot of claims. What happened? Tell me a bit about that. Oh well, we sold that business and actually all the liabilities went with to the new company or that was one driver and they’re no longer with us. Or we instigated a risk management program. Then you start to take a different view of those claims. That’s why I say the underwriting part is where you’ve got to. But the underwriter will be understanding more information that will fine tune the automated part of the journey to help us get to the best price. And we haven’t even talked about data in rating engines. But rating engines are every insurer will have them. They’ll be complex, they’ll be simple, they’ll be whatever they are. But they’re the things an underwriter has decided they need to understand some of that. And I think unfortunately in commercial, a lot of that information is not automatically available in some lines business. I think in places like marine, motor you can get some, but it’s much harder to get construction information on every commercial premises in the UK as you get a fair idea when it comes to personal lines information.

Juan de Castro: That’s a really fantastic overview, linking it back to your point about the consistency of risk selection – some of the risk characteristics that often drive appetite decisions or even rating considerations are things like client activity, like classification of client trade or activity.

Mandy Hunt: Absolutely.

Juan de Castro: That is an area that historically has been very subjective. As we know, some brokers spear the activity because they know that they will get a better rate. Some underwriters will do the same. Right. Have you thought about how to drive consistency in that type of peer judgement decisions?

Mandy Hunt: Yeah, so SIC codes are obviously a really simple way of doing that. Companies House data that gives you that information. The industry is full of trade codes and sort of trade specific codes that people will allocate rates and logic to how much capacity on a heavy food risk with loads of heat compared to a dry food processing risk where you might put more capacity down. But when you talk about that first stage, I am literally looking thinking companies house data would be really easy. We don’t probably all have it, we have it in pockets, we have it for some lines, we don’t use it in every line. It’s probably where I would go, it’s available. But the challenge is I think there are so many parts of data I probably don’t even know what is available fully. Sometimes it’s someone shows you and I’ve seen some of the opportunities to read a broker presentation, look at zip codes and go well, did you know on company towers it also says they are a motor repairer and over here it says they’re a double glazing manufacturer. We can pull that out. I think the world is full of so much information we also need people to tell us some of that. But for me we have to start with what do we want? Knowing as an insurer what information you want, can you map data points to it? And then if you could really just expand your brain and go what would make this even better? And then go and ask is there some data out there? Come and talk to companies like yours is can you get me something that will help me answer this question?

Juan de Castro: Fantastic. So one last question in your role at our sale, where do you start? There’s a huge opportunity in front of us, data can really help. What are the first steps you are taking in that journey?

Mandy Hunt: So the first one from my point of view was to map actually the process. Because I think at the beginning what we wanted to understand is where would we get the greatest benefit from which parts of the process we could improve. And that’s a data driven activity on its own. How long do people take in each of the stages? Is something automated? Is something that where is the dual keying aspect? How many systems do people have to talk to? Because then you can go, Well, I’m going to focus in this area, but at the same time as I’m focusing, improving this bit of the process, where can I automate? But underneath all of that, and if you haven’t got a company strategy that’s really clear about where you want to grow, which areas are your target, areas where you want to hold or even exit, you can’t make a strategy from a data and a risk and an underwriting perspective because you’re shooting in the wind. You don’t know what you’re going at. So, for us, we have had a really clear group strategy that’s come down through group, through the UK, through to each of the PNLs in commercial, and as a result, you can go in digital. We’re spending a lot of time improving flow, working on data that makes things go through the funnel quicker and with less touch points in regions. It’s about automation and how do we drive time for underwriters to make good decisions in that big, chunky bit of actual underwriting. And then in our delegated business, it’s an interesting one, right? Because how do you work with partners about what tools they’ve got and what ideas can you give them? But let’s be honest, they’ll probably give us some ideas too. They’ll be doing things that will be going, that’s interesting, I wish we had some of that. So it’s a good way of learning what else is going on. So, for me, get a strategy for your company, understand where you’re targeting. We want to look at the process that sits around it. Work out where your best wins are, because you’ve only got so many people and so many hours, and then talk to the people that can give you the automation and that ingestion, because ultimately you can automate the backend, but you’ve got to get stuff in the funnel first. So get that front end as quick as you can.

Juan de Castro: And educate your underwriters.

Mandy Hunt: Well, yeah, that’s definitely on my agenda. Maybe people think I’m a bit boring, but I think about the number of times people put things on my desk and they just expect an answer. And I talk to me about why, how, what if you tell me I can come to an answer? That’s not the difficult bit, but I actually want to see how well you’re thinking about things. And actually, that’s when the rubber hits the road. Have we taught people, have we trained people? Do they understand what they’re actually asking? So that’s when I think about data, we have to make it easy for people to understand.

Juan de Castro: I think this is a brilliant point to wrap up the episode. Mandy, I’ve really, thoroughly enjoyed the conversation. Again, thank you so much for joining me.

Mandy Hunt: It’s a real pleasure. Thanks for the afternoon. Take care.

Juan de Castro: Making Risk Flow is brought to you by Cytora. If you enjoy this podcast, consider subscribing to Making Risk Flow in Apple podcasts spotify or wherever you get your podcast so you never miss an episode. To find out more about Cytora, visit cytora.com.