How using data is improving insurance services

by Juan de Castro, COO, Cytora

How using data is improving insurance servicesThis is a shortened version of a bonus episode Making Risk Flow podcast.

You can listen to the full episode here

This is a bonus episode during our between-the-seasons break. This time Juan was a guest at another great podcast – Covered in 15. It was an interesting conversation about how data can improve insurance services.


Adrian:  Welcome to this episode of Covered in 15. My name’s Adrian Saunders. I’m a Commercial Director at Ecclesiastical. For this episode, I’m joined by Juan de Castro from Cytora, welcome.

Juan: Thank you so much, Adrian. Pleased to be here.

Adrian: Shall we start with you telling us a little bit about Cytora and your role?

Juan: Yes, sure. I’m the Chief Operating Officer at Cytora. We are a technology company and we offer a platform that enables commercial insurers to create digital workflows by digitising, evaluating and routing risks within their operations. Before Cytora I spent five years at Hiscox where I was their Chief Operating Officer in the UK. So I’ve been on both sides, on the insurer side and now as a technology provider to insurers.

Adrian: That’s great, it gives you that perspective from both an insurer, but then also from a technology side looking to try and solve some of the issues that we’ve got within the industry. There’s an awful lot of data that exchanges and flows between brokers and insurers. There’s an ongoing discussion in the market about service delivery, part of insurers’ ability to deliver services is the quality of data and information exchange that comes from the broker.

What role do you think data is currently playing and how can that be developed?

Juan: I think data and the lack of data in the communication between brokers and insurers is what actually creates a lot of friction in the industry. At Cytora, we work with both brokers and insurers, so we hear both sides of the industry. Quite often when you talk to a broker, they are quite frustrated with the service received from insurers.

When you think about the top three complaints that brokers have, the first one is: Where is my quote? We hear it over and over, brokers saying that they’ve sent submissions to a number of insurers and days later, they still don’t know who’s going to quote, or even if it’s going to be quoted. They have this feeling of a black hole. And I’ll touch on how this relates to data in just a second.

Another frustration is: why are the insurers so picky? Brokers understand that you want to do a careful risk selection to have a profitable business but most brokers think they have a profitable portfolio. Then I think it’s this challenge around not looking at risks as individual risk, but more looking at the broker portfolio as a whole. And if it is profitable, you should be able to write a broader set of risks. 

The third concern or frustration that we hear is around “why does it take you so long when I just ask you to change a detail in a given policy for the MTA?” And again, if you think about those three areas of frustration, that black hole of quotes, treating risks as individual risks rather as a portfolio, and the time to do an MTA – it’s all about the inefficiency in how data flows from brokers to insurers.

If as an insurer, you receive the better data from a broker, quite possibly you could get back to the broker right away saying, yes, this is in our sweet spot. We really want to quote this, we’ll get back to you right away. And actually, that is something that we are doing with you at Ecclesiastical and other insurers – getting those risk submissions automatically analysed so that you can get an answer to the broker as quickly as possible.

Adrian: I think I can’t imagine Ecclesiastical or any other insurer out there in the market who doesn’t want to turn around quotes as quickly as they can. We want to focus on where we think we can add value by quoting to the broker, and that means being very clear about those that we can quote.

I recognise that black hole issue, and I hear that a lot and I think sometimes what we also forget is that at the end of that, there is a customer. There’s a client who’s waiting to get a quote or there’s a broker who’s got a really great prospect who needs the quote to be able to close the sale. It isn’t just about the dialogue between broker and insurer. I think we have to look beyond that and think about who is the broker trying to talk to, service and what are their goals. I think this friction that you talk about is part of that. 

The portfolio view I know gets talked about a lot and without data that is quite difficult to achieve. So in the work, you’ve been doing, what do you hear from the other side, from an insurer’s perspective?

Juan: We hear a similar set of frustrations from insurers which actually is the other side of the broker’s issue – time to quote. The reason it takes a long time for an underwriter to analyse a risk is because the way submissions are received is typically in quite an obscure way. Some data is unstructured, sometimes it’s not even clear who the client is, that the information is buried in an email. That requires a lot of effort on the insurer’s side to even analyse the risk and say yes, this is something we want to quote or not. 

Some of our clients are expanding the service they are providing to the broker by firstly giving them a quick response on whether they are going to quote the business or not. But secondly, if they are not able to quote it, letting the broker know why, and what would it take for that risk to be quoted. For example, let’s take cyber as a line of business. You might analyse the risk and decide for whatever reason you don’t want to quote it, but the end client wants to know why and what could they do to get the quote or a better quote.

Another frustration from insurers is around being able to get a more complete description of the risk. And I think this is where you get to the dynamics of the industry, where brokers cannot accommodate all the information that insurers might want. Every insurer will want a slightly different set of data points about the risk. Right now, if you look at what underwriters do – they would receive the submission from the broker and then they will do a number of online searches to gather all of the information required for underwriting. That takes time, and that’s a reason why there’s that black hole of getting back to the broker. I think the question here is how to resolve that black hole – you can streamline that process in a way that insurers can automatically pull that information about the risk and be able to either decline or produce a quote as quickly as possible.

Adrian: I recognise this isn’t all just about the poor provision of information and we shouldn’t position it that way. I think one of the challenges from a broker’s perspective is that all of their insurer partners are going to want and need different types of data. We’ve got our own particular style, our own particular way of going about underwriting. We’ve got our own particular risk appetite. 

I think what you’re saying is if you can build all that in a way that actually all of that data flows through and you can source that data, that’s the sort of thing that can start to remove some of this friction.

Juan: Exactly. And I think now technology is available to do that automatically. So you are totally right. You would never expect the broker to provide all the potential or possible data about the risk. I think it’s on the insurers to identify what data would they require above and beyond what the broker provides. How can you automatically pull that data from third-party sources and most of it is available out there. And how can you embed all that data in your process in a way that is as automated as possible? 

I think that the problem here is that data is already being collected, but in a manual way by underwriters. Now, how do you automatically pull all that to make a decision on a risk quickly. That will allow insurers to get back to the broker faster and then the end client will get an answer in a timely way.

Adrian: I think the utopia, certainly from our point of view is that the inquiries that come into our underwriters have already been qualified against that data, and that’s been done in a very quick and slick way, perhaps not in the very heavy-handed manual way that it tends to happen now. So if we think about that data, what’s your experience in terms of availability of data and typically how many different sources do you think you would use to qualify inquiries?

Juan: The availability of data is there already. Your underwriters are doing this manually today, they’re pulling that information from third-party sources, eg. looking up a given charity, understanding how many volunteers they have or where their operations are or understanding if a company had any CCJs.  It’s just being first of all, manually accessed and quite often not recorded anywhere. How does Cytora enable these digital workflows? What we do is by extracting key information from the broker submission, which is enough to understand who’s the client, what’s the client’s address, what’s the product they are asking for, and then connect to all of those data sources automatically to be able to pull that information automatically before the submission gets to an underwriter.

Typically it’s between 10 – 15 data sources, most of them actually are quite common across insurers. So for example, we have an available data library with most of those data sources already pre-connected. It’s reasonably quick for somebody like Cytora to pull all that information together, otherwise, you would need to almost reinvent the wheel yourselves.

Adrian: That’s really interesting that you think about 10 to 15 different sources of data that might be used to qualify a quote. Think how long that would take an underwriter. If we can make that process more automated through the work that you’re doing, that has got talk to this issue around service and it’s got to mean the there’s better connectivity between brokers and insurers. The great thing about the conversation today is we’re not talking about something that’s just a concept out there in the future. You’re doing this with a number of clients including ourselves right now. I think this is genuinely really exciting from our perspective. We’ve been working for a while with you guys and it’s proven its value, but I’m excited to see where else we can go. In general, in the market and for this topic of improvement of the broker’s experience, finding a better way to harness that will be a massive win for us.

Juan: I agree with that. And actually, for us obviously, there are efficiency gains and risk selection gains in what we do, but probably the most exciting reaction to our technology is when brokers say, wow, we’ve seen a change in how an insurer that’s business with us. They now are hungry, are more responsive and provide better service. That is something at the core of your vision, Adrian, and hopefully, we are helping you deliver that.

Adrian: Yeah, absolutely. I think that’s a great message to leave our audience with, really. So thank you very much for taking the time to talk to us. 

Juan: Thank you so much, Adrian. It would be a pleasure to join you again. Cheers.