Log in

Forgot password?

Don't have an account with us?

Register Now


Podcast 38. Insurtech as a Service (Part One). Hitachi, Munich Re Digital Partners, Kasko, Shepherd and QBE

Another fast moving series of chats with more leaders from insurance, technology and data. In this episode, the first half of our 8th July evening at The SteelYard, Robin investigates how Hitachi (2:00), the 7th largest technology group in the world, is making it easier for established insurers to work with the emerging providers of data for trains, buildings and a whole lot more. Serif Otterburn, Business Development Manager, and Insurance Subject Matter Expert  “Everything Hitachi builds, from trains to medical equipment, requires insurance. A lot of insurers use historical data to calculate risk, but Hitachi is using its expertise in IoT to assist insurers to develop newer better risk models”. Joining Robin and Serif are Chris Gill from QBE (7:50) and Stephen Chadwick from Shepherd (16:00).

Munich Re Digital Partners (22:00) are well known for supporting a whole range of new MGAs and technology companies with capacity and sometimes investment. We get a peek behind the kimono at how they are getting on. In the words of Mark Dennis, European CEO “We’ve met with about a thousand InsurTech start-ups in the last three years. To date, we have about twenty live and wish ultimately to work with about fifty. Our job is to provide the scalability to get them to market as quickly as possible as InsurTech services.”

Kasko (34:20) was on stage at InsTech London 4 years ago at one of our earliest events, CEO Nick Sühr returns to talk about the work they are doing today with over 15 insurers providing plug-and-play ready solutions, enabling insurers to turn insurance products from idea or paper to digital in a few weeks.

Listen here to podcast 38. It is also available on iTunes, Spotify and Podbean. The second half of this event follows in episode 39 with more from Hitachi and their partners.


Transcript for this podcast

00:00 Serif Otterburn: We’ve got our own factories, we’ve got our own fleets, we’ve got our own trains. We build construction machinery, we build hospital equipment, and all of those things need insurance.

00:19 Matthew Grant: Hello, this is Matthew Grant. I’m one of the partners at InsTech London. And in this episode, we are bringing you the first half of our event from the steel yard on the 8th of July, when we’re talking about InsurTech as a Service sponsored by Hitachi. Now, one of the problems that many technology companies have, both new but also established ones, is connecting into the systems used by their insurance clients. And the insurance companies themselves often find it difficult to know how the technology solutions are going to link into their own applications. Hitachi have found a way to solve this problem with what they call InsurTech as a Service.

00:58 Matthew Grant: And essentially what they are doing is they are assessing each of the technology companies that their clients are looking to work with, and then creating a solution for connecting these two together. So we are delighted to have in this episode, Serif Otterburn from Hitachi talking to Chris Gill from QBE, and Stephen Chadwick from Shepherd, and that is followed by Robin Merttens talking to Mark Dennis who is the European CEO of Munich Re Digital Partners. And finally, we have Nikolaus Sühr, CEO and founder of KASKO.

01:41 Robin Merttens: So, on the stage, from left to right, Stephen Chadwick who’s the CEO of Shepherd, Serif Otterburn, whose role is roving at Hitachi Consulting. Stephen Gill, Head of Risk Solutions. And we’re going to talk about InsurTech as a service in particular, what Hitachi are doing, and where that fits with IoT as a service, which I think has fascinating potential for the insurance industry. Serif, I’m going to start with you. So, what is Hitachi Consulting?

02:18 SO: When you think about Hitachi first, a lot of you probably think about engineering and innovation, but we’re actually also the seventh largest technology company in the world, with a $4 billion business in financial services. And Hitachi Consulting is just one of those organizations of the 950 in that conglomerate, and at a really high level basically, we help companies transform their business by providing technology and digital solutions.

02:53 RM: So I know Hitachi for building trains, smart cities, electronics, heavy machinery and stuff. So in this day and age, that stuff has astonishing data sets attached to it. Do you, at consulting get the chance to play with those toys and to do exciting data stuff with them?

03:11 SO: Yeah, yeah, absolutely. We’ve got our own factories, we’ve got our own fleets, we’ve got our own trains. We build construction machinery, we build hospital equipment, and all of those things need insurance. And basically, we as Hitachi Consulting, we’ve digitized all of those things. We’ve optimized all of those things to make things safer, and we then provided all of that knowledge to other companies, other industries, other cities, other transportation industries and so on. So, yeah, it’s really important then to make sure that we know that there’s a risk associated to all of that, and what we at Hitachi are doing is making sure that insurers understand that risk and we provide that information to them.

03:54 RM: So, based on my knowledge of this stuff, there’s a huge, massive data set, dynamic data set that you then have to get down to something which represents something that insurers can understand from a risk prevention and a risk pricing point of view. And I guess that’s the magic, and where are you with that and how is it going?

04:16 SO: Yeah. So, at the moment, a lot of insurers are looking at historical data to calculate risk and a lot of averages, don’t know if you agree with that, but there’s so much data out there, and we are experts in IoT, and we’re also experts in ingesting data at large scale, and all of that basically needs to be fed into the new risk models and the new risk calculation, so that’s what we’re doing at the moment. So I’ll give you an example of… Two examples of what we’re doing and where we’re doing that. So, Hitachi, they provide trains and in every single one of those trains there’s 3,500 sensors. And you probably didn’t realize though that we at Hitachi Rail actually provide those trains as a service to a train line such as Virgin or Great Western Rail.

05:03 SO: And the whole point of doing that… So basically, if the train breaks down, it’s not Virgin or Great Western Rail that will suffer. It’s actually Hitachi Rail. So we have ingested as Hitachi Consulting, we’ve ingested all of that information in real time and provided it back to Hitachi Rail to say, “Hey, this is what you can do and this is how you can do predictive maintenance to stop those trains from breaking down as much as possible.” And another example is, we’ve actually partnered in Los Angeles with a company called Dispatcher, and this is really relevant to commercial insurance at the moment, to predict and prevent wildfires from happening or power outs just from happening before it occurs. So we leverage… Hitachi Consulting leverage machine learning and AI to actually make those predictions as to when the wildfire is going to happen.

05:51 SO: And then we’ll send off Dispatcher who are basically like field service engineers to actually go and assess that area and prevent anything from happening. And over the last year, we prevented about 12 power outages and about a dozen… Half a dozen wildfires, and one of those power outages would have actually left millions in danger. So, connecting that information with insurers is really, really important and making sure that that feeds into their risk calculation means that they can provide different services to their existing customers.

06:22 RM: Yeah, I was going to ask you that. So, if I may get a bit of free consulting from you while you’re here.

06:30 RM: How do you think this changes the nature of the insurers role and the nature of insurance as we currently provide it?

06:36 SO: Yeah, so I think there’s three things, really. I think the first one is reducing claims. So obviously, understanding all that new data in real-time means that you can start assessing risk differently before you actually place insurance, which means that you can make a better decision before placing insurance and hence, hopefully reducing significant claims. I think the second one is improving customer engagement. So, getting away… How many times I’ve had insurers saying, “We want to move away from talking to the customer, just at renewal, just at claim, and we want to start adding more value.” So you can add more value to them if they share their data by saying things like, “Okay, this is how you’re performing on your current business policies and this is what you can do to improve those policies or improve what your employees are doing to mitigate risk.” And then finally, I think, actually, new revenue streams. With all this data, and we’re partnering with your customers to be able to share data as well, you can start understanding new opportun… Identifying new opportunities that are out there and creating some new products that you can then go to market with.

07:42 RM: Trying to hear what insurers are really doing with this stuff. Firstly, let me congratulate you for being very brave because I’ve been in and around the IoT space for quite a long time and I know that insurers are struggling to come to terms with what it means. So, on a train front, this all sounds like a data set that doesn’t naturally lend itself to a grabbing spreadsheet and a rating engine. So how are you coping?

08:06 Chris Gill: Well, it’s a very interesting question. For insurers, we’re at a point now where data is becoming more and more prevalent. We’ve hit an exponential curve in terms of technology and data that’s now available to us. And customers first and foremost, are speaking to us and they are asking us how we can use their data and work with partners so that we can tailor and bespoke the risk pricing model to the actual exposure they present today. Insurance is made up of three key raw materials that we have to consider. We have people first and foremost, so the talents and experience that we have within the insurance company, we have the capital that sits behind us, and then we have the data. And it’s the data we really need focus on, it’s going to move us to the next level. And as I say, we need to become more relevant to our customers and our customers are, as we’re saying now advancing more than we are as an insurance industry whole, in terms of looking at supply chain, crisis management, looking at wearable technology, and how that influences the human behavior and health and safety, looking at telematics, so to keep up with the industries that we actually serve and support, we need to evolve as well quite rapidly.

09:26 RM: So it sounds like early steps on a long journey, how far have you got and what are the kind of… From an internal adoption point of view, how does that work?

09:40 CG: That’s quite a tough question. So QBE at the moment is going through a process of transformation. So we’re looking at this from three key areas as an insurance business. And probably not too dissimilar from our competitors as well. QBE sits up there in the top 20 global insurers and digital transformation is actually quite frightening on the inside looking out because there are startups appearing everywhere, there’s data available that we’re not familiar with. So I think it takes a very brave company to sort of step in, embrace and actually start to organically learn how to better use data as a company. We’re looking at this one of three different ways.

10:23 CG: The first way is we’re looking at the data which is available. Shelly, you mentioned at Hitachi that data is quite difficult to come by with insurers, to make sense of it. And when we’re looking at this from an internal perspective and we see this across the industry through our broker partners, 60% of… If you like, claims causation is actually unknown. It’s almost, I suppose, behavior within the industry, almost lazy reporting whereby the claim states, “we just don’t know where it’s coming from,” and so that data becomes almost useless. So we need to become more disciplined, more sophisticated and actually putting down claims analysis against that so we can turn it into information.

11:07 CG: The second one is actually what do we… How do we turn data into information. And so as a business again, we’re looking at, from an underwriting perspective, turning this data into information through working with third-party companies, we’re working with Cytora, we’re now looking to partner with Hitachi and Shepherds and a few others out there in the market for our ventures, to be able to scrape the external public domain for datasets which will be able to give us insights into businesses. And now, where we just started to look at claims and risks from industry to how many people they employ, looking at key losses over the last five years, we’re now becoming a much more forward-looking within the data sets to be able to give us predictive analysis and be able to better tailor the risk management programs.

11:55 CG: And then thirdly, it’s very much around sort of, if you like, the digital worker, which is keystrokes. So through the claims process, we’ve already started to digitalize our entire claims process, which looks at emails informing brokers, it looks at how we instruct solicitors and close claims and key being, sort of, in a period of a month, we get over 15,000 transactions for claims coming in. And through this process already, we’ve saved 70% on the process through digitalization. So we’re really trying to embrace this. And rather than taking a view that this is replacing jobs, we’re very much looking at it from a perspective actually it’s freeing people’s time up to better focus on the clients, to be more of a client-centric business.

12:39 RM: Good, thank you. So what does the partnership between you guys… What shape does that take? How does that work?

12:46 CG: So I think the pace has changed, which is to say it’s absolutely exponential at the moment, and the insurance sector will look very different in the next 5 to 10 years. I don’t know exactly how it’s going to look, but you get a gut feeling, instinctive feel that change is absolutely imminent and much need within the industry. So I think, first and foremost, we’re looking at proof of concept, so we’re working with Hitachi to find mutual customers. Hitachi are a very big organization, a very big outfit, like also QBE. So at the moment, just staying within the premises of the UK, looking at the customers that we have in common, and then trying to get a proof of concept of how we can feed data into the model, that also will influence the pricing model for the customer. Then, being able to get a cross-class view and feel for the customer across casualty, mortar, property, and financial lines, and then ultimately look to upscale across industry. So industries with an appetite… So construction, manufacturing, transportation, media and tech.

13:48 RM: But how does that affect the nature of your relationship with the brokers, and should brokers be worried? A difficult… If the questions get more and more challenging then, I…

13:58 CG: They’re getting more and more difficult. No, you don’t come here for an easy ride.

14:01 CG: It’s incredibly difficult to predict at the moment, and the views I can give are only mine and not as a corporate organization. So clearly, the tides are turning and the winds are changing with the likes of Amazon, with the likes of Google, and Hitachi on the doorstep. It only makes sense to create partnerships that are only going to advance the organization as much as possible. If you look at customer segmentation, which is very important to QBE, if you look at the top multinational organizations around the world, I think within our careers, our lifespan, they will certainly still remain a place for brokers as a risk advisory capacity. I think as they progress and collect more and more data, they may look for alternative methods for insurance and adding risk advisory to customers, so this could be superior reinsurers, other capital market providers within the mid-market, it could very much be… Brokering could become commoditized. And so clients with the availability of data may look to go more direct with insurers and then at the lower end of the spectrum, the lower end of the market, I think it’s only going to be an artificial intelligence solution. So with the likes of Apple, Amazon, and Google again, moving in other tech giants in the market, it’s just going to become automized for small commercial businesses and for the domestic market.

15:26 RM: This is an easy question. If it’s trains to start with, thereafter other extra classes, if it makes a difference, then you’re going to take it, right?

15:34 CG: Look, all we know is what clients tell us, and the clients are always looking for an easy solution where they can place all of their insurance in one place with as much convenience as possible, and clients only want to be understood as much as possible. So, the more data that we can collect, the greater the partnerships that we can forge and use that for our commercial advantage, the better, I think.

15:57 RM: Thank you. Thanks. Stephen, over to you. So we’ve had you here before and would I summarize rightly by saying that you do on commercial property what Hitachi do on trains, sort of commercial property IoT?

16:11 Stephen Chadwick: Absolutely, yeah, spot on. Thank you, Robin. So yes, Shepherd do leverage IoT devices to extract data from essentially dumb objects within the property environment. But as we know, data is only the beginning of the process. It’s conversion of data to value and it’s that route from data being analyzed into information, and information being shared to create new knowledge, and the ability to be able to delve into the issues, and the risks that are exposed from understanding how essentially a dumb property performs in its life cycle. So Shepherd is really around looking at the risks that are inherent from the mechanical and the electrical components, that sit within systems within a property, and being able to expose those and understand how preemptive maintenance could be effective, how the risk can be reduced by understanding the performance of the property, and how cost and value can be reduced in terms of the OpEx costs of the property in terms of reduction of energy usage, reduction in water usage, and be able to change the maintenance cycle of the property by understanding the performance and the metrics of the performance of the equipment that sits within it.

17:38 RM: Now you are at sort of 12 months, maybe a bit longer into engaging with the insurance industry, how far have you got and what’s been the experience so far?

17:47 SC: Thank you, Robin. So we’ve had really good traction with a good number of insurance carriers. We’re working very closely with Aviva, with Zurich and with Ecclesiastical. And we’re in conversation with a good number of partners along with QBE in terms of how we take these proof of concept ideas forward. We have in the ground some nine projects with the major carriers today. And these aren’t short-term projects, these are long-term, whole building approaches, to understand the performance of property with the value for the clients in terms of reduction of OpEx management of the property itself, and the value for the carrier in terms of understanding the risk scoring of a property and how that risk is reduced to being able to understand and exposing those risks, and manage them over time.

18:42 RM: So, in the course of those 12 months, how has your proposition evolved from the original one you brought to market?

18:52 SC: The proposition has changed in a number of ways. I think first of all is the route to market in terms of exposing the whole property. Originally we were looking at a single risk, so now we have two approaches to market. The first is a set of portfolio products to look at individual risk items such as leak detection, such as fridge freezer, cold store failure in academia, in schools, and also in retail. From looking at Legionella monitoring for a classic risk within any commercial building. Vacant property monitoring, and electrical subsystem monitoring. So those are the single products. But in the solution side of the business, it’s really been the traction with the insurance companies that has driven this whole building approach and being able to look at all of the risks that are inherent within the property and be able to take essentially dumb items and expose their lifecycle, their lifetime, the expectancy and be able to plan for when those items should be replaced in their normal operating cycle. That’s really driving the approach from the insurance carriers.

20:05 RM: So have you got to the point yet where you can offer what I call IoT as a service? In other words, the benefits of having the devices in place and your analytics in place are paying for themselves through increased efficiency energy use and so on. So that actually for an insurance point of view, it’s kind of the customer has already paid for it effectively as a…

20:29 SC: Absolutely. I think in terms of the original go to market customer that we had, it was very high CapEx because of the hardware for something that essentially customers didn’t believe could happen. This is taking an education within the market, of having to educate customers and end users that they can actually understand the performance of systems and assets within their environment. So, I’ve changed the approach to be Shepherd as a service, the real InsurTech-as-a-Service type approach, and combine the capital cost into the lifetime cost of monitoring, to be able to provide very low cost value for each individual asset within a property that is being monitored. If you monitor all the assets, we understand the performance of the system, say that be air conditioning, that be heating, whether that be the water flow without the energy usage within the property and be able to start to deliver business information for decision-making to really allow property owner and property occupiers to make decisions on what they will update, what they will maintain and how they will treat the performance of the property, which really touches upon the risk elements for the carriers within the insurance.

21:46 RM: Thank you, that was great. Call me an old cynic, but if you want to sell something, you better wrap it in wellness or environmental benefits and then they… It sells like hotcakes. Selling as an insurance proposition, you know, any… God knows. Thank you very much, indeed…

22:00 SC: Thank you very much.

22:00 RM: Everybody appreciated it, really enjoyed it.

22:07 RM: So next up, we’ve got Mark Dennis, who’s the Global COO of Munich Re Digital Partners. Those who come here regularly know, I’m an unashamed fan of Munich Re and what they do. It seems to me they’re one of the few who has a proper sense of R&D and are investing for the future in the knowledge that the insurance model is going to change fundamentally. So…

22:35 Mark Dennis: Thank you.

22:37 RM: No, no. Look, you get a free plug if you come and give up an afternoon to come with me. So first up, how long have you been going and how is it going?

22:44 MD: We’re around three and a half years old. Andy, Ray and I formed the business at the start of 2016. We’re now 80-something people globally, I think about 85.

22:56 RM: And in that time, what have you changed? When you came in, you thought you knew what was going to go on, how is that different now do you think?

23:06 MD: The fundamental pillars of what we do haven’t changed in that three or three and a half years. We still provide capacity, we’ve got global reach, we have a venture capital arm which is a kind of sister company within the group. We provide product and processing expertise and flexibility, if you like sort of an execution shop in some sense. I guess the market or the industry has matured to a point, so that’s probably where the change lies, but the business model itself hasn’t really changed which either suggests we’ve got it right first time, which is perhaps unlikely, or we’ve kind of evolved it as we’ve gone along rather than a step-change, perhaps, I would say.

23:48 RM: And I said you’re doing incredibly well, but have you got any metric… How do we know you’re doing very well? I kind of think you’re doing quite well, but are you prepared to share anything with us?

23:57 MD: Well, you just have to take my word for it.

23:57 RM: No, we’re simply not prepared.

23:57 MD: So we’re not very public on numbers, but we wrote around 100 million in premium last year. And I know in such esteemed company, it’s a small number, however bear in mind, that’s from kind of fairly immature InsurTech startups just starting to gain some traction. So we’re around 100 million last year, about two-thirds in the US and a third in Europe, which really means the UK first at the moment. We expect to pretty much double that, or more than double that this year, and then year-on-year.

24:34 RM: I want to ask you the lemonade question, which is: It’s all very well saying you are growing really fast, but if your claims are growing even faster, that’s not cool. Are you pleased with the results despite the growth?

24:49 MD: I think we’ve surprised ourselves a bit. So I think we wrote more business than we expected to write in 2018 and we launched in ’16, which was very much about experimenting, and then ’17 was about does that model scale. And then suddenly, we’re writing 100 million and it’s… That was a bit of a surprise. So as you’d expect off the back of it, there’s a bit more scrutiny around what does that business look like? So we continue to grow and we’re not putting the breaks on it in any sense. In fact, we’re accelerating still, but we’re learning lessons. So for example, we’ve built a lot of data infrastructure which allows us to react quite quickly to anything that looks a bit spiky in the business, but as I said at the talk, our core business hasn’t really changed. And one of the key tenets of that business is patience, so we work alongside InsurTech startups who have a limited runway, and our job, if you like, is to be first of all to get them to market as quickly as we can. And once we’ve done that, to be patient with them because there’s no use to me looking at the numbers after six months and saying, well, it’s not quite performing where I expected it to be ’cause in reality I didn’t know where I expected it to be.

26:00 RM: It’s all very well me being impressed with what you do, but are your owners impressed and what is it that impresses them? Is it the growth thereafter or is it what you can teach them from your effectively acting as their experimental lab?

26:18 MD: A bit of all of that actually, in truth. We were initially described as a strategic experiment, which sounds a bit odd, but really, as I mentioned before, it was, can this kind of large corporate work with InsurTech startups in an effective way? I think we proved that quite well in ’16. Then it was about scaling, and then it was about, can we grow that business in an upward trajectory without massively increasing our expense base? It starts to look a bit like a genuine business. I think there are three things that we aim to learn from the business. One is, can we do it? Can Munich Re work with tech startups, essentially, and my job in my business is to get that gearing in the middle right, so that all the pieces move at the right kind of time. The second one is actually about, can we actually build a business? Essentially, we’re a startup within a corporate environment. And I think we’ve done a pretty good job of that. As I said, we’ve grown from two to 80-something people, to about 20 partnerships over those three years. We’re definitely doing something right. And I’ve forgotten my third one now, maybe there’s only two lessons, is it?

27:34 RM: Anything out there at the moment that you find exciting, in the InsurTech world, obviously?

27:41 MD: We don’t talk about the others. Yeah, I suppose we’ve touched on it a bit in the initial session, there’s a lot of stuff around data and connected things, whether that’s at home or cars or other inanimate objects, perhaps. I think that’s clearly a thing. By the way, I should tell you, I should give you an idea of numbers. We’ve met something like 1000 InsurTech startups, over the three years, we work with about 20 or we got 20 live. We’re working with maybe 50. It’s definitely a thing still, there are emerging themes, I think so data’s clearly at theme, I don’t need to tell anyone in here. Subscription-based models, probably a theme. Platform plays is definitely a theme and probably that’s going to be a topic for discussion, perhaps later on. There’s a sort of consolidation to some extent, I think maybe there’s… I haven’t seen anything… I’ve got to be careful what I say now, but I haven’t seen anything massively exciting in the last three, four months, perhaps.

28:45 RM: That seems to be… Is a theme, would you agree? Yeah, no, I think it’s all become very partnership… It’s become a bit cosy, and I don’t see people coming out with deeply disruptive models at the moment.

28:58 MD: Yeah. Except this… The thing I have noticed changed is that insurers are I think… Forgive me for speaking for an entire industry, but three years ago, I think they were slightly suspicious of the InsurTech trend because, “Are they encroaching on our territory, are they going to make us look slow and terrible?” I’ll hold my judgement on that. I think now there’s definitely that much more of a partnership focus, so there’s much more of an emphasis on actually how can we work together? And if I look at Munich Re’s competitive landscape, I can’t… And it’s hard for me to define that now, because are Swiss Re a competitor, well, yes, kind of… But actually, we collaborate on a lot of stuff together and we actually, we have done for a long while. Are InsurTechs encroaching on territory. Or are they actually creating an opportunity? And you can argue either way depending on which side of the fence you sit, but I think there’s much more of a kind of collaboration approach now compared to where we were say, three years ago.

29:57 RM: I think I was… I can just speak for the whole industry as well. We see ourselves, InsurTech, as an opportunity. Look, Mark, thank you. There’s a couple of minutes for questions if anyone’s got one.

30:09 Audience question: You were one of the earlier investors in Trōv. Are you able to comment on Trōv in the UK, ’cause they recently announced that they were pulling out of offering their services here. And does that have any implications for your wider portfolio plays?

30:25 MD: Trōv remain a strategic part for us. In fact, they’re in my office tomorrow. We’re having a half year review. You’re right in the sense that they’ve kind of withdrawn from the single item cover Trōv-branded product in the UK, but their focus is more kind of distribution deals through the same platform, so they still back, if you like, the idea around single item cover. That will play out how it plays out. They also actually have a bit of a focus now on fleet business, so fleet as a service, as they describe it so they start to diversify, and in a sense, that’s a maturing of them as a business, but certainly, from our perspective, they’re a strategic partner still. We’re fairly heavily invested in them as well, which is a separate decision from my part of the business, but again, that remains a key part of our relationship with them. Did you have a view on single item cover and whether that’s still a thing?

31:33 Audience question: Hello, I’m James. Does Munich Re Digital Partners invest or is it the only the venture funding team that does? Do you insist on taking equity share in the collaborating partners you work with?

31:50 MD: No, we see that as complementary, it’s a part of our offering, but it’s a separate part of the Munich Re business. It’s called… Was Hartford Steam Boiler Ventures, it’s now called Munich Re Ventures. And of the 20 partnerships that we’ve got live, I think we’ve invested in maybe seven of them. Roughly a third, you could say. It’s not pre-requisite, the partnership from our perspective is paramount and then investment may or may not follow but is a separate decision.

32:19 RM: Time for one more, if we got it.

32:21 Audience question: Just curious how Munich Re defines success in partnering with different types of startups ’cause you mentioned there are thousands and thousands of them, and you have 20 partners, 100 million in premiums, how do you select these people, on what basis?

32:37 MD: A lot of it… If I think back to the early days, it was a lot of it was about a were we wowed by the idea. This, I’m talking three years ago now, so it’s fair to say we were easily wowed, perhaps. And then it becomes much more about the team around the idea, have they been successful previously with an exit, and potentially in another sector.

33:00 MD: Much more now, it’s about a much kind of more rounded view of that proposition. Do we feel we can work with these guys? Do they have some kind of guarantee of distribution? Have you got customers, or you’re just having a great product without a customer? Do they know how to get stuff over the line? Do they have insurance expertise, believe it or not? And I think it’s fair to say some of the startups we met, not ones we’re working with, but some of the ones we met in the very early days were trying to disrupt an industry without any kind of knowledge of the industry. And you could argue that’s a strength because then they’ll be bringing new ideas but in a highly regulated industry that’s difficult to navigate, having a knowledge or at least an appreciation of that is helpful. So financing is another factor. So have they got enough money to get across the line and launch before they run out of runway? So it’s quite a lot of factors. In fact, I’ve gotten some of my team in the front row and they haven’t heckled me yet, which must be bonus season. So they’re around. If you’ve got an idea to pitch to them… Rich, you’ll be here, right? We’re always ready to… We’ll have at least one conversation with anybody.

34:10 RM: No, so if you’ve got a pitch tech ready, now you know what to put in it. Mark, thank you very much indeed. Really appreciate it, thanks.

34:25 RM: So next up we have Nikolaus Sühr from KASKO. Now, we had Nikolaus here… So, we formed InsTech London in April 2015, and Nik was absolutely one of our very first pitchers come about September, I should think, September. You were using the word MGA as a Service before it picked up the same currency that it has now. You’re still going, which is a good sign, and you’ve been very successful. What’s the secret of your success?

35:05 Nikolaus Sühr: First of all, thank you for having me. I think at the time we actually did call it InsurTech as a service, so I was really happy to read the headline of today’s event. I guess, so what we do is we enable insurers to bring products to market, scale them up outside of their own IT infrastructure. And I guess what we are really proud of is with, I would say, it’s not really a shoestring budget, but mostly self-funded, we managed to get a lot of insurers onto our platform. And I think there’s two sides to it, one is to say you’ve gotta understand your customers, solve a problem, product/market fit, da-da-da-da. And I think there’s another truth to it. And one was sheer luck. So we started out 2015 and that was when everyone wanted a new shiny thing. So we were able with very limited, I would say, technology bandwidth to just scoop up these innovation… Stuff… I don’t know. Watch insurance with photo AI. Would not build that again. But that kinda got our name on the map. And I think the other one is we made a decision very early on to… So for the sales cycle, which is really complex in an insurance company over the building a product. We have a product, but for us it was really key to sell to the product managers and make the scope so digestible, that no one would even think about doing senseless RFP. And that’s, I guess, the secret.

36:53 RM: So how do people use you? Are you providing a digital journey end-to-end or are you doing what I call sticking digital lipstick on a legacy pig?

37:10 NS: I think we’re the latter quite frankly, because we don’t want to reinvent the wheel. I think there’s good reason to migrate to new infrastructure. So, yeah lipstick on a pig, I guess is the right word to put it. And really what we’re doing is we are digitalising an insurer from the touch point out. So we don’t actually start with the consumer, we start with the user who’s usually an intermediary, and figure out how do we bring a product to market, how do we onboard partners, how do we scale volume, how do we improve profitability on the underwriting, how do we then do the OpEx? So yeah, we kinda, if you like, we put a Chernobyl-like fencing around their core and make it work for the future.

38:02 RM: No, that’s nothing wrong with that. That seems to be where the industry is right now. So where does your best opportunities lie? Are you talking mostly to insurers, to MGAs, to brokers? Who hunts you down if they want to use this?

38:18 NS: Insurers. So classical SME and retail insurers who have more opportunity than IT capacity. As a caveat, we talk to MGAs because we have an MGA stack. We talk to re-insurers who want to help bring the capacity to market. We talk to very large distributors like banks, but it really… 95% of where… Of money comes from is insurance companies.

38:44 RM: Interesting. So I don’t remember many pitches but I remember you doing one at the DIA where you were showcasing something you did with Baloise on watches which I thought was really cool. Are you still doing that or has that died a death, and what have we learned from that generally?

39:07 NS: Yeah, okay, so that was the aforementioned AI-powered watch insurance. The one thing I didn’t want to build, but I’m really glad that we did. So I basically started out with a showcase, got a lot of stakeholder involvement, made Baloise, I think for the first time in the history, the most innovative insurance company in Switzerland, now fast forward, we rolled out a single item scheme. We went for distribution as well, and if you were to extrapolate Swiss numbers to UK numbers, we’re writing something like 7,000, 8,000 policies a week with that program. And let’s just say it all started with this bloody watch insurance and then just kind of cascaded out. So you gotta start somewhere.

40:02 RM: Yeah, but you’re insuring Swiss watches, whereas if you insure UK watches, you’d be doing something fundamentally different I think. So how do you do your marketing? You talk about cash strapped. How do you put the messages out there and why do you come here when you had to get on a flight and come a long way and present yourself?

40:27 NS: So I think because we’re B2B business, it’s really important to meet people to understand the respective decision makers, their portfolio of initiatives. One of the best thing to do is an event. Not any event. Events such as yours, which to us, just makes it really easy. You’ll just filter, you filter everything out for us so when we decide for an event, we’re actually more… This is what’s interesting to the audience rather than the speakers, so that’s how we bring the message out. And the second one is customer referrals within the group, within departments, has probably like 100 people at Allianz you can talk to and they talk to each other but events and referrals.

41:21 RM: Trick question. How many of those 100 people at Allianz have you spoken to so far?

41:26 NS: 97?

41:26 RM: Alright. That’s about right. Two more questions, quickly. Do you think that you’ll always be just lipstick on a legacy pig? Or are the aspirations to start being digital digital. To some extent, you’re not in charge of your own destiny there, but presumably, this is a step on a journey where at some point you’re going to go the whole hog.

41:54 NS: No, and so of course we believe that whatever end-to-end is… But I think at that inflection point where you then go, once you’ve maximized scale underwriting profitability, which I think happens outside of the classical legacy core and we heard about data and IoT, and all that. And at that point you’ll then decide whether you actually… Well, your policy administration system might be really good at administering policy, but really bad at data connectedness and workflow. And you can make that decision. But at that point, you can also pass into a greenfield carry system. But the nice thing is, at that point, you know exactly what that looks like. So you build your target operating model based on operating in your target market, rather than guessing without the use of the market and the rate of change in mind. So we don’t plan on going down becoming full policy administration system like Guidewire, although we have some capabilities, but creating that connected layer, where insurance underwriters… Ideally, if we do our job well, then in five years time that job will look much like that of a hedge fund manager, who gets a lot of data, publishes products, disperses them to the different risk pools without really caring about how the plumbing works. So we’re here to do the plumbing.

43:29 RM: And then what next? Are you flush with money? Are you looking for more money? What’s the rest of the year look like for you?

43:38 NS: So I think we have two key milestones. One is we need to… Well, three. Continue the business, grow, grow, grow, grow, grow. And deliver. But two caveats to that one is especially growing our existing customers. So we showed that we’re really good at landing. But now we need to expand and that just takes time. I cannot buy an insurer’s trust with five more sales people going to the same account. So that creates… So we need some time and the other one we’re publishing more and more self-service tools to open our platform for external developers, especially so that either the insurers, IT vendors or IT consultants can program on our platform. Yeah. So that’s it.

44:29 RM: We got time for a couple of questions for Nik before we return to the bar.

44:35 Audience question: Hi Nik. So this is probably part of your secret sauce, but you claim that you launch a product in four to six weeks. What does your tech stack look like? Are you able to share that?

44:52 NS: In terms of the functionality across the value chain, or what do you mean by tech stack?

45:00 Audience question: Just is it in the Cloud, AWS, Microsoft?

45:04 NS: Both. And it’s all shared instance as well. So we can leverage a lot of DevOps investment around that. And the other secret sauce, is that you’re basically what insurer would call a front office application, you move a lot of policy administration functionality in that front office application. So you do the whole quote of a bind, renewals, cancellations, refunds, payments. You move that into the front office application. With that you have a controlled environment, and you can then just build out the APIs and the customer journeys, all of which you control, and you just data dump back into the insurer.

45:47 RM: Nik, thank you very much, indeed.

45:49 NS: Thank you.