Who captures value in the next insurance stack? - Nelius Strydom, CEO at ⁠Seamless Insure⁠ | Eps.73 -Reinventing Finance Podcast

Speaker A: Foreign.

Speaker B: Hi everyone. Um, welcome back to another episode of Reinventing Finance podcast. As usual, I'm not all by myself today. I actually have um, three guests, um, in this room. Let's start with my co host, Ron. Ron, how are you today?

Speaker C: Hi. Great. It's a fantastic day today.

Speaker B: Um, with me I have um, Nilius and Matilda, um, from seamlessinsure. Very, um, happy to have you on the podcast. How are you today?

Speaker A: Thank you very much for having us and we're excited about the uh, opportunity to join you today.

Speaker B: Perfect. Um, before we get started, um, and um, what we're going to talk to, to you, um, amongst ourselves today is um, who actually captures value in the next insurance stack. Um, but before we jump into that and try to dissect that, um, for ourselves and hopefully for our listeners and viewers as well, um, why don't you briefly introduce yourself and your company and how that informs, um, the way you think about, uh, insurance, insurance value, insurance technology, etc.

Speaker A: Perfect. Uh, thank you very much, Nicholas. Um, so very quickly, um, I'm the CEO for Seamless. Um, I joined the company about two years ago. We um, are a native cloud core system provider. Um, what we have done over the last two years is to expand our offering and to be much more driven by ecosystems and platforms. And uh, that's really to adapt to the different need that we see from customers which have moved away from traditional core system, which is transactional in nature, but rather looking for something that is much more, you know, uh, ingrained into their business in terms of how they operate and how they actually go after revenue pools. Um, so we've done quite a lot of work around the infrastructure and the way that we set up the platform to enable our customers to really kind of mine new ways of insurance pools and clientele, um, and really kind of off at the forefront of how technology is really shifting uh, the whole insurance landscape.

Speaker B: And if you, before we kind of jump into the wider story, if you're saying it's more ingrained, uh, it's less transactional, it's more supportive of capturing revenue pools. Um, could you compare and contrast or give an example of what that um, would mean before and after or.

Speaker A: I think the traditional uh, model of core systems is to create different distribution channels which the clients then would somehow uh, kind of empower that channel, whether it was through an agent or through a direct website or whatever the case might be. What we do today is that we would integrate into providers. So as an example, we would integrate into a uh, parking garage or Events provider, uh, and actually capture the clients at that point for them. So rather than kind of wait to sell the insurance, kind of find the insurance. And to an extent, this is not new because embedded insurance has been around for quite a long time. But what we see now is that actually it's not just the traditional travel insurance that we do in that channel, but whether it is a camper van organization or whether it's a big events company, um, by the use of technology we can actually capture those pools and then you capture the pool as a whole and not as an individual client. So probably the biggest shift we've seen is that our clients want to capture pools of clients rather than capturing individual clients. And, um, to an extent, what we've done from a technology perspective is that ability to actually capture those pools and drive those pools in a far more constructive way maybe than it was done in the past.

Speaker C: So two minutes in, we already left that embedded insurance matters where, uh, we went all the way to profit pools and we went into the question, who actually captures value? So, niljas, let me, let me start with a question. Does the future insurance winner even need to be an insurer anymore?

Speaker A: I think insurers will always play a big party in terms of the risk taking and the risk offsetting as well as the pricing side of it. So I think the technical capabilities will never disappear. And I'd like to maybe use a practical example of how I would see this. So if you think about something like home insurance or SME, property insurance, or where you would need the carrier to basically construct the product, assess the risk, um, what happens today is that the carriers would need a lot of external data, whether it's flood mapping and things like that. So what we find is that a lot of those services are transitioning out of the insurer, but into the third party that is actually doing the distribution. So distribution has become more than just selling for commission, it is becoming selling for value. And I think that's one of the attractions for the insurance companies is that they get that selling for value. So do I have to be an insurance company to mine a pool of potential insurance? No. Do I need to be connected to an insurance company to provide the service fully? Yes. I don't think we're seeing the merge between reinsurance and insurers. Definitely not. But we do see that third party sellers and distributors are now doing much more than purely selling for commission. Uh, it's a completely more profitable share business rather than just a commission business.

Speaker C: So let me rephrase what you're saying if you only have to be connected. Are insurance carriers slowly becoming just nothing more than regulated balance sheets?

Speaker A: I think that is an extreme, um, I wouldn't say that's 100% correct, but I think it's certainly moving in that direction where you find that it's far more about the risk management and offsetting than it is about insurance itself. Uh, we have a number of clients and I'll use this client, uh, it's a Swedish client called My Safety. They predominantly sell protection products. Um, but their whole business model is about identifying pools of insurance and going after that. So whether it's parking insurance, event insurance, uh, or cyber insurance, what they would do is that they would find this distribution point, they would create the product and then there's a component in the product that they would need to insure or reinsure and they need the insurer for but 90% of the product. They've constructed how they sell it. They've constructed, um, the way that they would monetize that is not commission anymore, but it is a different fee. So they might have a giveaway, the underwriting fee, but there's different fees that they would keep on the product. So I think it's a bit maybe too quick to say that carriers are going to become just a legal entity like uh, a reinsurer. I don't think we're heading there. And the other caveat that I want to make is that a lot of this is not true in certain business lines. So when you look at, you know, specialized lines like Captives Marine and things like that, you know, what I'm kind of off selling now does not exist. So I think insurance is quite kind of nuanced. So I don't think everything that fits into a personal line product would work for a commercial or a marine product. So I think there's a, there's still a big role for insurance to play in certain more complex risk and things like that. But if you think about the personal line book and you know, even the plain M vanilla commercial book, I actually think that the product is becoming largely irrelevant. It's how you sell the product and you know, how you construct the product. Um, and one more comment and then I'll stop there. A lot of our, uh, MGA clients as an example, what they do is that they would negotiate limits, category limits with the carriers and they would then construct their profits within that limit. So a flood limit, theft limit or whatever the case might be. So we see more and more the third party sellers, uh, taking a much bigger role in the product and, and, and what they really need is a technical premium and then they run with it from there in terms of how they would finish the product.

Speaker B: I mean, and I, I would argue that, I mean we've, we've now kind of talked about embedded insurance and I think that's a very um, easy to, to, to imagine but the same mechanics, even though slightly different, um, differently wrapped you have with you know, employee benefits programs that are voluntary. Um, or even if you're looking at, I mean you mentioned Mariner speciality, you know, think about a real time uh, war or you um, know, dynamic product that is a group policy or just a fleet of trucks, ah, going somewhere. Um, and there are um, I would say that it's just, there's different, I would say limitations but also opportunities to design a program and we'll go into what the, what that might mean, um, per, uh, channel. And I think that variance just then requires different skills and specialization that traditional insurers who might sell home insurance through their 3,000 tight agents, which you might still have in continental Europe. Yeah, where the benefit of the insurance company, the assets, probably not the license, but the tight agents and the trusted network as they keep producing really profitable business. Um, you know, and I think it's just very um, very interestingly nuanced. Um, but trying to kind of um, pick that apart where value might be sitting in that. Um. In the preparation to our conversation you state something that the infrastructure is laying. Um, the infrastructure layer is winning the insurance race. Um, care to elaborate?

Speaker A: And I think this kind of fits with what we were talking about is that if you think about um, the technology you would use or the infrastructure you would use for embedded insurance or the, or for a scheme insurance or anything like that, in essence the technology is the same thing. So it is a set of modular services with a set of APIs. Um, and that hasn't changed dramatically. Um, what has happened though in the last two years, certainly from what we see from clients, is that the fast growth that we've seen in associated services. So uh, whether it's underwriting or claims, there's an ecosystem of small startup and providers now that provide really specialized services at certain areas. And traditionally in embedded insurance those were excluded because you were selling a simple product in a simple way through a seamless channel. But what we see today is that we're starting to sell more complex products where there's a need for external data that needs to come in for your dynamic underwriting or your distribution and things like that. And I Think why I'm saying that the ecosystem or the infrastructure is winning is that if you've got the ability to not only just kind of run a basic policy, like a travel policy through an embedded challenge, but I can sell you any personalized product in a channel and I can enrich that with additional data for pricing, etc. And things like that, then I've got a big kind of model that is difficult for anybody else to replicate because not only would they need to, um, you know, have the product, but they would also need to have this ecosystem that they can connect to their platform. And I think for, for me this is where we seeing the biggest growth, where people create these, these massive infrastructures that they then connect to insurance. So it's not a insurance product. And I'm going to go back to the my safety example. Their biggest, you know, kind of driver of their business growth is a front end and a middle layer that they've created that access all these pools and then bring them back into our platform. Um, but they've now got this kind of, you know, selling network which, you know, unless you've got the technology, you can't replicate it. It's not a tied agent where you replicate it through a relationship or a commission structure. This is only working because they've got a multiple set of ecosystems that they've connected and orchestrated. So I think that is becoming more and more. And we're seeing the product set that is being sold in these kind of channels becoming more and more complicated and therefore the infrastructure that you need around that is becoming more and more complicated.

Speaker B: Out of curiosity, because infrastructure is something similar to platform, um, that where you know, depending on who uses it, depending on context, there might be a different, um, interpretation. Am I right to assume that your definition of infrastructure is not just the technical infrastructure? It is and not just the technical capabilities to integrate. Would it also go into selecting different partners? So let's say I have a line of business, I have a, whatever. I have a cyber line of business. Not all of the data is local. Now which of these Darknet screeners do I use? Rather than API, you select, um, up to, um, how do you see the different chains working together? Um, and is there a specific preference like a right to play? We talked about the insurer could not the insurer also organize those, um, like, like what is. Where do you kind of see these things, um, um, working. What does it take to, to win? I guess.

Speaker A: Yeah. Uh, I want to start off with your question around infrastructure versus platform and so forth. I Think what we have seen is that that line between technical capability and business capabilities become completely blurred. So when I talk about infrastructure, I'm not talking about the set of APIs and, you know, hosting on AWS, but what we have seen is the ability to construct your technology in a way that it can interact with the outside world in a business like that is what I would define as infrastructure. So it is the ecosystem infrastructure, platform. I think it's maybe a combination of all of them, but we are seeing the lines of that blurring tremendously. Um, I think we're living in a world where having a modular stack with APIs is no longer enough. I think it's a deem that everybody has it now. You need to have something new. So I think the ability to create these ecosystems where you can plug and play a range of different kind of data, uh, vendors and at all times into the ecosystem, um, I think that is super important. And that's the bit where I think insurance companies, unless you may be the top 10 insurers in the world, I don't think they have the ability to replicate this. If you think about your tier 2 insurers across Europe and the Nordics and even in the Middle east, uh, for them to have a modular stack with APIs is already an achievement. They know where they actually can run ecosystems. And I think this is the bit where I think the third point he providers are really accelerating and they're really kind of building this connectivity at a much faster rate than what the insurance traditionally would do. Um, one last comment on this. I think we all fully painfully aware on how slow insurance companies are changing their core systems. So not something you do every five years, it's something you do every 15 or 20 years. And I think this really kind of contradicts the infrastructure play, which is about the ability to move quickly around different integrations. You know, if a new, um, uh, pricing, ah, engine comes on the market, I want to be able to connect to that within two weeks. I don't want to wait two years to connect to it. And I think this is where the infrastructure play in my mind is really kind of accelerating and moving past what insurers are doing from a tech perspective at the moment.

Speaker B: Moment I would. Sorry, Br. Go ahead.

Speaker C: Um, first of all, I was a little bit afraid that I get outnumbered here with two tech people on the other side. But I'm very, very happy that you almost took my side already, Nick, saying, well, isn't it about delegation, isn't about orchestration? Who actually holds things Together. So I would like to start actually with one question just to get a little bit clearer for myself. Why do you think embedded programs actually fail most often in reality? Because that brings, uh, us to what is actually the hard part of holding it together. And is it really the infrastructure? Let's see them. Why do they fail?

Speaker A: And I think this is an extremely, um, important point and probably something that is going to counter my own argument a little bit. But the two main things I've seen embed insurance failing on. The first one is underwriting losses. In other words, we're trying to sell a product so cheap that we can't just commercially it's not viable and eventually we shut it down or we try and increase the price and then the market is gone because you've come in so low and now you want to increase the price. Um, the second thing I think is more to do with the ability to force products down a channel rather than it being really embedded. So in other words, I'm trying to sell you motor insurance on a food delivery platform. Not sure it's going to work. Um, so I think those are the two main things, in my opinion, that has made, you know, kind of these models fold. But I want to go back to the point that I made earlier is that I think what we see is that embedded insurance is almost becoming old news. And what we're seeing now is much more sophisticated distribution platforms. And I don't think those will fail on either of the ones that I mentioned earlier because I think the pricing sophistication, the ability to bring in external data, uh, uh, the ability to have customer data means that you can run a profitable product. Um, and a lot of these providers are not incentivized by commission. So for them to have a product that really performs well is imperative to their commercial existence. Um, and I think this does change the argument away from it being product failures rather than technology failures. But I think your point is valid. I think traditionally these have not failed because of technology.

Speaker C: And let me rephrase again. If it is like this and it moves away from this, who ultimately owns the customer pain once it failed? Is it the orchestrator? Is it infrastructure? Is it what you said, the underwriter, um, or the distributor or a mix of it, who basically takes ownership, then

Speaker A: I think it's a combination of it. I think, um, as a starting point, the traditional model of the insurer bearing the pain. I don't think that's true anymore. I, um, think what you find is that a lot of these, and let's call them third party distributors for a lack of a better description. A lot of these are now investing large sums of money in the infrastructure and in their product knowledge and their ability to sell. And therefore there's a big financial imperative for them now to actually make sure that this is successful. If you think about the earlier embedded models, um, and I used to work for an aggregator in Singapore um and that kind of didn't work out but the person standing holding the hat at the end of time was the carrier because they took all the risks, they took the underwriting risk, they took the financial risk um, and the aggregator largely didn't have to take any of those risks. So they failed and you know they didn't carry the responsibility. So I do think it is changing a bit because of the investment that is done at the third party level and it's not. They now got skin in the game. It's not a question this is going to fail and you know we don't worry, we're going to find another pool that we're going to run to and try and sell to. So I do think it is shifting but it'd be silly to say that the carriers or the insurers are not carrying the bulk of this risk.

Speaker B: Yeah, I would, I would maybe add to that Ron that it's um, also it depends a little bit on your perceived history of, of failure. Um like I, we do get quite a few requests especially from like strong commercial um so if you're very strong commercially um, but you work for a large broker or carrier who still have large stakes in affinity embedded insurance, you name it, um, then there is a history of we couldn't get this program off the ground because of technology. Like there's just a long arc of history. And in that regard there is a, the programs either didn't get off the ground or what. We also know Ron, it's almost like this. You set yourself up for false negative. It's to say why don't you sell our standard product first and if that has legs then we'll invest into customization. So I think there is a certain level of, of that that plays into it. Then equally I think commercial um, misalignment um, especially doesn't um, have to be bad for Ryanair squeezing yet another carrier um, by putting volumes or you'll have the same thing with large um, price comparison websites who can um, if you can deliver large amounts of volume then you own the margin game and then depending on how you, how you play that um, then the program might be a uh, failure for the previous insurer but then the next insurer might price it slightly smarter without having to recoup past losses. I would then say, and I think you, you mentioned that as well, that in the middle a lot of it is um, I would say maybe product process design and activation kind of coming together and really understanding. And by that I don't mean a single person needs to own everything but needs to at least check. Have we thought about this and understand the um, implications. So which touch points do I have? What marketing activities would my platform even be willing to do? Which experience do they have in insurance? Is this what, how should the product be priced? Uh, what information do I know about the customer? If I go into more let's say retail like programs, certainly if I have MTPL product in the UK where I can just go on Money supermarket and get like a really cheap quote versus let's say more opaque markets in central Eastern Europe where you don't have strong, that you know all of these things kind of influence those aspects. And I said and um, at least in my experience the um, let's say the maturity of either the platform, you've got the third party um, distributor which you said potentially an intermediary, the MGA or the, the broker or potentially the carrier. It's, it's very varied the level of experience and expertise that you have um, on a given line of business. Um and so yeah, it's a bit like wild wild west out there. It's cool. Um, but someone has to bring these pieces together I think.

Speaker A: Um, sorry, I want something to that really illustrate that point is uh, I've been uh, unfortunate or fortunate enough to win through two telco insurance builds. Um, the first one was basically a build where basically we replicate the underlying carrier and we sell their product and that was it. Um, and that was semi successful. The second one actually the carrier hired a number of underwriters and what they did is they designed the product so they had different cashbacks and things like that. So they really build a product. You know this was a proper insurance product with the right benefits and the right structure. Uh then actually, actually testing the models to make sure that it will work and things like that. And that was substantially more successful because not only did they have a, the same you know, captured pool that they were selling to, but they had a real thought through product that they were selling. They were not pushing the old school commission model where I'm just going to be, you uh, know you're broke and I'm going To sell the products. They actually did quite a lot of on the design and the product. And to Ron's point, now for, for, for this telco, there is. I don't want this to fail because I, I've built this insurance business, I cannot afford to let it fail. And that's a big protection for the insurance company because, you know, they do up to technical premium and you know, the telco would do everything after that. So I do think there's a big shift away from selling, you know, what I would define as selling for commission and selling for value. And I know selling for value is a bit of an overused word, but if you actually really take care to build a proper product and you then build the infrastructure to manage that product, um, then I think you've got a proposition that, you know, A, you'll make good money and B, it'll be difficult for people to, you know, kind of off take that from you and I totally buy this.

Speaker C: So if we bring back why do embedded programs fail? I, uh, mean both of your points were perfect, right? Uh, Ryanair. Somebody has to own somehow the economics end to end. That uh, doesn't have to be the insurer. Also doesn't have to be the infrastructure player. But activation is important. Activation goes hand in hand with probably product iteration cycles. Uh, you said the insurers are slow, so I'm still not 100% sure. But I do understand against the two of you probably today, I won't win right away. I still challenge whether the mode in embedded insurance is really the infrastructure itself, but rather the ability, at least that's what I think, to continuously coordinate change among all those different players faster and better than everybody else. But I would like to go back. Unless, uh, you, Nick, have no.

Speaker B: I mean, I think we all agree, um, and the interesting thing is all of these things, um, wherever there's friction on one part, the whole flywheel just grinds to a halt. And, and my, what I find really interesting is um, there's probably an even more competitive race right now and that's all of these AI labs and then Claude code just, they leaked their, you know, they leaked their source code which is like the holy gray, the holy motherload of something. It's like it doesn't even matter. The source code is already updated two months, you know, two weeks in, the source code is already out of date. So whatever we thought that past experience, that past knowledge, that past relationship, all of those things matter. But it's really, if you can um, escalate, um, faster, um Iterations. Also stopping certain stupid AB tests just because someone thinks an a B test is great and you should test certain things, you know, if you can kind of come with shortcuts, um, and kind of putting these things together. And that is all of that.

Speaker A: Right.

Speaker B: It's understanding the platform, the economics, the product, the process, the uh, technology, the value chain. What does it take to win on a given product? Because a product with really low, um, bad economics is not really sexy for everyone. And some place this needs to be orchestrated. And I would argue that whoever plays in this game, everyone needs to play a dual game. They have a product and a pipeline and that might be the insurer, that might be the API provider, that might be the distributor computer. And the further they understand everything that that happens and are willing to insource that in terms of a orchestration without making everything stop. More expensive, um, trying to play a gatekeeper, traditional. I'm sitting myself on this value, you know. Yeah, orchestrate everything. I don't really do anything. I just get 10%. Um, then I think that is where a lot of the differentiation and ultimately the moat comes. It's, it's almost like the continuous reconfiguration of what happens if you'd start fresh or you know, uh, today. I think that's really where we're, where we're heading. So after even forget having a mode, what would you do if there wasn't a mode? What if you just had to show up every day and really work through it again and really have to prove yourself again? And what if that was the way you would conducting yourself in a business environment rather than reaping of past returns? I think that, I think it's in that direction where the future mode comes.

Speaker A: Yeah, I am, um, Ronan, I know you wanted to move to back something. I just want to make two very quick comments on this. I think the infrastructure enables the ability to move quick. So I fully agree with you. I think success is the ability to continuously iterate, move and adapt. But I think the infrastructure allows you to do it. And if you haven't got the infrastructure, you can't do it. The second point, which is a bit more subtle, is a lot of when a third party goes out and he's looking for capacity to sell, um, a lot of the carriers now place a huge amount of value on their infrastructure. How are you going to sell? How does your system work? So traditionally they would only give capacity looking at the risk in the portfolio. You know, risk and portfolio is still number one. But the actual infrastructure that you have around. Your proposition is becoming more and more and we see this where clients are now actually taking us with the, you know, to the carrier conversations where they're going to talk about capacity. They actually say no, no, no, you need to come along and you need to, you know, sell the technology angle to the carrier as well. So I think the carriers are looking to that infrastructure, technology, platform, whatever. We want to define it as a component in the decision making process to release capacity. So I think it is an enabler not just at a dimensional but at multi dimensional levels.

Speaker C: Okay, one question we will probably not answer today, uh, is then if infrastructure there's that important, then why do so many infrastructure players still struggle economically if this is supposedly where all value accrues? Not sure about this. Yeah, I would rather like to come back to your example because it's one which is close to our hearts as well. Telco assurance and their distribution comes into place. Right. And in telco you see what we just said, uh, activation beats product sophistication. Products are fairly simple, contextual elements are important, timing is important. Journey integration is definitely better than having a marketplace on screen number three of your app somewhere. At what point does this kind of embedded insurance then stop being a growth strategy? And I'm taking the insurance perspective today. So being growth strategy for an insurer and just becomes outsourced commoditization, if you want to say so, because the products are really simple.

Speaker A: I mean, ah, if you're talking about, let's use travel insurance or handset insurance for the telco as an example, it's already a commodity. Um, you're not selling an insurance product, you're selling a, you know, add on to the phone. It's like I'm, I'm going, I'm signing up with a telco, I get my handset, I get a, you know, earbuds and you know, maybe I get a screen protector. The insurance is just a commodity like that. No, it no longer adds value to the insurance component of it. And I think travel insurance to a large extent in my opinion is exactly the same. Same thing it is. You know, I can select my seat and I can order my meal and I can buy insurance. So I think it's already commoditized.

Speaker C: Do I have to be cynical then and saying, well we as insurers are um, not building ecosystems anymore. Ecosystems are basically just taking our margin away.

Speaker A: Uh, and I want to go back to a point I made earlier is that I, I do think that it's changing quite rapidly away from commoditized products to a more complex product. So want to go back to. So uh, um, uh we've got a client that basically sells personalized insurance to a Swedish client as well. Um, what they've done is that they've gone to small associations, uh, the Campervan, I think I m mentioned it earlier. So they've gone to a number of these associations and they capture that association and then they go back to the carrier and say I've got a pool of 10,000 clients. These are the six products they're going to buy. Uh now let's work on a pricing. So it's not a group scheme, so it's not being sold as a group scheme. It's being sold as individual policies and the customer can voluntary take those up. I think the carriers like those because it's not a margin game, it's not a pricing game. It is a game of customer acquisition cost at a much lower rate than what they normally can do. And I think maybe the right way to look at this is to say that embedded insurance is evolving and becoming a grown up product rather than a simple product which is commoditized. And that's certainly what we see is there's a huge demand for group schemes to take multiple products uh, as a scheme. And the Nordics is called the scheme so they've got all these different schemes that they run. Um, and that seems to be growing the fastest from an end market perspective. Those schemes are certainly the fastest growing pool that we see.

Speaker B: I mean I would, what I would say is that um, if you look solely on embedded insurance, on selling hand sell insurance ticket or God forbid PPI or you know, some, some other um, products which by their definition you know if you carve out 50% of a commission, um, out of it, you can question um, the attractiveness of that product. Um, I think that's a simplified version. I think what you'll really learn um, in embedded insurance more so than others that insurers have been traditionally built products and then distributed them through a series of smaller agents or brokers. Um, and most of them actually did not want to have, they want to have standardized processes, they wanted to have comparability, the agent wanted simplicity, the broker wanted to have comparability. Um, and that was how you designed your product. So you would get, it would be crazy to continuously update them now, um, where we see the market moving, um, you know, non traditional, you know, embedded players. Okay. Um, but also you know you have MGA's coming so there seems to be an increasing demand for product customization based on a specific scheme or something. And to me embedded insurance is just an already existing value or you know, revenue stream or customer stream where you can learn how to do these things well, which will then be propagated much more. So um, once we're getting, you know, once people actually start buying and shopping through Claude and OpenAI and you'll have mass customization and you know I would, the next time I'll go oh, uh, I'm now my daughter got a bike. Should I update my bike insurance? I'm not going to, I just included it. But if I had an agent run something to save me 20 bucks, like I'm not going to do that if I'm saving myself 50 bucks or. But if I can instruct an agent and go continuously save this thing and continuously update, then they think that embedded insurance is a really good bridge towards that um, um, towards that future. And I think that is ultimately insurers will continuously um, demand um, mass customization truly of their, of of their products. Um, and where they're going to really make their money I believe is embedded insurance, commercial insurance, employee benefits, affinity schemes. But, and then you know, once the tide, um, the tide breaks on um, agents actually buying insurance, which we don't have today, um, but I'd find it very nearly as you said, ah, a core system. People buy something for 15 to 20 years, um, and it's like how do you even make a technology decision for 15 to 20 years when the ground is shifting like weekly and you know, you have a new plugin from Claude and all of a sudden something writes off and you know, guidewire already lost 50% of it um, of its stock price, um, and that's a crazy space. I just think embedded insurance is an actual applicability of these things. So that's where I would.

Speaker A: Yeah it does. I know this is slightly off track but I think in my opinion what this does is that the smaller Tier 2 insurance companies are for the first time in, you know, in the 20 years that I've been in this game in a position where they actually can really attack the big guys and steal market share because they haven't got these long cycles, they haven't got these models that if it's something small, break the model, break. They can be much more agile in terms of what they do. Uh, they're more open to different distribution models and things like that. So I think what we have is that the infrastructure is leveling the playing field to an extent and giving the Smaller guys, a much bigger ability to punch way above their weight because of the fact that they've got this agility and technology. So I think it's a super interesting time. Um, I'm expecting lots more challenges coming up now over the next two years merely because as you say I don't need to build a system for three years, I should build it every six months. Uh, and I can be much more competitive than somebody that's got the stable system for 10 years. So I think it's a super exciting time uh, where technology will really determine success. And hence my opinion that I think infrastructure is the key driver for value at the moment and it puts all

Speaker C: of us under pressure if you want to say so. Because if we summarize that activation and I'm, I'm not yet taking aside whether agent activation is enough, I still, still think we do need customer activation. Um, if that beats product sophistication, um, and the ecosystem actually owns everything, right? The customer, the interaction, the data and the activation, what is left for the insurer, be it tier two, be it tier one. We have to really make sure that we position ourselves rightly and that definitely is away from product first, it's definitely activation first. But there are also other competitors there who also want to own the activation. And then we have uh, the entirety of Nick and, and uh, agents actually doing all of the purchase.

Speaker A: Mhm. Yeah.

Speaker C: Difficult.

Speaker B: I, I mean I would, I don't think that uh, insurers not owning distribution, I think that is the norm rather than a new era. Um, I would argue there really isn't

Speaker A: any,

Speaker B: really isn't any direct ownership of a customer. Unless maybe you have your own employed salespeople which most, you have some of them in Switzerland but most of um, even tight agent organizations are um, independent. Um, and you see that in terms of how they treat the CRM and marketing opt ins and how certain things go and what happens when someone all of a sudden switches from AXA to alliance even though it's alliances, you know. Um, so I don't think that has changed. And I would also argue that if you go through Google or Facebook or price comparison website that you're actually going through a platform as well who kind of uh, dictate certain things. So I don't think anything has changed And I think insurers just need to continue to as they have to treat the different channels differently. Which is why you don't have m multiple insurers dominating all of the channels because the channels actually dictate so much about the economics, the activation. Um, and so I think that's just a logical progression. Um, the one thing that I think we're. And that was the point I was trying to make. I think the skills you'll need in embedded insurance where it is, um, you. Because the difference is you kind of orchestrating these programs with um, a large multiplier rather than many smaller, um, um, tight agents with that. I mean like, um, actual humans, not AI agents. Funny, um, that you need to say that these days it's like cow milk or um, oat milk. And you know, and that was cow milk. Um, uh, I now need to specify, uh, it's cow milk or. Anyway, um, but it is that element of, let's say, scheme design, product design, the holistic design element that someone needs to own. And I think that is one of the biggest transformational benefits of that channel over other channels. Because the other channel, um, is you're outsourcing a lot of that and I think now you need to keep insourcing it.

Speaker A: Yeah. 100. I couldn't agree more. I think that that growth and sophistication and that distribution channels is, is. Is a key bit that, that you know, we, we certainly seeing the insurance kind of, of shifting away from that and allowing a lot more of that to happen outside of what, what they normally would do.

Speaker C: And I, I would actually add now another dimension I, I think we discussed embedded and activation is important and products may be a little bit simpler, but also maturing, like you said, isn't the hard thing. And I see that every day at the platform partners, launching one embedded insurance product or product channel activation combination, if you want to call it like this, that's one thing. But isn't the hard part launching 50 across 10 countries, maintaining your operations in a, in a consistent way and after all, at least in the midterm, being profitable while doing all this. Isn't that the heart? Isn't that the hard part of it? And how does technology now getting basically back to you? Isn't technology then what could make a difference here and there? And being one of the reasons why money goes more into technology, although it's less scarce than into insurance.

Speaker A: I mean, um, if I go back 10 years, there was a term called country layer. And you know, everybody had a different country layer. I don't particularly like the term and I think Nicholas is smiling, so he probably, uh, shares my skepticism about it. And to an extent, the shortcomings in technology was masked by having country layers and country layer was nothing else than, you know, everything that I couldn't do in my platform. I put in the country layer and then my platform was okay. Um, so do I think technology can breach that gap? I don't think so, but I think there's a smart play to look at the world through different dimensions. Um, and I'll use the Nordics as an example where, you know, if I look at Norway and Sweden, the technology landscape is identical. So my motor registry APIs are the same. So there's certain, certain spheres across the world. So if you look at the Nordics, you look at the Middle east, you, uh, look at Central Europe, where actually the, the ability to replicate technology is there. And it's quite feasible to actually not have to rework or create a country layer, but to build on your platform as it is. So I think it's more about identifying where those regions are, uh, that you can use a bit more consistently from a technology perspective. Um, so that's the first thing. I think that is an opportunity, and I think it is something that we certainly have done quite effectively. The way that we split our Middle east functionality from our Nordic functionality, um, uh, is something that is doable, um, you know, maybe similar to the old country layer. Um, the second thing, and I think this is something that is fundamentally new, is now is that with the use of AI, um, the ability to adjust code for regulation differences and things like that is really made easy, um, you know, for me to, if I take Dora as an example, um, you know, if I have to give my supplier list and you know, kind of things like through that, uh, using AI to actually create those kind of different buckets in the code and things like that means I can actually kind of skew my code towards a specific set of regulations quite easy. Now, I don't need to rewrite the code, but I can say in this country, the GST needs to be treated in this way and it needs to be paid in a different way or calculated or shown in a different way on an invoice. Those nuances traditionally you would do for configuration or code. Code makes it expensive. Configuration means it gets so complicated you don't know what you're configuring and you make lots of errors. But with AI now, you can control that much more consistently. So I do think the pace at which technology changing is maybe kind of driving that a bit down. I do not want to for a moment pretend that assurance is consistent for every line and every country, because it's not. We know it's an extremely unstandardized industry across the Globe. And you know that that makes technology quite difficult to manage.

Speaker B: I would kind of add to that that I agree that um, the advent of agentic coding um has allowed you um, especially once you've gotten through the hard problem of properly contextualizing your platform um so it knows um, how to interact with it and doesn't probabilist things that you're not aware of. But once you've kind of went through, through, through that I think it's, it's, it's very powerful. So what it, that allows you to let's say pick up some, on some let's say um inaccuracy and planning or um just unforeseen events. So that's great. Um, but ultimately um, I think wherever technology, wherever business expectations weren't met with technology most of it was at the planning and expectation rather than the execution stage. Um, not to say that there also errors happen but it's like that Chinese whispers thing. And so to your point Ron, I think the biggest um, um de risker of unintended um complexity down the line as you multiply certain things is having that at the basically when the promises, at the very early stage when the promises are being made to your distribution platform partners. Um and that is basically Neili said it himself when his partners to the insurer kind of have him on board. And it's not just to say um, on what the technology can do etc. But it's also to kind of have that at the design stage because otherwise someone, a non technical user will design some really weird process that kind of makes sense for them. Um, but it just doesn't scale as easily. And that goes for everything that goes for how you design an insurance product. So that actually a claims agent has a rules based not um, a magic wand based of how to evaluate something. These things don't necessarily happen if you have a law firm design your underwriting criteria um, without any claims experience. And so I think it's putting that together and ultimately making people aware of certain design choices very early on. And I think that is the first thing is and I think it's, it's someone needs to feel like it's their job. Most people don't, most, most technologists don't feel it's their job to design and inform the actual commercial and um, maybe because of that the way that they, the business people don't want them in the room because equally you cannot have a naysayer and who is just like this doesn't work, this doesn't work. Give me, you know, you don't want them to kill the deal, um, either. Um, you know, same like a, like a, like a Harvey Specter lawyer rather than other lawyers that we.

Speaker C: I was just about to say this

Speaker B: is why I left.

Speaker C: This is why I left Big Law.

Speaker B: But you know, it's, it's, it's the same thing, right? You want to have that um, solution oriented piece. I would then say that the other advent that we find, um, where AI is tremendously helpful, um, is not just on the coding side, but it's on trying to conceptualize all of these really complicated, multi faceted things about product, legal, marketing, jurisdiction, you know, all of these things and just trying to put these together for the different decision makers to do. Because that's just a lot. Um, and um, then coupled with the technology boost that we all see through agentic coding and modern platforms who are already microservice so it's not that hard to put an MCP on top, um, is I think how you'll solve it. But it's actual intelligence, Ron, is how you solve that at the early, at the earliest. Um, I would say.

Speaker A: Yeah.

Speaker C: Is that where the value goes in

Speaker B: future and actual intelligence as we just said, right? I think so. I think so. Collapse.

Speaker C: We have AI. So where does the value migrate to?

Speaker B: Yep, I would say, um, it sits. I mean that coupled with. Okay, let's, let's play this out. So technology costs collapse some more sorts of. I don't think it's the end of SaaS. I don't think you can vibe code a core system because um, you still need to maintain things, etc. But let's say, um, it becomes less, your code base becomes less of a defensibility. Um, I would still say that relationships and identifying opportunities and making other people see opportunities and finding a collaborative thing, I think that there's a premium on that, just like it always was. Um, so I think that that remains, um, and I think that orchestration, that intelligence layer that kind of keeps those things together in a, uh, in the right form, um, is able to pull and push and pull from the different parties so they don't try to, um, kind of overextend themselves. Um, but take everything that's already there.

Speaker A: Right?

Speaker B: I mean we know this. If you want to run an international um, MTPL program, well, you're probably not going to recreate it from scratch, right? You'll take whatever is local. Um, and even if it means putting lipstick on a pig because it's just too difficult. Um, and just because you have a cool API doesn't mean that, you know, fraud differences from Romania to Switzerland to Singapore.

Speaker A: Right.

Speaker B: It's just not how it works. Um, and I think then it's that incessant curiosity and mindset to create ever shifting coalitions, um, and be humble enough to um, not try to kind of dictate where other people don't dictate. Yeah. So it's pretty mushy. Um, but that's where I think this is going to go.

Speaker A: Yeah.

Speaker C: Brings us back to coordination, orchestration, partner management orchestration. Maybe a little bit of governance, but

Speaker B: with intelligence I was about.

Speaker A: Yeah, sorry Nicholas.

Speaker B: So what I mean is it's not going to lunch is not what I mean. Right. That's not the uh, orchestration. It's like putting people together and then delegating everything away. I think it's, it's doing all that but coming with a meticulous footprint that other people can then say and assets and understanding. So, and I think that's, that's what usually has not been combined. Um, you don't really have an, a commercial organization and an engineering organization all in one and in one person in the same meeting. Um, that's, that's what I propose here.

Speaker A: Yeah, I 100% agree with that. I think you can never take the domain intelligence out of the technology. The uh, technology by itself is pretty meaningless if you don't infuse that with the domain intelligence. What you get is something that functions, you know, hollowly without any value in it. And I think um, luckily what we're seeing is the opposite is that people want technology with that business intelligence in it. To run a business, it's not about a technology that allows you to issue a policy. It is about the ability to issue a policy that will drive value for both the client and the carrier. And I think that to me is the bit that means technology is never going to be the sole provider of value. You're going to have to infuse it with that intelligence to make it work. Um, I do want to make a very quick uh, point on the shrinking of coding budgets and the speed to which you can develop now using modern tools. Um, in our case we um, certainly seen a drop in our development costs, uh, uh, quite a substantial drop in our, through the stuff that we've implemented. But the counterbalance is that my R D budget is balloon. So you know, it's great that I save on the coding, but I spend so much time now looking at new things, testing, because it changes so rapidly. I cannot Wait six months before I look at something. If something comes on the market, I need to be there tomorrow and see can I use it, is my competitors going to use it, what I'm going to do with it. So yeah, we see reduction in engineering budgets but we see a higher change in R D budgets on the back of the pace of which things are changing at the moment.

Speaker B: But that would corroborate I think that investment in. Because you know, intelligence or uh, connections etc. You know you need to work for those and that's basically reallocation of those budgets and you basically need to invest in those, figure things out, um, take your um, senior people's time away from that, you know, to, to um, away from kind of coding to, to those things. So I, I would say that's a, I think that's also one of the reasons why, you'll see why it's so hard to measure efficiency gains on these things because what do you then do with the free time and you'll iterate on something else. And so I, I think we are um. Yeah, R D is certainly try and error is certainly um, going to become much, much bigger.

Speaker A: Yeah.

Speaker B: Gentlemen, conscious of time is there? Um, if we try to wrap this up, uh, do we have. I mean I think we've identified certain roles and aspects, you know and let's just. I think, I think orchestration is actually a good encapsulation of it.

Speaker A: Right.

Speaker B: Um, do we feel that different market players and maybe we can say an insurance company, commercial broker, an mga, um, a distributor, um, like a platform, either a technology provider, maybe consulting player, I don't know, maybe that's a third party administrator. Um, which of these have a. Do we. Maybe starting with you Nils. And then would love to get your view on that run as well. It's like do we think that someone by the nature of their um, organizational DNA historically, um, or type of legal entity, do they have a righter right to play than someone else or is it just a strategy? Was it just a choice?

Speaker A: Um, I mean that's a difficult question to answer because I think behavior across different jurisdictions and different products are different. So I don't think there's a uniform. You know, this is better just from, from what, what, what. What I see is that in The Nordics the MGAs, um, because of their ability to do understand underwriting and pricing much better. I think they definitely are doing better at the moment. They are faster at identifying technology pools. They haven't got the legacy of the big carriers that they need to deal with. And we see them as really at the cutting edge of using technology to drive businesses and profitability. So that's in the Nordic space. Um, in the Middle East I still see embedded insurance as a big driver. Traditional old embedded insurance that to me is still, you know, the telcos, the third party resellers, dealership insurance, uh, we still see these things leading, uh, even though that's maybe not cutting edge but I still see that as probably the most dominant growth pool in that region. Um, Europe apart from the big markets. But if we look at the Baltics and central Europe, um, I think it's probably more brokers and agents, uh, rather than kind of MGAs. Um so I think it's quite dispersed across the different regions. And then per product, um, I mean our biggest business is in the Nordic so that's kind of where we have the biggest footprint. And right across the Nordics we certainly see the MGAs and the tier 2 and 3 insurance companies actively, you know, building themselves out around technology, you know, new methods, new uh, pools, uh, as you know growing quite fast.

Speaker C: Ron, I think we talked a lot about commodities, commodities right, or future commodities. Um, tech costs going down. M Tech's not the differentiator carriers. Uh, what do we still own in an ecosystem everybody actually becomes, for all the right reasons, a little bit anxious to get replaced in a way or not. So I think the future winners, and that probably brings us back to what we started this discussion with. Um, the future winners in insurance are probably not the ones with the best products. This is what we said. Probably also not the ones with the best technology or infrastructure. I don't know. But the winners may just be the players that can coordinate increasing complexity, the 50 products in 10 countries and uh, X product lines. The players who can do this faster than anyone else. So um, if we meet again in five years, Nelius, uh, I would be interested to see who of all those players have become commoditized.

Speaker B: Yeah, yeah, yeah, I would, I would also um, you know there's, there's obviously what looks simple on a kind of two by two strategy matrix and you know then real life with opportunities and inefficient markets and opportunities and all that. Um, but what I would say from, to answer my own question, I would say as a entity footprint I would, all things being equal, which none of them are, I find the mga, um structure because it has this um, ingrained this like commercial motion product design, motion, acquisition motion. Um, so in its design it has this like commercial trajectory um, and I think that's ultimately what it really comes down to is taking a stake into the successful outcome of this for all of the stakeholders involved. As you know, it sounds cliche but that's really what it is. And that's not what a TPA normally does. A TPA is not normally necessary for the growth of the program or the acquisition of it. An insurer, um, sure, to, to a point, but not really. I mean they let it for the distribution partners, a broker again, um, you know they, they think more about the deal making on the structure rather than necessarily on the operational side of things. A technology vendor normally also just said great, give me your problem, I'll give you a technology solution. Um, and so I think it's really with that, with that mindset and um, how to either um, hit yourselves towards such entities so you can then support them, enable them and benefit from the increasing capturing of market share. Um, you know, becoming one yourself or playing a dual play as well. I think, um, all of these you need to be one thing. You only need to be one thing. I think in a world which is around orchestration and complexity, um, I think becomes less strong because you need to understand most, most of the aspects. And for that it actually means having tried to solve it yourself at some point. Um, but yeah, I would say thinking more like in mga, um, I think is a good mental footprint, um, for how to um, play in this orchestration game.

Speaker A: Yeah.

Speaker C: Take ownership.

Speaker A: Yeah, yeah, yeah. I think in, in, in, in in essence all three or all four of those things are important and the question is which one is the enabler and which one is the driver? Uh, because the one can't go without the other. So I think it would be interesting to see what drives the most value over the next five years. Um, we know that all these things would need to be in the pot for success. Um, you know, the question is which drives and which enables. Um, uh, to Ron's point about, you know, is a technology driving or is it enabling? Um, I guess doesn't really matter which one it is, but at the end it's a core component for, for success.

Speaker B: Absolutely. Great, gentlemen.

Speaker A: Okay, thank you very much. Well, thank you very much. I really enjoyed that conversation. It was a, the best use of my afternoon in a long time.

Speaker B: Likewise. Likewise. And um, for everyone listening, we went willfully off script. Um, we just. And this is the best conversations, um, you know, where you just follow um, on, on what your colleagues, um, have said. So thank you very much for both of you for kind of sharing your, um, your views. And hopefully this was something, um, that helps everyone who is listening or watching. Um, you know, it's, uh, a. It's an evolving space. So watch this space. Um, but, yeah, um, you know, think about orchestration and taking ownership to take, uh, Ron's final words and making them mine.

Speaker A: Yes, everybody, thank you very much.

Speaker B: Thank you very much.

Speaker A: Bye. Bye.