Defining a Data-Driven Product Strategy - Konrad Heimpel, VP Product & Data @Getsafe

Welcome everyone to another episode of the Product Bakery. Did you know that insurance can actually be simple? No. I didn't. Me neither. But after a long break, today we finally had a new guest again in the session. So it's not only me and Christian talking, but we also had Konrad Heimpel with us. And Konrad is the VP Product and Data for an insurance company, GetSafe. They do call themselves digital natives. So you can imagine it's mostly based on an app where you can manage all your insurances. But more interesting than that, Konrad actually talked a lot about how they set up their organization and the role that data plays in the organization. Deriving a product strategy from a good understanding of data was something that blew us away today. And I can highly recommend tuning in to learn how to form an organization, how to form teams based on a good data strategy, and also how to understand what your customers really want by not looking at the NPS and rather looking at the PMF. So with that said, lastly, as always, we love to hint you to our social media, follow us on LinkedIn, check out product-bakery.com slash episodes to comment on our episodes. And with that, I would say, let's get started. Hi Alex. Hi Konrad. It's nice to have a guest again in our show. Alex and I love talking to each other, but it's a pleasure to have you here today, Konrad. Hi there. So as always, we start with our lovely guests, and I'd like to introduce you quickly to our audience. So you actually have a very interesting background because you studied physics and math. I guess numbers and data are your strength. And you used to work for a couple of companies in consulting. You used to be a self-employed data scientist. And these days you are working as data and head of product for a company called GetSafe. So as far as I know, to not spoil it too much, you do stuff with insurances. And I remember I applied, I think four years ago for a company doing insurance comparisons. And I was like, okay, it's quite boring to me personally. So I'm just wondering how do you end up in the insurance space? I think insurance isn't like really the industry with the best reputation at least, but I think there's probably a good reason for you to be there and to work there. Yeah. What a warm welcome, right? Happy to be here. No, just kidding. Yes. Hello. Happy to be here. Happy to talk to you today. How did I end up in insurance? To be very honest, I would never have thought that I would end up in insurance. I did start my career in financial business as I was excited about it, wanted to have a look into it. And I found it to be quite boring. I like banks specifically, those old banks. And still I'm here. And I think this was a decision I took around two and a half years ago. I joined GetSafe around two years ago. I took the decision around two and a half years ago. And I struggled a bit with it. But in the end, and I think this is really something that's also part of GetSafe's mission, is that insurance itself, and I really believe in that as one of the greatest achievements of society, basically making sure that individual risks that you can't control the person are socialized. This is really great. It fulfills one of the basic needs of every human being. And I really would love to be part of bringing the whole area back to that core of where it has basically started. Yeah. And I think that sounds great. And I think that's also a little bit like what I see being the mission of a lot of tech companies. And you mentioned finance as someone who spent probably more than six years now working in finance. I see this all the time. The reputation there is also, okay, it's not really the most exciting fields to work in if you compare it to, I don't know, working for pitch making, redefining the way people work on presentations or something like that. But in fact, it's tools that people use every day. It's tools that people have to interact with. And if we have the opportunity to actually change this, improve this, make this better, it sounds like a great opportunity. Yes, this is totally true. And maybe you mentioned a very relevant point, products people interact with every day. I think that's even a bit different for insurance, because in the best case, you don't really interact with your insurance provider. Was just about to say that. Exactly. And this is one of the great challenges from a product perspective, by the way. But I think if people then interact with their insurance company, often something really awful happened to their lives. I just read a review of a lady that basically her house burned down, and she basically lost everything. And we were able to, that's actually one of the good cases, that we're able to help her out really quickly. And that is just amazing to read, because it makes a real difference in people's life. That's actually a very interesting side of how to make the negative moment where you have to reach out to the insurance company to something a little bit more pleasant or take the negativity out of the situation. Yes, totally. And I think we are already quite good in that. It's super simple with our app, we'll probably come to that in a second. We do have an app where you can basically manage all your insurances or your coverage that you have with GetSafe. And it's super simple to actually file a claim and then get the reimbursement. So we're not basically waiting for a fax coming in that took two days to be transferred. And in the end, that's a huge benefit for our customers. I'm just wondering, you just explained right now a good case, but I can also imagine that there are sometimes tougher cases. So what would you say are generally speaking, still with the revolution in mind, the biggest challenges you are facing right now in the insurance industry? Yes. So I personally think, and that is based on the experience I had in the last two years, this industry is incredibly slow. And I think you can see that because basically the big disruption didn't happen so far. Some players and the competitors or colleagues in the US are like just did their IPO, so they're becoming, I think, more relevant in the insurance space. Still insurance, it's such a big area. It's such a big business opportunity and no new player really took it so far. And I think this shows how incredibly hard it is to get started and get traction in that field. And I'm very happy to be in part of one company that is in good shape, which is GetSafe. But I think the challenge really is, and maybe to give a specific example, we are about to get our own license as an insurer. And this process alone took now more than two years to finish. And you alone need to have pure cash to be able to pay money to people to just sit there and wait for two years. No, I'm kidding. But I do the process for two years to actually get a license, which will allow us to basically control the whole value chain, which is a big step. What would you say, especially comparing also to traditional players in the industry, like how do you actually try to break those habits or the processes? Is having the own license and controlling the value chain the first step in that direction? Yes, totally. We have taken first steps successfully also in a lot of the past years. So we currently have around 250,000 customers in Germany and the UK who are quite happy, which we're very proud of. So we are already able to provide them a very good experience, overall user experience. And this is in the end what we believe in. We believe that it is possible to actually create transparent, fast, flexible, and simple solution or access to insurance. But we also believe that we can basically raise it to a whole new level if we control the whole value chain. And we've seen that in numerous examples. And having our own license, especially around that, what we call moment of truth. If someone happens to our customers, if they have a damage or a claim, especially in that moment, it's critical that we have the freedom to decide ourselves to opt in for a perfect user experience and not only for maximum margin, how others might have done that. That's nice. So you're saying that to some extent, obviously, like the perfect user experience also comes at a certain cost, which could be the cost of having lower margins. But that's very interesting. Yes. And I think this is very special. I think that's always the case, right? Good user experience costs money. That's pretty clear. But I think insurance is a bit special in the case that it might happen that you file a claim and you're not covered against what you thought you'd be covered. And then still, we might actually, because it's a bad experience if that happens, it's quite frustrating, invest quite a lot to be very transparent and simple in explanations upfront about what you've covered and whatnot. But still, that might happen that you get something rejected. And then maybe we can still help out the customer. But actually, that's money. That's more money than we would need to spend if we just pay out according to the terms that we have. Yeah, it sounds really interesting to me. And I'm actually super curious on how you do that, how you execute the scenarios that you just mentioned. Maybe you can tell us a little bit more on how you structured at GetSafe and what are your product values and principles that make you driving toward this big disruptive change that you want to bring into the market? Yes. At the core, there are a few principles. I think one of them for me is that we totally believe what I just stated, believe in superb user experience. So we believe that a better user experience in the end will make a difference and make customers actually choose us and not others. And I think another principle that we're following is that we really believe in, and that's also that fits to our target group, who are digital natives mainly, we really believe in having our app, the GetSafe app, as the central hub for everything coverage related and insurance related. And that's something that we're following quite strictly. And I think another principle that we believe in is that we want to scale and we think we can scale through technology. So everything we do is really, in the end, a piece of technology. And we already showed in the UK that it's possible to very quickly start in a completely new market and we will continue to do that. And we have a strong belief and invest a lot in it that it can all be made by building the right technology and automation. And that also gives an indication how we're organized within GetSafe. For the product organization, we're actually working very closely, are very closely connected to the engineering department, which I think is not surprising. But what we do is we are running product squads and we're covering different areas of our product, both customer facing, as well as more like partner facing or like sales partner or risk partner facing. So what's very important for us when we decided how to run the teams, and by the way, we'll come to that, it's something that we need to reflect on and iterate on over the time because we will grow a lot and we will need to adapt. And we decided that we basically have teams that cover a very fixed area of the product, which is basically buy, manage, and claim. People, basically a squad that is optimizing the buying experience, that is optimizing the manage experience, which is like managing your contract in the app and optimizing the claims experience. And then there are other, as I said, more technical squads. And for us, it's very important that we have this kind of fixed area of responsibility so that we can give the teams the possibility to create a longer term vision and idea of their respective field. And that is both true for the team which leads, like the product manager who leads the squad, the product designer, which should also create kind of a design idea and design vision of the respective field, and also the engineers in the field who should become experts in the technology that they're building basically. I think this actually brings me to a question or something that I was like recently also working on, which is a little bit around the growth and the scalability. And I think everyone who followed a little bit the news around Gatsby, you're obviously like also growing a lot. And so thinking about like also the scope of the team and fixing that and making sure that you have a fixed scope. At the same time, one thing that I've seen is that as a product grows and develops, you also have many features that constantly are added to a specific team. So if we think about the claims team, obviously, as you grow and you add more functionality, the team has a much wider responsibility. So how do you manage that growth? Would you then split the team or how do you approach that? Yes, in general, we would split this team. And I think, to be honest, we don't see it as like black and white. We're not 100% divided yet how to do that per team. But I think in general, you can split it horizontally or vertically. So you can either say, OK, now this team, like in claims, there are now two teams. One team is focusing on filing a claim and the other team is focusing on processing the claim, which is a long customer journey. You could also think about, hey, one team is still responsible for the overall customer facing part and the other team is responsible for building tools for agents that handle claims that make everything faster. That's, I think, really a decision you need to take when it's time to take it. So when you see that kind of the business or the user or something outgrows the current capacity of the team, then you actually need to make that decision or if you can foresee it. And that's, I think, also very important. It really also needs to fit the rest of the organization. So we always see that our squads also make sense in the sense of fit to, I don't know, operations teams, insurance team, marketing teams, you name it. I hope that a lot of companies who follow reorganizations by simply copying, let's use the Spotify model and then ending up like with too many teams with no real scope definition. I hope that those companies are taking notes now because I think that's a very important point. And it's also like a common theme that we're discussing. Splitting up the domain and managing growth is one big issue, but what goes in hand with that is also the whole definition of the roles within the teams. And I'm just curious because your title says data and head of product. So I'm just wondering how important is the data part and how are you letting it flow into your day-to-day business and also into the day-to-day business of the teams? Yes, happy to answer. So this is to clarify, I see as I started as GetSafe as head of data and then took over the product role. So I'm within data. I do have a lovely colleague, Susanna, who is head of data now and takes care of the team. Then I'm more part of strategic discussions. But still, I think that's also something I strongly believe in that product, but also in general, business decisions should be always very data-driven or informed, however you would call it. And I really, that's one of my core beliefs, I want to make sure that we also do that at GetSafe. So I think product and data should go hand in hand and both product manager and designers, but also engineers should understand the kind of impact and the metrics they're moving. And on the other hand, also data people should really understand or data people and engineers, basically what they're building, what value they're creating for the customer, for the business. I think it's quite a connected field and both sides can contribute from each other. Do you have also prod analysts in squads? Yeah, exactly. Partly, yes. To be honest, we're not a hundred percent strict there right now because the demand for an analyst can really vary over time. So it's really hard right now to put someone in there fixed. Also the topics can vary. Sometimes it's a very data engineering driven topic, just get me that kind of a data set and it takes the time. Sometimes it's more like exploration and you need a data scientist. So that's why we currently do not have this a hundred percent fixed basically relation between data people and squads. But yes, in the long run, I'd love to have that, definitely. And if you say, obviously data needs to be a central part, right? Me coming from design and working mostly with designers, do you have any advice of how to educate around data or how to, yeah, how to build in a data mindset also in organizations that maybe are not data driven or functions that by nature are not so data driven, for example? So like spontaneously, I think two things come to my mind. One thing is really just living this kind of data drivenness, which for me is always asking if there are proposals and ideas out there, hey, what's the data on that? So please bring some data. How many customers does this affect? Which metric does it affect? Why are we doing this? And then also if we're discussing launching a new feature, asking how good is it validated? How did you validate that actually? And is there a way maybe to run an A-B test? I'm a big fan, huge fan of A-B testing. And to be honest, it is always a hard decision being very honest whether to A-B test or not, because you might know that you're thinking of a feature and you attach a certain potential to it, what might happen if you launch it? And then someone tells you, yeah, please only launch it for, I don't know, 50% of people. You might think, well, but why not launch it for everyone right now? Because I really want to have that potential and want to make that thing better. But from my experience, it's always better to run an A-B test to get any kind of hint on what the real impact of something is. And just like coming back to the original question, just really living that and having people always remind everyone of that, I think is one thing. And the other thing I think is telling stories about what great things you can do with data, just success. Where basically you looked at data, you found something and you did something great for either customers or the business. I think that helps people to understand why it's so important and basically also helps removing maybe the barriers in people's mind that it's so complicated to look at data because it's actually not, I think. Can you share a success story with us? Yes, I can. I need to be, yes, I can. So I think one of the biggest success stories I saw at Gatsafe is when we say analyzed prices of competitors and we did a basically educated guess how we could improve our prices to be more competitive. And this was a pure data science thing. And we could basically overnight, we could increase our sales that we had with certain channel by a factor of, I don't know, eight to 10. So that was really amazing because it really like this on and off switch. That's impressive. Yeah, I really love that. I wish we could reproduce that more. Every team, please reproduce the success. I wish it would be like a much nicer story about some kind of user experience hack that we found, but actually the business. But looking at the business, and I think it's also a nice bridge that we can just make to the whole topic of strategy because you also mentioned for the data team, you are there as like the strategic consultant, but generally speaking in your position as the product lead, as the head of product, I guess the strategy part is just super important for the business. And I mean, as we know, the health insurance market maybe is not the fastest and innovation is maybe not as fast as it might be in FinTech, for example. But still, I guess having a good product strategy in place is very important. And I'm just wondering, how do you approach the topic of being data driven plus defining a good strategy? Yes, happy answer. I need to clarify one thing. We're not selling health insurance yet. I'm always switching back. Yeah. I'm going to cut that out. Okay. Okay, good. Sorry. No problem. So yes, happy to go into that. Basically what I'm not sleeping well if I base my business in the end and as we're a product driven company, product equals business. I'm not sleeping well if I don't base my strategy and my bets on data. And that's why I really tried and tried to come up with a very data driven product strategy that in the end minimizes the risk of taking the wrong bets, taking wrong invests. And what we did is we used some, I think best practices and combined them. And so what we did in first place is we started to collect the PMF survey, send out the PMF survey and started to measure the PMF score. Everyone who doesn't know about it should go and read about it. There are tons of blog posts and the original blog post about it. It's really cool. And so the key idea basically that we also use is just asking people how happy they are with GetSafe and then putting them into segments and then asking them like, why are you happy? Why are you not happy? And even more in particular, asking them like, how much do you like the speed of answers? For example, how much do you like that? We have an app, stuff like that. So really testing our use piece with them. And then basically what you can do is you can start and go and collect data and do analysis on what the impact of different fields of the product and different use piece that you have on actually the happiness of the different segments is, right? And what we also did is we added another question, which is, hey, how likely is it that you're going to buy your next insurance with us? Which for us is like the most important question about loyalty that we can ask, which is the underlying assumption of our business model. And then we could basically directly connect answers and feedback from the customer of the different segments to in the end, this kind of business outcome that we want to, that we wanted to have, right? The business or product outcome, basically that people are loyal and start to buy not only one insurance, but multiple insurance. And this gave us a very good direction of what the important topics for our customers are, where we need to improve to move towards that direction of actually following or making our business model come true. And that was the first step basically, right? And that led us to certain, I think there was nothing surprising rather that customers really telling us that we, what they really care about is for example, that moment of truth, what I talked about. So they really care about that. Even early up front, they really base also their first decision about whether they have the impression that we really take good care of them if something happens, right? Even more than the price, for example. Yeah. Yeah. And I think using the PMF survey is so much better than looking at the NPS because it tells us so much more and helps us so much more to understand which area of your product, which feature is performing or not, instead of just becoming a feature factory and building things and using data that actually tell you nothing, which is more event metric than something else. Yes. We also really liked it. It was really telling. And what we then did, this was also the basis to form our squads, to also basically set the product areas that we wanted to cover. And then what we as well did, so let's take again the example of claims, because I think it's such a telling example. What we then did and still also doing, because I think it's a never ending job, is we were trying to find out also from asking customers, analyzing feedback, what in the end are the key metrics, the leading indicators, the leading metrics that we need to move to really make customers happy. And it's really important, as you said, NPS is super lagging. Just asking people around, hey, how much did you like the claim process? Yes, it's important, but only if you then ask why. And if you really go down and analyze it every single day, only then you get a hint on, hey, what's really important. And also, it's no surprise, but it's good to have the validation and it's good to quantify it, is speed is quite important to people, how fast we are. And then next question is, what is fast to customers? What does it mean? Is it like getting it done in one day, in 10 days, in 30 days? So you're going to ask customers about that. And then what we did and are still doing as a next step, at the same time, while really professionalizing this kind of customer feedback analysis, we also started doing a lifetime value predictions. So trying to predict what kind of customer segments are going to stick for what amount of time, what's their bundling or adding other products activity, what's their referral activities, and so on. And then we are into basically combining those, what we have hypothesis for leading indicators with lifetime value predictions. And that is really amazing because we can specifically show already for some metrics that actually people having, for example, speedy processes, they have a higher lifetime value than people seeing like a slower process. And we can really quantify it in the sense, okay, how much more is the lifetime value, which directly also gives us as a business, again, an indication how much should we invest actually in being faster in that, right? And to be honest, we are not there yet that we did that for all of our business, but definitely it's a goal of mine to do so because in the end, going back to what I initially said, it let me sleep well at night because I know that actually I have proof that where we put our money creates value for the customer and creates value for the business. I'm wondering with such a strong focus on data, especially like on the product development, is this something that also translates to the whole organization? Are all teams that focused on data or is it mostly like something that influences the way you develop your product, but not so much maybe the way marketing launched their campaigns? So I think it is, I think GetSafe in general and all the teams are very data driven. I think it can always be more, but I also, I also believe that you won't in the end win the game, especially if you're trying to disrupt an industry with only relying on existing data. Also need to put in kind of your own core beliefs about the product and the business and you need to wrangle it, right? So I'm a fan of saying, Hey, there are decisions that we need to take where we can basically open a Google sheet and get any data and basically any result we want. That's something that I learned as a consultant. I can turn you any business case into a positive one by just looking at the data long enough and like adjusting some assumptions. It's possible and you shouldn't be doing that. So you should use data where you like being honest to yourself, use data where it makes sense. But for other decisions, you also sometimes need to follow your gut feeling and you need to also make like entrepreneurial decisions that no one will tell you, you just need to. How do you balance that? What is an entrepreneurial decision versus what is like a decision where we heavily rely on data? I think that's actually, that's a very good question for me, like personally, but this is very subjective. I always try to go and say, Hey, like how, what kind of validation do I have and what kind of validation can I get about that? And if I can get validation quite quickly, then I'd always do it, right? Because it's basically for free, right? It's fast. But if I know, Hey, this, getting this kind of validation is, takes really long and the large and vast, then I know it's a decision where it may be a need to follow or just take a decision, right? Rather quickly. And I'm also a fan of, of always separating decisions into that from Jeff Bezos, like those one door, two door decisions, right? You should be more careful about those one door decisions where there's no way back, be more careful with it. But actually if you're, if it's a two door decision, then you can be a bit faster in taking it, right? That's a very personal view, but for GetSafe, I think for me, key is still just talking about it and being transparent. And I think it's part of the culture also being transparent and, Hey, there is a big decision and any of the teams that's up front of me. And to be honest, I don't know which route to take. And I think talking about that very openly and it's actually not a weakness, but a strength because then you can use others to reflect on that. Maybe someone else has experience, has data on it. However, I think it's a lot about openly talking and transparently talking about those kind of decisions within the team. Everything you explained, I'm just trying to digest the information because I'm just asking myself, where does it start? Because what you explained to us is actually the way it should work, right? And the fact that you check the product market fit, which is for every company, the most important thing you can do and start then understanding where you can optimize, where your, I say cash cows lies or your top performance features are, which then impacts also the organizational structure by saying, okay, we form teams around our products, which are performing or that we want to perform in the future. It's like a huge step. And to what you've said just previously, I think all those decision making processes and the transparency and the culture derive automatically from a good, healthy understanding of data and also the motivation to work with those. And unfortunately, including myself, once you get thrown into a new team in a running company, you don't have that mindset always, and it's hard to establish it. However, I believe whether you are a product manager or a head of product or a CPO, no matter in which position you are on the ladder, you can always have an impact by making the change towards focusing on data. And sometimes it's just important to start understanding is product market fit given or not. Yeah, I really liked your approach. And I'm actually surprised to hear this amazing process because many times we talk about it in a more abstract way. So I'm very grateful for the nice elaboration today. That's very cool. But maybe to also make that to be realistic, we're not there yet across the board. So there are also topics and squads where I'm like being very honest, I'm not sure yet what the specific metrics are that we need to move. Because sometimes there are also competing metrics between users and business and stuff, efficiency versus user experience and so on. To be honest, in some areas, not 100% sure yet. But I think it's really important to keep on reflecting and iterating. That's like, the way is the goal, right? If you keep doing that, you're minimizing again, risk that you're doing it wrong. But you know what, if we're going to talk in two years again, you will say the same thing. Yes. And I think it's also the beauty of our jobs, right? If you could answer everything now with data, automate it. I think it also comes down to intuition, it comes down to a certain vision, taking the right decisions, having smart people around. I think the cultural aspect of reflecting, talking about things, learning from others is as important as actually like focusing also on the data, right? Yes. And that's a super good point because actually, if you watch my calendar, what I'm doing the whole day, most of the time I'm not spending on strategy and data and metrics, but actually making sure that teams are running smoothly in communicating, optimizing communication. Because I think that's one of the biggest challenges when growing is making sure you keep communication efficiently, right? That you adapt communication, increase communication if you grow, but not to an extent where you get slow like a corporate. So I think that's actually one of the other next to knowing what to do and getting that right, how you do it, running the teams is really challenging, but also exciting. And if we speak about communication and spending a lot of time in meetings, talking to people, I know it's probably been a long day for you. And after multiple meetings, we're happy to also release you into your well-deserved firearms, the way we would say here in Germany. So if there are no final questions, especially looking in Christian's direction. I'm satisfied. Beautiful. Then Conrad, it was really great talking to you. Thanks for being here. A lot of great things to actually take with us and apply into our day-to-day jobs as well. Thanks a lot for having me hosting this. That was really nice. Everyone come check out GetSafe and find GetSafe. We will put the link in the description and also the original link to the PMF survey. So whoever's interested can also reach out to you via LinkedIn and all the other social media channels these days, whatever your kids use. And just as a side note, I'm not advertising anything here, but as GetSafe is growing, I'm sure they have a lot of like product or data roles open. So if you're interested in working with Conrad and with Conrad's team, I'm sure you find something there as well. Always hiring, reach out. Beautiful. Then have a beautiful evening. You too. Bye-bye. Bye-bye.