October 10, 2025 | 14:41
Karlyn: Thank you so much. Happy to be here.
Dani: Happy to have you and very excited to dig in. I wanted to ask you first you know from your perspective at Celent what are you seeing broadly in terms of trends and opportunities for insurers that maybe are gonna you know be big in the next 12 to 18 months or things that insurers perhaps should be paying more attention to that they aren't.
Karlyn: Oh gosh there's so much going on out there I mean if you have walked the exhibit hall here out of Insurtech Insights, there's one thing that just continues to smack you in the face, which is how dominant the issues or the opportunities for AI and Gen AI and agenting AI are. And these are terms that I think aren't always clear to everybody what the differences are and how they can be used and what the use cases are. But these new technologies, I think, are fundamentally changing what we can do and what we should do. And we said that Gen AI is like a Gutenberg moment, because the Gutenberg printing class redefined how human beings interact with a printed word. And Gen AI completely redefines how human beings can interact with data. Now, Gen AI isn’t a very good at making calculations, but AI and machine learning is, and when you combine these things, when you combine the Gen AI and the heavy duty processing, you create some incredible opportunities. Now you throw the agents on top of that and being able to break tasks into tiny little pieces and assemble them quickly. And I think that this has a fundamental transformative moment for the insurance industry. Now I don't think we're going to see all of the impact of this immediately. Like we're still trying to figure out what the heck it all is.
Dani: Absolutely. I was going to say to your point, some folks don't necessarily even know the difference in terms of maybe the best places or touch points in which to integrate that technology and I'm also hearing here there's a potential outcome for greater efficiency likely profitability if you're doing better with your resources things like that so I want to ask you and grabs also where tech fits into this conversation insurance is a for -profit industry where do you see money being left on the table?
Karlyn: Well it's definitely in that world of efficiency there is still so much processing that is being done in human beings that are entering data or manipulating documents that technology could clearly impact that level of operational efficiency. The legacy systems that are being used, we go to insurance carriers offices and we see people with three monitors sitting in front of them because there's so many places that they have to go. When you think about the cost of that not just in terms of the human beings, but in terms of the time that it takes and that ability to to create the insights from all of those different aspects. And so I would say one of the huge areas where money is left on the table is operational efficiency. But I also think we're leaving a ton of money by not using the data that we have. By not being able to pull the data easily by not being able to connect it, by not knowing how to apply it effectively. And so, you know, how do you solve these things? I think it starts with having a bigger vision of what you're trying to accomplish, right? If you go back to the fundamentals, insurance is about assessing risk and transferring it to somebody who's willing to take on that risk, right? It's not about having the screen where you enter data to enter the next screen to enter data to enter the next screen to enter data. And if we begin to think fundamentally about what insurance is, that creates tremendous opportunities for using data more effectively to assess the risk to determine what the price of taking that risk is and finding different sources of capital to do that. Now I will say there's one additional aspect to operational efficiency that we sometimes don't think about, which is picking which accounts to work on. If I have a close ratio of 20%, that means 80% of what I'm touching, I'm not writing. If I could simply pick the ones that I'm most likely to write and write those ones first, I'm going to have a huge impact on my efficiency. That 80% of effort is not wasted. So I think that there's a variety of things that are being left on the table, but it all revolves around these manual processes and the ability to access data.
Dani: Fabulous points, and there are a couple things I wanna ask you about there. I would like to go back to consumer trust, but also talking about here, kind of picking which accounts to write or not. A question would be, what if insurers could more confidently trust the information that they were getting from consumers? How would that impact their process, their people, their business?
Karlyn: It's something I hear from carriers all the time. Insurance companies have moved to buying a lot of third -party data, and I think they need to. I was interviewing a CSR once for a project, so we were talking to one of the folks at the insurance agency, and we were talking about the process, and she said, well, I just don't know why the insurance companies asked me such stupid questions. Like, what is the construction of the building? Like, how am I supposed to do that? My initial thought was girl you don't know your job and then I realized wait a minute how why do we expect that they know that you cannot tell the different string construction class 6 and construction class 5 by looking at it and we can buy that data and it will be more accurate and so the the ability to use third -party data I think has increased our ability to trust to the information but even there we don't have high accuracy rates And so when we talk to insurance companies about third -party data, they're constantly trading off, do I buy it up front when I'm quoting, it costs me money, and if I don't write it, that's very expensive, do I buy it after the fact and then potentially have to reform the policy? And what if I don't believe that data in the first place? How do I think about the accuracy of that data? And then if they get it wrong and have to reform the price, well, that damages the trust that both the agent and the consumer have. You told me it would be, you know, $12,000 and now you're saying it's 14 and that that very moment of erosion of that trust because the data coming in wasn't accurate is a huge issue.
Dani: Absolutely. I'm hearing here are very real tradeoff when it comes to the data and I mean from a Clearsped perspective we often talk about the tradeoff between speed and security like you can either move really fast but a lot of fraud gets there You can move super slow and put everyone through all this stuff, but then how long is it taking to pay things out? So there are a lot of trade-offs here. It sounds like even from a data perspective, to your point, buying third-party data, you also run the risk of potential bias built on historical models, like incomplete information. So there's all of that too. So we're kind of in this situation where something has to change or be different or shared, perhaps, or other technologies that can help us along.
Karlyn: And if you try and buy it, and then they don't actually have it, so you can't even have the fill rates that you need. Well, now you have to come up with other techniques to process out of business, to assess the risk, to determine if you're willing to take that risk on. And now you've introduced bigger sources of a potential error, so it becomes complicated.
Dani: Yes, let's go to the consumer trust and side for a second. I mean, consumer trust has taken a big hit. Whether or not it is the full story, we see the headlines for example, the echo chamber of social media, there is a perception problem I think we can say. Where do you think that the industry has fallen off and what can the industry do to rebuild?
Karlyn: Trust is such an emotional issue. We tend to think of it as transactions, but trust comes in every interaction that a company and an agent has with that policy holder. Whether it is, you can trust that the premium I'm offering you will stay, whether it's the denied claim or a complicated renewal or some other life altering event. Every one of those moments have the ability to create an emotional connection. We run a risk of over-autorating in some cases. Now that said the claims area we've just done a study on this which has not yet been published. So hot data, you heard it here first. Yeah we surveyed 324 folks who had automobile claims and we chose auto because they're very homogenous. So we're looking at that claims process not the claim itself right and we found that for those and we looked at them for catastrophe claims is non-cat claims. And we found that 39% of the non-cat claims, the ordinary auto insurance claimants, loved their insurance company so much after the claim that 39% of them said, I love them. I'm willing to stay even if it costs me more. On the catastrophe side, it was even more. 44% said, I will stay with this company even if it costs me more. And so that claim moment has tremendous opportunity to build that loyalty and that emotional connection. Now, we do that a lot with human beings. And there were some other things that were a little bit surprising, which is that the claimers also thought there was extremely high levels of fraud. And I don't remember the exact number. It was like 25% felt that there was fraud in their claim case by the medical providers or the body shops or the other party. And there was a large proportion that also thought they were being overpaid. But it might be that the lifetime value of a customer is worth overpaying a claim now then. I mean, I don't know, that's not something for the actuaries to calculate. But I think building trust requires insurance companies to do a couple of things. First of all, I think you have to be very transparent. You have to make sure that it's very clear what you're covering and what you're not covering so that it's not a surprise when you claim isn't covered all the way. I think you have to have great service with that emotional human touch. And it doesn't mean that it has to be a human. You can simulate human behavior using technology. You have to put humans in the middle. But it has to be consistent. It has to be empathetic. You have to show up for the customer. And I think those kinds of things are the sorts of things that then begin to build trust.
Dani: I think that's wonderful and also very aligned with some of the feedback we've heard from senior insurance leaders today, knowing that there's an educational component, there is the empathy aspect, like all of that is very critical to building that trust. And to your point, you know, these moments in time, these moments of truth happen typically when a person is vulnerable. So, you know, the most you can do for them is be there for them and kind of create that safe gap.
Karlyn: Know who I am and know that I'm not trying to cheat you, help me protect myself help me recover after the loss and let me say that is not necessarily a check right I mean I think about when the town of Paradise burned down because my sister lost her home in her business and let me just say everybody got checks right away they have not recovered they were still living in tents they were still living in in RVs and the insurance carers were done they'd walked away and so there is more that we can do to help that recovery. I think people want to, you know, talk to me the way I want to be talked to. Let me talk to a human if I want to and don't make me if I don't have to. Be fast and easy and of course they also all want it to be, you know, make it cheap, right? But there's a wide variety of things that we can and should do to drive that customer experience to be more, more focused on what they're actually trying to accomplish.
Dani: Absolutely, and to your point, sure, make it fast, make it accessible, make it cheap, but based on the study and the numbers that you just shared, you don't have to make it cheap if the other two are kind of hitting the mark. So to close this off, Karlyn, if there's one thing you think insurance leaders who are listening should know or should be thinking about or should be doing right now, what would that be?
Karlyn: I think I would say be bold, pragmatic about change, be human about how you implement change. This industry, there are strategy documents out the Wazoo and let me just say I kind of feel like strategy doesn't really matter that much because if strategy mattered everybody would have the same one. It's about how you execute on that strategy. And so you need leadership that can create this decisive vision of who we are and rally your team around who we are and then put all of the resources around delivering on that policy. Whether it's the the way that you handle your technology, the way that you handle the customers, the way that you handle your intern employees and how you build talent, you can't lose sight but if you because if your people are brought along for the ride It doesn't matter what you have on the whiteboard, right? People are the ones who are going to be delivering what you're doing so I would say pick an area where your organization has always said this is how we do it. We always do it this way and then challenge that. Look at some small aspect of it and think through what if what if that were not true, what if I could change that smart start small, learn fast, abandon quickly, or scale up but I would say that if you look out here again at the amount of change that is happening in this industry I've never seen change as much as what we're seeing today and it is fueled by data it's fueled by AI it's fueled by this incredible level of level of competition and that's I guess that would be my advice: be bold and pack up.
Dani: Wonderful. Be bold, pragmatic and also folks take a look internally, look at the thing that's been the thing that you've been doing and think about it differently. Thank you so much Karlyn for being here.