“Ask once remember always” – Build a consistent personalized customer experience
Listen to the Insurbreak podcast on:
Brian Poppe worked for Lincoln Financial Group for a period of time before joining Mutual of Omaha 13 years ago. From pricing and product development at Lincoln Insurance, he worked his way up in risk management at Mutual of Ohama. Brian also created and led the innovation practice for a couple of years, led one of the PNLs, worked on tech modernization, a bridge between business and IT strategy. He is now Chief Data officer – we can say he’s exempt in the data world!
In this podcast, learn about:
- Friction between data privacy and personalization. How do you protect customer privacy and what are the legal implications? How do you make sure you’re not intentionally/unintentionally biasing some of the models that you’re using to either recommend or automate underwriting?
- Personalization of the customer experience. How do you help your customers pick the right policies? How do you earn their trust to be able to provide them with the right recommendations?
- Whether human interactions will continue to be crucial for consumers’ decision-making process and how AI is impacting insurance and potential future applications
- His Pilot AI with chatbots, and the experience of an interactive agent call center
- The future of insurance and the implications of embedded insurance
- How inflation is impacting insurance businesses and consumers
- Diversity equity and inclusion in the insurance space
Ash: Where do you see the future of insurance?
Brian: Yeah, one of the trends that I’ve noticed the last couple of years at InsureTech Connect has been a concept called embedded insurance.
Ash: You’re tuning in to the Insure Break podcast. It’s the podcast about the latest and greatest trends in insurance. I’m your host Ash, and I invite you to join us as we interview experts and executives at Insurance covering innovative practices, technology advancements, and insight into the future of insurance.
This podcast is sponsored by Zelros. Zelros is an AI software solution for insurance to hyper personalize the customer buying experience with insurance recommendations across all channels, boosting client acquisitions, cross sell and upsell.
In this episode, we chat with Brian Poppe, chief data officer at the Mutual of Omaha. With over 15 years of experience in insurance, Brian draws from his experiences in various roles, including the Lincoln Financial Group and the Mutual of Omaha to answer questions about the friction between data privacy and personalization, how AI is impacting insurance and potential future applications, and how inflation is impacting insurance businesses and consumers. Make sure to stick around until the end to hear about Brian’s predictions about the future of insurance.
We are joined today by Brian Poppe. He is the Chief Data Officer of the Mutual of Omaha. Brian, thank you for being with us today.
Brian: Yeah, happy to be here Ash.
Ash: So to begin, can you give us some insight into your background and your journey? I think you grew up in Omaha, Nebraska, is that right?
Brian: Yeah, I grew up in a small town in central Nebraska called Hastings, about 25,000 people. If you are a fan of Kool Aid, that is where Kool Aid was invented, Hastings, Nebraska. So, fun fact for you. At the end of every summer, they’ve got a Kool Aid festival. So you get to see the Kool Aid Man and drink unlimited amounts if you buy one of the special mugs. So I’m an actuary by background. I’ve been around insurance since I graduated. I worked for Lincoln Financial Group for a period of time and then came over to Mutual of Omaha a little more than 13 years ago. And I’ve done everything from pricing and product development on long term care products to work on some group products at Lincoln.
I worked in risk management, created and led the innovation practice for a couple of years at Mutual of Omaha, led one of our PNLs, worked on tech modernization for a couple of years, kind of as the bridge between business strategy and IT strategy, and then have been Chief Data officer for about three or four months now. So you can pretty much call me an expert in the data world.
Ash: Very cool. So what are your responsibilities now?
Brian: Yeah, my job is to make our analysts smarter. So we are in the midst of moving from an on-prem analytical environment to a cloud based analytical environment. So we get to keep some of that data that we might have consolidated in the past because we just didn’t have enough compute or storage power. We’re going to keep that as granular as long as possible and make it really easy for our analysts to both find and uncover insights that can help drive our business forward. So in addition to that, I’m working to better enable machine learning. So I’ve got a team of data scientists that I work with pretty regularly and help keep them pointed at business problems.
Again, I’ve had the good fortune of being all over the place in Mutual of Omaha, so I know a fair number of problems that’s like, man, it feels like we should be able to solve this one. If only we had a way and data to be able to look at this particular problem in this particular way, we think we would be able to get there. So helping the data scientists give them the type of environment that allows them to be successful as well, and of course doing that all while making sure we maintain our customers privacy, meet all of the regulatory restrictions and privacy regulations and so on, that we absolutely need to do from a company standpoint.
Got it, yeah. And kind of on the topic of customer privacy, I know that you want to personalize the insurance experience for customers, but also you have to make sure that you’re not crossing lines. So when it comes to customer experience or personalizing insurance experience, what does this mean to you?
Brian: When I led one of the PNL centers, I led the final expense business unit. We call it departing well. One of the rules that we try to enable I don’t think we got all the way there, but we try to enable is like, “ask once, remember always”. So if you ask the customer a question, you should remember their answer. So instead of asking them again for address right. So, hey, give me your address, and then later on, if they buy another product from us or we’ve got another form that we need them to fill out, pre-populate that address for them. You’re not trying to be too creepy about the personalization, but you are trying to make it easy for the customer.
Right. That’s a personalized thing of like, oh, they do know me. They remembered my address from when I gave it to them when I applied for insurance. And now I’m onto this other product and I’ve already got that type of thing, just small little things like that. And then once you’ve kind of learned and earned their trust, then it can be a little bit more personal.
Oh, by the way, you have this type of product, people in your situation also buy this other product. And whether that happens automatically via recommendation engine or whether that happens from their local sales rep agent, whatever, like, that is exactly the type of thing that we were trying to do from a personalization standpoint.
Ash: Yeah, I just remember literally like a couple of weeks ago I was adding a card on my insurance and I had to refill out like all my information again. So that would have been that would have saved a lot of time for sure. Okay, so personalization technologies, smart recommendation tools, decision support engines, these have seen a growth of over 200% year over year. Why do you think there’s so much attention on such initiatives?
Brian: So one of the things that we and I suppose many others in the insurance industry have learned is, as much as you want to give an automated recommendation to a customer because it’s such a personal and monetary decision or financial decision, customers want confirmation that they’re making the right choice. So we can put a recommendation engine out there that helps customers pick, for example, am I getting Medicare Advantage or Medicare supplement? And if I’m getting a Medicare supplement plan, which plan is the right one for me? Customers will still want to talk to a person to just sort of verify that they’re on the right path.
Same thing in the final expense business. They’re like, hey, I need to buy a policy to cover my funeral costs so my family doesn’t have to after I pass away. I still many times will want to talk to a human just to get confirmation that I’ve got enough coverage and that my family is going to be taken care of and any other pieces of information that they want to share with us, we’re happy to take that too. The insurance companies have done a better job of understanding maybe the customer’s psychology than we have in the past of why are they buying these things and how can I reaffirm that they’re getting the right coverage for them? I think the insurance industry has done a pretty decent job of product-market fit. They just don’t know maybe the psychology behind why that product fits that particular market. We’re getting better at that and that is leading to the types of personalization and in doing so you start to see a higher close rate if you’re selling things or a higher persistency or stickiness of the customer once you kind of affirm that they’ve made the right choice.
Ash: Do you think that ever in the future there will be a time where there is no need for human interaction for those decisions or there will always be a need for human interaction for those decisions?
Brian: I think for some products. Certainly we don’t see that for everybody. It’s not every customer that buys from us. We actually have a human talk to them. So there are some products for sure. The insurance industry has done a pretty good job of making things somewhat complex and as we unwind that and make things a little bit simpler so you can do more of the DTC or the personalization or make products better understandable, I think you’ll need less and less of it. But there’s still things like if I’m buying an annuity, that’s probably a five or six figure financial decision. And generally I would want to talk to somebody about making that type of financial decision. Even if they’re trying to sell me on it and I know they’re trying to sell me on it, I would still want to talk to somebody to get confirmation that we’re doing the right thing. So there’s probably things like that that I think are going to require some sort of human interaction for a while. Like if you’re making, again, six figure financial decisions, in general, people don’t do that completely blind. They want to talk to somebody about something like that. Even some of the insurance products have tax implications, like at a personal financial level. And so if you’re trying to have some sort of tax play in conjunction with the life insurance policy you just bought, that’s certainly something that you would probably want to confirm that, hey, does this work the way that I think that it does?
On the other hand, you’ve got folks like in the auto and home industry, home insurance industry. Those are relatively straightforward. I know I’ve bought plenty of those products in the past and not talk to a person and not feel like I needed to. So, as either the market gets more comfortable with those types of products and the insurance industry gets better at simplifying the actual products themselves, I think you’ll need less and less of that human interaction. But there are some things that even when I gave the example of Medicare earlier, when you turn 65, there’s a lot of things that happen, including potentially retirement.
And so as part of that, you’ve generally got some questions because you’ve gone from either accumulating assets through your working years, through having probably insurance provided by your employer. So now all of a sudden it’s a switch to a government sponsored program and whatever else you want to do from a supplement or advantage type product. It’s confusing and you haven’t had to go through it before. And there’s a lot of material, but it’s not exactly always clear on how to use that material.
Ash: Yeah, life insurance is the big one I’m thinking of. Like, if I’m getting life insurance, I want to talk to somebody for sure.
Brian: Yeah. Regardless of the premium that you’re paying, which may not be a significant amount of month, like done right. Again, that’s a six figure, seven figure, depending on your situation, maybe an eight figure payout. Like that matters, right? You want to make sure you get that, right?
Ash: How have data analytics and AI played a role in this shift to the consumer experience?
Brian: One of the things that we have implemented at Mutual of Omaha, and again, I don’t think we’re unique in this is trying to do a better job of taking what we know about the customer and then feeding them things that they are interested in. So for example, somebody hits our website from either a Google search or some article that we posted, right, well now we know something about maybe they clicked on a we’ll use Medicare again. Like this is maybe they clicked on a Medicare article. Well now I know something about them. They’re probably not a 35 year old working mother or something. They are probably somebody who is nearing retirement age and I can start to tailor their experience for them.
So certainly that’s where your data and analytics come in from that standpoint where I’m beginning to personalize based on the little bit that I know about that person, even if I know nothing else, I know they clicked on this link and they are interested in this particular thing from folks that are customers we’ve implemented like an onboarding recommendation engine. So again, based on things that we know about them, if they’ve named a beneficiary on their life insurance policy we won’t bother them and ask them to name another beneficiary.
If they’ve already set up auto pay, we won’t bother them to set up auto pay. So those are the types of things that I guess came out of some teasing of the data of like well, one, why are we doing this? But then two, we’re not really seeing the type of adoption that we want or we are seeing the type of adoption that we want. Let’s make a different decision from that.
So AI is an interesting one. We’ve run a couple of pilots with chatbots to varying levels of success. I would love to know if any of the viewers out there, you can contact me, you can find me on LinkedIn. If you’ve had success from a chatbot, if you’re a carrier out there and you’ve implemented a chat bot and it’s been wildly successful, I would love to know how that worked out for you and the types of lift that you’ve seen and whatever you were trying to measure. Again, we’ve had both ups and downs depending on where we place that. The closest thing probably that we put in place is chatbot experience and we try to make that apparent to our customers as well as on the other side.
On the other side, we do have it’s like an interactive agent in our call centre where you call in and it’s asking you questions, trying to get you to the person who can help you the fastest. So if you need to pay a bill, like we try to ask you questions about that and you respond verbally to get you to the right person, that’s probably another, I suppose, a good example of AI that we’ve got in place.
One of the things that I think the insurance industry had not done well in the past is, they were product focused. They created the customer experience solely based on what the company needed to execute. So shoot, we need more information about your health, don’t really care what that means to the customer. We’re going to send somebody out to your house. We’re going to ask for physician statements. We’re going to get a full set of medical records, whatever. And you need the insurance product so tough, you got to deal with it as a customer. Over the last probably five or ten years, you’ve seen the insurance companies switch to be, well, there’s probably a better way that we can do that. We don’t have to do that for every single customer. We don’t have to send somebody out to their house to literally poke and prod the customer to be able to make sure that we can insure them as a carrier. What if we were to do it this way? Seeking that better customer experience, so it’s like having empathy for the customer, thinking about things from their side and then ultimately building that experience around that of like, okay, I get it, you want the product. We’re not going to make a bunch of friction simply because we need it.
We’re going to try and make that much easier for you as the customer to get to your end goal faster. We know you’re not waking up every morning excited to buy insurance. You’re doing it because you need it for whatever family coverage to protect either yourself or your family. And we want to make that as easy as possible for you. So the data analytics and AI are all in service from what I’ve seen to help improve that customer experience.
Ash: Where do you see, maybe an application of AI that could be like, a potential solution or that could be an application that’s maybe not out there yet?
Brian: This is a very good question. Maybe one of the interesting ideas would be, could you maybe like, better personalize/tailor, whatever you want to call it, a coverage for a customer? So an example might be, if I’ve got homeowners insurance and for the last five years, ten years or whatever, the real estate market has been going up. If I haven’t moved, I’ve got a particular coverage amount on my house. Could I have a moving coverage amount? You could do it for potentially life insurance too. Like hey, as my salary increases, could I ratchet up life insurance without having to buy a whole new policy? I am earning more money. Could I get more coverage just because the first time I bought it, I bought it as a way to help cover a particular situation, right? I haven’t seen that happen yet, but that might be an interesting one to consider.
Ash: Okay, on to legal. So how do you ensure the protection of consumer data privacy while serving their needs for a more personal touch?
Brian: This is one that we are actively struggling with, in my role as Chief Data Officer. We have clamped down based on source systems. So you can imagine we’re a hundred year old insurance company. We’ve got a whole bunch of technology systems that back policies that go back probably 50 or 60 years. There’s some of them. And we said well this, this IT system has some sort of private information in it. So nobody gets access to it unless you are administering one policy at a time.
You can imagine that drives the analyst nuts. Certainly that drives the data scientist nuts. So one of the things that we are doing is to say well some of that data is useful but I, as an analyst or I as a data scientist, I actually don’t need to want to see a customer’s Social Security number for example. But I do want to know one, I need to know this person is also this other person in this other system because I can tie that information together. And now I may know something that if I want to make a recommendation to them but I don’t actually want to know, as the person who’s managing the process, I don’t want to know anything about the individual. I just want to know, here’s how the customer flow might happen. So the people that do need to know are the frontline associates. So customer calls in like obviously they need to see actual individual customer level detail. Most of the other folks do not.
So the thing that we’ve tried to do is say if you have an individual operational need to look at this data we are going to protect that in the way that we have. You will have login, we’ll have logs of who accessed it and what they did with that information and so on. From an analytical standpoint we are anonymizing that data and then we’re trying to say well I can group that in a way that I can’t uncover who that person is but I can derive some insights from that, that is going to help tailor not only that customer’s experience but other customers experiences down the line. So it’s like you think about it at an individual level basis and I’m limiting very tightly who can see that. And I’m trying to anonymize the data either through consolidating it in a way or tokenizing it or masking that Social Security number and saying like hey, there is a Social Security number here. No, you can’t see it because of the privacy rules that you asked the question about but know that it’s there, it will still function. You can connect that to other Social Security numbers and in doing so this is the type of system that you’re building. If it is maybe a recommendation engine for example. That’s the real trick that we’re uncovering. And thankfully the technology industry has greatly advanced the ability to do that in a way that protects both customer privacy while still allowing you to do the types of personalization that customers are coming to expect from other industries.
In the past it might have been like well the best I can do is just group these so you’ll never be able to match the individual together. Thankfully, now we’re at a place where I can do that at scale and bring them in a way that I can’t unwind down the line.
Ash: Yeah, I’m just curious, from your perspective, what’s the speed of change for all these legal side? Do you feel like it’s changing at a fast rate and every few months you have to make sure you’re up to date with the privacy laws or how do you feel the speed of changes in that?
Brian: The legal world probably changes a little slower than the technology realm, but I will give the regulators credit. They’ve done a good job of getting on top of some of those systems or some of the changes that have happened along the way. I mean, the White House put out an AI bias paper probably about a month ago now. It’s not a law yet, but you can definitely see, hey, we expect probably more and more states to adopt some sort of regulation that looks like that or a federal regulation to come in that helps prevent bias in AI type models and are actively working to do that. So certainly faster than I’ve seen the legal and compliance industry operate in the past.
And it’s addressing exactly the types of things that you’re asking questions about, which is how do you protect customer privacy? How do you make sure you’re not adding to bias or intentionally biasing or even unintentionally biasing some of your models that you’re using to either recommend or automated underwriting. How can I shortcut some of the underwriting steps that I mentioned earlier and make sure that I’m not, incidentally, perpetuating some sort of racial bias that would have been in place in the past. That’s exactly the type of thing that the regulators have put in place and what we’re trying to navigate through as an industry.
Ash: So you obviously have a lot of experience in this space. I think you’ve been at the Mutual of Omaha for, what,14 years now? Almost. Where do you see the future of insurance?
Brian: Yeah, one of the trends that I’ve noticed the last couple of years at InsureTech Connect has been a concept called embedded insurance. So I think that is probably going to take off at least a little bit. If you buy a ski lift ticket from many of the resorts, I don’t think it’s all the resorts. Many of the resorts in Colorado, you’ll get presented with an offer at the end of that purchase flow of, hey, do you want to get like, an accident coverage? So in case you’re injured while you’re on the mountain, we’ll cover the medical expenses associated with that. Check this box here. That’s an example of embedded insurance. We’re the backing carrier for that. We work with a company called Spot who does the little widget to place in that checkout flow. So I think you’re going to see more of that. We’ve already talked a little bit about personalization. You’re definitely going to see more of that along the way. I think you’ll also see probably simpler products for some of the things that the industry has made complex.
Again, in pursuit of that customer experience that we had referred to. There’s still going to be the tax plays, there’s still going to be the annuities that require a little bit more complexity. But I think you’ll see some simpler, both life disability and health type products come through in pursuit of reducing that friction from a customer to buy it once they realize that they need it.
Ash: That’s interesting, yeah. Embedded insurance, I see that more often too now.
Brian: Yeah. Travel is probably the one that I think really kicked that off. So if you’re in the midst of buying a trip, like at the end, almost always, you’ll get an offer for, hey, do you want to protect this trip? Same thing with a concert. Oh, do you want to protect your ticket price or whatever.
Ash: So, just curious, I mean, inflation, right? How is inflation affecting insurance?
Brian: In general, Insurance companies like higher interest rates. Most companies or like most companies, most people, whatever, they dislike high inflation rates. So with interest rates going up at the Federal Reserve raising those, and I think most insurance companies are generally happy about that. Living in a 5% or 6% environment is better for an insurance company than 3% because you got to remember, we collect those premiums, we then invest those premiums to ultimately pay claims. And if we are investing at a higher rate, we can theoretically reduce the overall premium costs. Now the challenge is if you’ve got inflation in conjunction with that, everything else is going up.
So all of the overhead is going up. So you might not see that actual premium reduction pan out because of inflation going along. There is a tipping point though, if rates get too high, insurance companies do not like that. So there’s kind of a sweet spot right now. Most folks in the insurance industry are probably pretty happy with the Federal Reserve increasing rates, although probably less happy about the inflation rates at the moment. Like anybody else, the cost of doing all of the other types of businesses rather than just the investing has gone up for us and many others.
Ash: Yeah. So then how do your priorities change based on that?
Brian: Yeah, that’s an excellent question. So in the immediate term, one of the things that we’ve seen as those interest rates rise is an uptake in annuities because people are like, oh, I can lock in, but last year I could have only got a lock in of 2% – 3% and now that lock in is higher like if I am looking to park some of this cash that I might have accumulated over the past few years into something that’s going to pay me a return for presumably the rest of my life. If you’re buying a lifetime annuity, we’ve seen those go up for sure. So making sure that that is as easy and seamless as we’ve got with regard to top priorities, I mean, we’ve got real budgetary pressures on, are we focusing on the right things? So one of the discussions that we’ve had recently at Mutual of Omaha is are we maybe spreading ourselves too thin? And would we be more effective if we were to focus on a few big things rather than maybe lots of little things?
It certainly maybe sharpened the things that we’re working on because with inflation as high as it is, that might not pan out in the way that we wanted to. Our investment team is like, hey, if you can generate some more cash, we can lock in these high rates for you, which is, again, ultimately going to help both us and our customers out for the long term. Bonds are an interesting move. At the moment, I can get rates of 6& or 7% or 8% as a forecast, whereas before that doesn’t make any sense. If I’m only getting like 3% or 4%, I should probably be investing in the stock market.
Ash: We always ask each speaker about the diversity, equity and inclusion. What are your thoughts on DEI in the insurance space and do you have maybe a personal story you could share?
Brian: Yeah, I’ll actually pull a recent story. So since we, and I assume most companies in the financial services have gone remote, I think we’ve seen diversity increase. Now, whether that’s carried through to equity and inclusion, I would hope because you’ve got a more diverse group, so people who may have not the majority identity, feel a little bit more comfortable because they’re not surrounded by folks who they may not identify as like, themselves.
Certainly with the remote world, it’s more difficult maybe to build community than the traditional way of building community. But we’ve done a decent job of better utilizing teams to have either group happy hours or we try to start many of our meetings with some sort of icebreaker. Whereas before we might have just gotten straight down to business in the office.
So the personal story that I was going to tell is that we bring together my team, the Enterprise Data team, roughly twice a year. So maybe two or three weeks ago, we brought together the team. I was lucky enough to have dinner with several members of my team and looking around the table like I was in the minority. And you can see I’m a white male and I was in the minority racial group. I was in the minority from a gender standpoint. So certainly we have increased DEI. And my personal story was just like recognizing that, right? That might not have happened before we had gone fully remote as a company.
Ash: Wow, that’s really. Cool. You’re saying that’s interesting because the past speakers we’ve talked to are saying that industry is a little bit behind, it’s still lacking in diversity. And you’re saying that when you guys had that dinner, there was a lot of diversity. Wow, that’s great.
If you can go back 20 years and talk to a younger Brian Poppy, what advice would you give your younger self?
Brian: I actually might have spent maybe a bit more time formalizing my tech knowledge. I have always enjoyed being around computers, but didn’t really focus on that in school or even out of school. It was more of a fun thing to toy around with. Yeah, it’s fun. It’s fun to play video games or manipulate Windows 95 to do whatever it is that I wanted to, but I might have been a bit more intentional about coding or thinking about maybe how the Internet might change things.
Ash: This podcast is sponsored by Zelros. Zelros is an AI software solution for insurance to hyper personalize the customer buying experience with insurance recommendations across all channels, boosting client acquisitions, cross sell and upsell. Thank you for tuning in to the InsureBreak podcast. Join us next month as we interview another insurance executive to gain insight on innovative practices, technology advancements, and what the future of the industry looks like.
See you next month.
«So one of the things that we, and I suppose many others in the insurance industry, have learned is as much as you want to give an automated recommendation to a customer because it’s such a personal and financial decision, customers want confirmation that they’re making the right choice. So we can put a recommendation engine out there that helps customers pick a policy, they will still want to talk to a person to verify that they’re on the right path» But how to ensure the protection of consumer data privacy while serving their needs for a more personal touch? « The technology industry has greatly advanced the ability to do that in a way that protects both customer privacy while still allowing you to do the types of personalization that customers are coming to expect from other industries (…) for example through anonymize the data either through consolidating it in a way or tokenizing it »