Transcription:
Patti Harman (00:07):
Hello and welcome to the DigIn Podcast. I'm Patti Harman, editor-in-chief of Digital Insurance. Science and technology, engineering and mathematics, or STEM careers, are providing many opportunities for today's professionals, particularly women. According to the National Science Foundation, the number of women entering the STEM workforce increased by 31% through 2021, twice the number of men entering stem STEM jobs for the same period as technology adoption permeates every industry. It's creating new opportunities for young professionals in the areas of analyzing data and design writing code, and finding new ways to implement it into daily tasks. Joining me today to talk about the use of data and technology in the industry, STEM opportunities for women in insurance, how the skill sets insurers need are changing and more is Jamie Warner, managing Director of Data Science at Plymouth Rock Assurance. So thank you for joining us today, Jamie.
Jamie Warner (01:15):
Thank you for having you, Patti. Excited to chat.
Patti Harman (01:17):
It is, it's really nice to catch up. Technology is changing so many different aspects of the insurance industry. How is it changing the skill sets that insurers are hiring for today? Is it making it easier or harder to find new employees?
Jamie Warner (01:35):
Yeah, I think this is a great question and I think it really spans different job families. We're seeing different types of things. So first, obviously we're seeing a lot more roles that have to do with data, and part of that is just that data is becoming more available. Insurers have been the greatest data focused companies for hundreds of years, and actuaries are doing these paper calculations back a hundred years ago, even a couple decades ago. And now we have so much more robust data where we don't have to make as many assumptions about the risks. We can actually go and look at data about the risks. So we're seeing so many more roles and data and analytics and those roles are definitely difficult to hire for. People in those spaces aren't as used to hearing insurance as the hot job family. But we do actually have so much really interesting data, and especially now that we've got the large language models, even more data is being able to be pulled because a lot of our data storage today is in forms that are in paper or files and PDF, and a lot of those models actually allow us to pull information from those so we can do analytics and do analysis in ways we couldn't before, not even five years ago.
(02:42):
The other place where it may actually be making it easier to hire is places like in claims and underwriting and the call center where it can be used in training. So a lot of these large language models, we can create bots and training tools that read things or identify, find different information so we can train up new employees more quickly. With some of this technology though, I would say the actual workforce constraints around hiring those types of employees have increased. So it might be a little bit of an even trade off at this point in those areas.
Patti Harman (03:13):
It's really exciting to hear how the industry is changing because insurance wasn't always known as being on the cutting edge of technology, and yet that has just really changed significantly in the last couple of years. Your focus is on data science, which I think is absolutely fascinating. And when I've been covering the industry probably for about 30 years now, that was never a topic before, but can you explain what that is for our listeners and how you're using it in the insurance industry today?
Jamie Warner (03:46):
Sure. So data science first of all is a really tricky topic because a lot of people define it in really different ways. And so I think it's helpful to just think about the ways in which data and particularly models help the insurance industry. And there's a couple different areas where that happened and it's really evolved over the last decade. So we've seen crazy evolution in this space. When I started in this field, it wasn't even called data science, it was called statistical modeling. Fewer people applied to those roles, but essentially there's a couple different areas that this becomes really important in pricing. This can help us build effective pricing models, which is we take all the information we know about someone applying to the company for coverage, and then we use that to predict how much we should charge them. Traditionally this was done by actuaries and a lot of actuaries are still really involved in the process, but now we have a lot more data to help us make a more accurate prediction.
(04:42):
So previously you might've sent an actuarial table where it had gender, age, and a couple other factors. Now we might have thousands of different factors that help us understand. The second place would be claims. In claims. We see a lot of modeling around fraud. It helps us predict and identify characteristics that help us find not just instances of fraud, but also fraud rings, like groups of companies that are participating in fraud. So different types of symbols, other areas where it has become really important in the claims and underwriting space for training and document summarization. We've seen that a lot in legal lately summarizing out key points because we have things like condo documents or really, really thorough lengthy pieces of information that tell us the constraints for the policies, and those can be really hard to get through and this can help summarize and identify key things within them.
(05:33):
So it's really all across the cycle. And then when we think about the marketing agency side, it's a lot of target. Who should you target? Why should you target them? Who's a good customer? Who should you try to retain? How do you prioritize? And we see that in both agency kind of marketing and in claims. How do we triage or how do we prioritize who to go to next, who to service next? So there's just so many broad opportunities for data across all those spaces. So yeah, we get to touch all of those, which is really exciting. Never a dull moment.
Patti Harman (06:05):
What a great way to personalize though what you're providing, the services, how it's priced, all of those sorts of things. It just makes it much better for both the companies and for the policy holders then.
Jamie Warner (06:17):
Absolutely. And the goal is really to use that type of information to get closer to personalization. How do we personalize this? How do we make sure that we're not just proxying what is right for this customer, but we make it as personal as possible?
Patti Harman (06:32):
And that kind of goes into my next question, which is why is it important for carriers then to adopt and really utilize data science these days?
Jamie Warner (06:42):
Yeah, I think it's just a great way to make sure that we do the core of what our business is. And at the end of the day, the whole reason that we have insurance is so that we can pay what we owe. So we want to be able to hold the right amount of dollars in reserve to pay what we are going to owe our customers. And the best way to do that is have really accurate prediction of what those losses might look like. And so this allows us to make those predictions more and more accurate so that we can hold the right amount of money, invest the right amount of money to grow it, and also where we need to request more money if we think that someone might be a more severe risk.
Patti Harman (07:17):
Okay. Carriers, you mentioned this earlier that they have collected vast amounts of data for years. What role does that play then in helping them plan and develop new products now, especially since they can access it and really synthesize it in new ways?
Jamie Warner (07:34):
Yeah, I think this is one of the greatest things is some of the newer models, all that they do is turn text into something that's actually readable by a computer. So one of the things I'll point out is electronic health records or even just general health records, it used to be you have doctors writing things, and we couldn't really use that as data unless we sent hundreds of people to try to type up what they thought they read. Now there's models that actually look at that and translate it to text, and then we can use that to create indication. And so if we're thinking about all of that data, it allows us to do a lot of different things we couldn't before, especially looking at the historic data. We can compare things like, Hey, the call center volume last year versus this year is different.
(08:17):
But we can also say, oh, the calls, the transcripts from those calls, the tones of voices during those calls, people were angrier, they were happier, they were more satisfied. So all this data can give us a lot more information there. I think when we're creating new products, some of the new data isn't just the data internally, but it's actually the data we can purchase externally to better understand people's behaviors. Because a lot of times insurance is about sometimes some risks like weather, climate risk, those ones we can kind of predict using national really great collected data behavior is the other component of risk. And so how do we understand more about people's behavior and what kind of needs they have? A lot of data has opened up the fact that people have different needs than necessarily they had a decade ago. And we've seen a lot of startups come into that space of things like different types of renter's insurance or whether you own or don't own your car, can you get insurance coverage or things like Airbnb and asking about Uber, what are the changes? A lot of external data can help us create better newer products that actually fit the lifestyles of people today versus decades ago.
Patti Harman (09:26):
I don't think that the general public really understands how much the insurance industry has evolved because I remember when Airbnb and Uber and all of those services first came into being and insurers were like, well, how do we cover this? How do we price it? All of that sort of thing. So it's interesting to see how all of this additional data is really making a big difference. Then how is technology changing how you manage data and even do your daily job then?
Jamie Warner (09:57):
I mean, that piece is huge. I think as a data scientist, I'm a statistician by background. When I first started, all I had to know was statistics. Now I need to know it. And that goes for pretty much all of my peers in the space is suddenly we really rely on IT infrastructure and it helps us understand how much we can compete if we don't have really good IT infrastructure. So for example, that's investments in cloud environments and for those not familiar with cloud environments, those are just like you want to think about it as a really, really big computer that you can size up and down however often you want and you can turn off when you're not using it, where you can store as much or as little data as you want and get as much or as little processing power as you want just on demand.
(10:43):
And so those types of environments with the scale of data we have today have become really, really critical. And it means that it has really had to step up and transform from what they have historically had to do at insurance companies. And the second piece of that is a lot of this data is private data. It's data about people, it's data about health conditions. And so we really have to work with IT to protect it effectively and make sure we're putting in governance at a level that a lot of other companies don't have to. So that is really, really critical part of this process. And my IT team, I talk to them every day, so their vision around how to make that work and then also how to bring that out to our customers. So can customers see every aspect of their policy? Can they see the information we have about them? Can they request it? That stuff has become newer things that we didn't have to worry about before.
Patti Harman (11:36):
Just a lot of different advancements and just better ways of doing things for everybody all the way around.
Jamie Warner (11:43):
And they do have a different expectation, right as well? Yes. They've used Amazon, they've used these other tools. So they say, why can't my insurer do that? Why can't they show me what's my cart? Why can't I see all the details in my policy? So we've really been pushed by other industries to innovate because they see it in the rest of their life.
Patti Harman (12:02):
Yeah, it's so funny you would say that. I literally just had that conversation with somebody yesterday and they were saying, is there an impact from other industries? I'm like, yeah, there definitely is because when you look across the entire ecosystem, technology is changing everything and you expect to be able to track a lot of your insurance policy or your claim or whatever, just the way you track your Amazon package or your pizza delivery.
Jamie Warner (12:29):
Yeah, text alerts, they want it all and we want to be able to give it to them.
Patti Harman (12:33):
Yes, very true. We're going to take a short break right now. We'll be back in just a few minutes. Welcome back to the Dig in podcast for chatting with Jamie Warner, managing director of data science for Plymouth Rock Assurance. So Jamie, you're an adjunct professor at Northeastern University. Why is it important for you to help educate the next generation of professionals?
Jamie Warner (13:01):
That's a great question, Patti. Number one, I love teaching, so it's fun for me. But number two, I think they actually teach me a lot more than I teach them. And the reason is because insurance as an industry historically, we haven't had to evolve so quickly and now there's all this new technology and new information coming out, especially in the field of data and data science. And so being in the education environment, I get to fit in front of the cutting edge of that and then take it back to my company. So the questions they come in asking, they're like, I heard this new technique, and I go, oh, I got to go search that new technique. And so really getting to keep a pulse on what's going on outside of the industry helps me bring great information back. And I would say that if you can explain something to a student, you can explain it to an executive. So it's also just great practice for anyone in data if you're trying to translate hard topics. I think getting up in front of a class of students and having to explain that topic is the perfect prep.
Patti Harman (14:01):
Well, and you're a great representative of the industry, so they get to see somebody who is actually working in insurance on a day-to-day basis, and they get to see firsthand how what they're learning in the classroom applies to real life, so to speak. So how do they view data science and other STEM related careers then?
Jamie Warner (14:23):
Yeah, I think they're absolutely shocked that there's so much data science and insurance. And that was actually something for me when I came into the industry. I didn't realize, I thought, oh, these people will pay for grad school and then I'm out of here. And I got here and I looked at just how much data we have and how interesting the problems we solve are. And also because we have regulators and we have customers that we have to answer to, we have to really understand our stuff. We have some of the top statisticians I think in any and insurance, which is so thrilling to be able to work with people so educated in this space and really try to think of innovative ways to use data. So that piece has always been really exciting for me.
Patti Harman (15:04):
Yeah, it's interesting because I think, like I was saying, people have this certain perception of insurance and I love the opportunity to talk to students and to other people and just let them hear, no, this is really what it's like to work in the insurance industry. You have no idea. And I talk about insurance fraud and mean as a reporter, I get to cover a lot of different things. So it's nice to be able to share that. What opportunities do you see in STEM related careers, especially for women?
Jamie Warner (15:36):
To me, there's just a huge opportunity in the STEM space, especially because you've got all these people in pretty much any industry that have grown up with all of the business knowledge. You've got all these people working with tons of business knowledge who want to improve things but maybe don't know the how of it. And then you've got all these folks coming out of school that know tons about the latest technology, the latest programming language, but don't know that much about the business. And so really bringing that skillset and then marrying it with this question, that questions that people are already coming up with, I think is just a huge opportunity to be able to make someone's vision come to life. That's a lot of what my team does. The product team has a vision of how they want things to look, and then we figure out how to make that happen logistically with data.
(16:25):
And that is just such an opportunity, and it's been really fun to watch as analytics teams spring up how heads of these departments they're going, wow, I never got to see the numbers on this before. I had a gut feeling, but I never really knew exactly what was happening. And now we can see into that. And so these folks that are in these STEM careers can really help drive that understanding and especially if they're willing to learn the business basically unstoppable. So I think it's a great track to be on. And I don't say that just because I'm a little bit of a math nerd.
Patti Harman (16:59):
Why should women in particular pursue opportunities in these areas, do you think?
Jamie Warner (17:04):
I think that sometimes we're dissuaded and maybe someone has said, oh, you're so great at English or literature. And I think that a lot of people, because they have great communication skills, they say, oh, well, I can't also be great at math. And I think actually being in STEM and having great communication skills is what we need more of. Being able to be empathetic and try to help people understand and teach is so critical, and especially in an industry like insurance where we're highly regulated, being able to not just understand what you're doing but explain it to someone else is the most critical part of the process. You can build the best model, the best technology tool, and if you can't explain it, it doesn't matter. It won't get used. We can't use it, right? Especially with the NAIC, putting out lots more regulations around modeling. So I just think that everybody should, if they like it, obviously pursue the path because there's so many job opportunities and there's also a real desire for people to understand.
Patti Harman (18:05):
Yes. I agree. As someone who knows that this is not my area of expertise, that's why I am so impressed. And it's funny, I'll laugh with some other reporters and we're like, yeah, we just know that math and science, that's just not our thing. That's why we're reporters. But it's such a fascinating field and being able to cover it from the outside, it's very exciting to me to see how it's changed and the opportunities that have arisen, especially over the last couple of years. What excites you the most about how technology is changing the insurance industry and the job opportunities going forward?
Jamie Warner (18:41):
Yeah, I think the thing that really, really excites me is that we just have so much potential in so many areas. I know at the start, it almost sounded scatterbrained when I told you all the different areas. Data science touches insurance, and it's because it really is, we have obviously huge challenges ahead. We have a lot of climate related challenges. We have a lot of technology related challenges. How do we ensure people and make sure that we're there kind of the worst day of your life to make you whole again? And this data can help us do that, not just a little bit better, but much better in everything from actually calculating the right price to making sure that we do your claim faster. So it's not, oh, no, this happened. Who do I call and it's 15 days till someone comes out? All this different stuff. I mean, I'll give you a quick example. Now when a claim happens, people can take photos on their phone, send that in, sometimes get an automatic approval to get their claim serviced. A couple years back, that would've taken 60 days, right? Oh, easily. So that sort of stuff, just how do we make people whole again really rapidly is so exciting to me, and it's really rewarding to be in an industry where you're having an impact on people's lives and everything you do with data and with technology makes it better,
Patti Harman (20:06):
Right? Yes. It makes a huge difference for them. I know I had an insurance claim within the last six weeks or so, and the ability to use data and technology, it was amazing to me how quickly everything was able to go, and you can really, from a customer service perspective, that just really kind of highlights the importance of having insurance and you're able to service your customers so much more quickly. So that's great. Does anything about the adoption of technology in the industry concern you?
Jamie Warner (20:44):
I think there's always a concern of what are we adopting and who is behind the adoption of it. And so I'm a really, really big believer in education and lack of silos between the business and the data groups. The biggest thing that scares me with data is that you have a bunch of folks that know data and only data going through the data. And the reason that I say that is because, for example, if you go to school and you learn about missing data, you learn about something called imputation, you just fill it with the average in insurance when information is missing, a lot of times there's a reason it's missing and you don't necessarily want to treat those two records the same, for example, if you don't fill out your credit score, it might be because it's bad. And so we need to understand that when we think about your fiscal ability to pay or have a certain deductible.
(21:40):
And so really understanding the business side of things is where I think this technology and this data can make a big impact, and I think they can make a big negative impact if teams are building and don't understand the business, don't understand how the data's created. I think our industry is so cool because almost all of our data is what I call survey data. It's a claims adjuster's opinion, it's an underwriter's opinion. It's the person filling out their form trying to get a policy's opinion. And so we do have to treat it a little differently than if it were just numbers and counts and different sorts of things like that. And that makes it very fun and interesting to analyze it, but it often makes it dangerous if you don't know what you're doing.
Patti Harman (22:25):
Very true. I can see that we've covered a lot in the last couple of minutes. Is there anything that I haven't asked you that you think our audience should know about data science or utilizing technology or anything else related to insurance you want them to know?
Jamie Warner (22:40):
Yeah, I think I would throw out just a quick come as you are to it. I think a lot of people get nervous or scared around math, around data science because they feel like they don't know all the terminology or they don't know all the language. And the thing is this industry has changed so drastically in the last five years. Even people in that part of the industry don't know all the language, all the terminology, and so come with your curiosity, come with your questions. If your vendor or your partner can't answer them, that's on them. That's not on you, and it's really important that we just stay curious here. The second piece of that is for the folks that are thinking about STEM and say, maybe I didn't do an undergrad, maybe I didn't do whatever. What it really takes to be good at data is curiosity. If you have that curiosity in your interest in this field, don't let the fact that you didn't have the formal education or whatever it is, stop you. The coolest thing about the insurance industry to me personally is that we have people from the mail room that are the CEO at a lot of companies, and I think the same exists data. If you understand the way our business works and you're curious, don't hesitate to come over and start exploring this field.
Patti Harman (23:54):
Yes. I agree with that. There are just so many different opportunities. Thank you so much, Jamie, for sharing your insights with our audience. Thank you for listening to the Dig in podcast. I produced this episode with audio production by Adnan Khan. Special thanks this week to Jamie Warner of Plymouth Rock Assurance for joining us. Please rate us, review us, and subscribe to our content at www.dig-in.com/subscribe. From Digital Insurance, I'm Patti Harman, and thank you for listening.