INN: What are the most important tech initiatives for Progressive this year?
Ray Voelker: We're going to continue to expand our Snapshot program. We have some experiments, some of them going on at our
INN: In Europe they've been doing that for a while. How is it different or why is it different in the U.S. vs. Europe?
RV: I'm not sure what the difference between the European approach and our approach is, but one of the things that Progressive has always prided itself in is very tight segmentation. The actual experiments that we're conducting right now are side-by-side comparisons of the accuracy of the chip, which is obviously very accurate because it's plugged right into the vehicle, vs. mobile devices, so that we make sure that our drivers are earning the best discounts. So it's really a matter of how that technology is evolving and what point you started from. Progressive was the leader in usage-based insurance; in fact, a lot of the European users licensed our intellectual property to get it done. So once you start out with a very highly accurate data collection device, and you're looking at others, it's important to calibrate and understand what -- if any -- differences there are. Clearly mobile technology has evolved over the years and that's made it more practical to run some of those experiments.
INN: A lot of the insurers I talk to think of big data as something they're doing on top of’ or in addition to.’ What’s different about Progressive’s perspective on big data?
RV: Well, I think that it depends on which domain we're talking about. Certainly we have -- and have had for a long time -- traditional data and analytics technology. It's also true that as we have expanded into big data, and Snapshot helped us with that, obviously; we have so many billions of records that are accumulated there, and given our focus on segmentation we've always looked pretty far back in history. So we've had large data stored. But as we see the data horizon out there, we're keeping an eye on the ability to -- I suppose mash-up’ is the word people would use -- our traditional data with our big data, just like everybody else. But as the technology has emerged over the last four or five years, we certainly didn't wait for the integration between traditional and big data to emerge before we jumped into the big data to make sure that we could understand.
INN: What does that mean?
RV: Well, I think that as things like hybrid clouds are starting to evolve, and as some of the traditional analytics vendors are starting to try to understand the markets and the opportunities with big data because, if you think about it, the very early Hadoop and other implementations all they basically do is sort and count, which is a lot different than regression and some of the other more-sophisticated analytics algorithms. So you could have taken a view point several years ago saying, "Until we can run something like that against the large data stores, what can we really do?” Or “Until we know we can inter-mingle our traditional data with the big data, what can we do?" But our viewpoint was that it was better to get started, recognizing that if we chose to we could look back four or five years from now, and say, "Oh, if I had a do-over I wouldn't have done this, that, or the other thing."
We felt it was important to get to the game early by creating teams of high-end analytics on the business side and some of our better data architects and data people on the IT side, before the tool sets were all there. It didn't matter because we had coders and analysts and the raw technology, and we could start to understand some of the promise of that technology so we wouldn't be caught behind.
INN: What's the one thing insurance technologists need to know about Progressive's big data strategy?
RV: [Laughs] We're way ahead; they might as well just give up. How's that? [Laughs]. I'm not in the business of giving advice to my competitors, but I will say this: Just like in the past, it's going to revolve around what your smart analysts can do with the data that's out there. Whether it's usage-based insurance or other emerging data sets that are coming out in the Internet and other places, access to that data is going to be ubiquitous or easily obtained one way or another, just like before.
Somebody started with MVR data, and they were the only ones that had it. Now, if you're an insurance company, you can obviously get MVR data. Some of the things we are pioneering now will be just as ubiquitous in the future. So we're going to make sure that we go after the stuff right now, before it is ubiquitous, but continue to work on the skill sets of the IT and the analyst folks to make sure that we're developing professionals who can deal with that data and turn that data quickly into insight.
INN: I'm really interested in the idea that all of this third-party data is going to be, if not ubiquitous, then increasingly generic. What's the next step when everybody has it?
RV: We say that as if the next step’ will be a week from Sunday! That next step, from my perspective, is when the Internet of Things takes off. By the way, the car is a thing,’ and it's a pretty big and important thing’ for us. As we are now in the homeowners’ business, and we've been selling homeowners with partners for quite a while, increasingly there are smart devices in homes. But there's not a lot of standardization around that data and a lot of thought around that data. So I think that it's probably going to take us a little longer than maybe some people expect for that data to really all be there.
My guess is as Google and Apple and Microsoft compete for being inside the car and their various platforms there, there's very little thought given to the fact that none of those three is going to win all of the cars. That just doesn't make sense. So the consumers of that data, whether they're insurance companies or others, are going to need some standards to emerge. Those three tech giants probably right now are thinking that they'll get all the cars. And there is no real thought about interoperability of that data, etc. So I think there's a lot to be done still in mining through that data and understanding and rationalizing that. We have no desire to only have insight into cars that happen have the Apple operating system, or cars that run the Microsoft operating system; we want to insure everybody. So we're going to have to work through that.
INN: I thought the Apple and Android auto data initiative was mostly centered around on-board entertainment and where's the next gas station where I can get Starbucks.’ Is that just preliminary, before they would move into driving behavior?
RV: I think that all of those companies that are getting a foot in the door and competing for what operating system is going to predominate the car are innovators, and they're always looking for new markets to be in. Sometimes they go into those markets to compete with the vertical players that are there. Sometimes they go into the markets to sell data and enable those markets. I think multiple combinations of those things could happen. You think about Waze - Google owns them, I think. You should know them now because that's the app that tells you where the speed traps are and everything else. It's sort of crowd sourced traffic, etc. They're owned by Google.
INN: Google is now doing that insurance aggregation project: Google Compare. How might that impact Progressive? You're aware of them, obviously. What do you think about them or how you might respond to that?
RV: I think we have a pretty public position on that; we're not going to participate in that.
INN: Does Progressive have challenges recruiting tech talent?
RV: All tech hirers have some challenges; that's just the basic demographics of things. The number of people entering the field does not equal the number of people leaving the field. But we think between things like this Innovation Garage and our focus on big data and our strong focus on mobile, we have the kind of jobs that tech people who are interested in large companies, and large company benefits, are interested in.
INN: Does Progressive do much talent development? Internships, partnering with any of the universities, anything like that?
RV: Yes. We've had university recruiting for a long time. We probably have between 65 and 80 interns every summer. We are starting to sponsor even high school contests. There's a high school coding camp this year. We're trying to make sure that as many kids as have an aptitude toward IT get exposed to it early on so that they can pursue that. Again, if you are coming out of school and you're dream is to be living in a house with your 20 co-workers and coding on a kitchen table, you probably don't want to come to Progressive. But if you like technology and you want to get into some of the cooler stuff like the big data, or how are we going to use mobile to make better experiences for our customers, and you like the large company atmosphere, we think we compete very well, both here and in Colorado Springs.
INN: Have you given any thought to drone technology? That's been all over the news the last week or two:
RV: I personally haven't. But I do have an Amazon Prime account, so maybe a package will be delivered to me relatively soon, that would be good!
INN: How would you characterize Progressive's tech strategy overall?
RV: Your most important takeaway is that IT strategy is really indistinguishable -- at about 90 percent -- from the business strategy. My entire time that I've been here, we've focused on the fact that we have a product that is technology-enabled. Sometimes folks say, "We're really a technology company that sells insurance." I think that kind of gets it backwards. The truth is we can't sell our service without technology. As we introduce new strategies -- if you read the
