little bit about where we're going as a company, mostly for us, what we consider to be the subtleties of what it takes to win in this business versus just competing. Hopefully, that'll come through today. For those regular attendees of this sort of a meeting, you'll recognize that we don't go to a lot of trouble to reformat data that you already have. We're going to talk more about what we're doing as opposed to what you can reformulate for yourselves. After a few introductory comments, I am going to take a leaf out of some of your books. There's been many times on conference calls where I've wanted to grab a whiteboard and expand on my answer or grab some examples. Well, today, I'm going to do asking some questions. Hopefully, they would be the questions that are on your mind.
I recognize I won't get them all, so we'll have plenty of time for other questions later. That'll give us a chance to really expand on some of the things that are happening in our business, and hopefully, we'll get a good number covered. Let me do a real quick introduction of a couple of our speakers today. On the front row here, we have Pat Callahan. He's going to be one of our figures. John Sauerland, Tricia Griffith, and Jeff Charney at the back. I'd also like to acknowledge the CEO and the CFO, John Auer and Trevor Hillier of ASI. Really great to have them with us for lots of reasons. If I really get tough questions on homeowners, it'd be nice to have them here as well.
I said I wouldn't reformulate a lot, I'm going to have two slides, one on some relative measures, one on some absolute measures before I get into some comments I'd like to make about the organizational strength of the company. On a relative basis, I've shown this graph many times. In fact, over the years, I've done a lot with this particular plot. I've animated the movement quarter to quarter, nothing ever comes out dramatically different, and I'm proud to say that I never get bored looking at the fact that Progressive is one of the somewhat permanent residents of the shaded area there.
I say that with some respect, some humility, also some seriousness in the sense that there are over 300 companies trying to do at least one of the two things, profit or growth, and to find yourselves in that shaded area takes some skills that we think we actually focus a great deal on. We're going to talk about that, and hopefully, you'll see that come through in a lot of our comments today. On some absolute measures, just a real quick base setting. The distribution of our business hasn't changed a great deal in our commercial and auto. I'm just going to focus right now on the Progressive piece of the business, not so much with the ASI addition. About a 90/10 split commercial private passenger auto. Most of our comments today will be on private passenger auto. Commercial is doing very well.
We focused a little bit on that last year. There's just not enough time in any one meeting to cover everything and do it with some reasonable respect. You will see some trends towards our direct distribution, and you can fast-forward that a little bit, and it's getting closer to a 50/50. So be it, and we're very proud of that. On the other hand, it's really critically important to interpret that 50% of our business is through independent agents, and that's the part of the business that we think we can do even better. We can be an even more meaningful solution to independent agents long term. We love that part of the business. This is not necessarily a trade-off between one and the other. It's how can we make the most, and how can we represent consumers in the way they want to be represented.
I think you can see there that we've actually done a pretty good job of dropping our results to the bottom line. Our net income, comprehensive income, and return on equity, all pretty solid. Yes, we would love to see interest rates help us just a little bit, but we'll clearly stay disciplined on that one. Return on equity, if it started with a slightly different number, like a two, I would be even happier, and we'll just have to wait and see if interest rates give us some additional push there. Brian and I, after many conference calls, have talked about our return on capital and how we make sure that if we have unlevered capital, we return it in a responsible way. Last year, we did an acquisition. It was clearly something that we think is very good for Progressive's future.
We've also done dividends, variable and extraordinary, and share repurchases. You can see there that we've actually done a fair amount of it over a reasonable period of time. I'm not going to dwell on any of those issues. The real issue in the room is can we do more on growth. 99% of what we're going to talk about today is can we do more on growth. You can see from the first plot on a relative basis, yes, we're doing well. Can we do more? We think the answer to that is yes. Let me take time now to go through a quick overview of what I consider the operational strengths of Progressive. If I had a pegboard in my office, I actually don't, but if I did, it might look something like this.
I'm going to take you through half a dozen different elements of our business and just comment on them. The little check marks there are a long vestige of a physics teacher I had in high school, and three checks means the obvious, you're doing really well on that. Two, you got some movement. You can do better, but you're getting the hang of it, and one means pretty obvious. Get on the stick. Price to risk. Overall, I'd give this a three check mark. This is absolutely a hallmark of Progressive. The overall might be three, but there are a couple of areas in there that we can do more and we want to do more. The first is really continuous product evolution. To bring a new product to market, and I'll discuss why we might do that in just a second.
To bring a new product to market requires a lot of component pieces inside the company, whether it be from the R&D group to the technology group to legal. All sorts of pieces have to work in harmony, and we're good at it, but we could be even better. This could be our Swiss railway. This is really the thing that makes us very efficient at getting our new information to market. When I say new information, we're not getting as much new information on the non-standard or SAM, for those of you who remember our nomenclature there, into the market. On the preferred, we're starting to get more and more, and richer and richer data sets. We want to be able to incorporate that into new product design. We want product design to always be a continuous function in Progressive.
You'll hear today about our current product design that's going into the market. Our research and development group is already working well into the next design, and the next one is already in the pipeline. This is really an important element of what we see as being able to ingest new data and incorporate it. Also, having that sequence, that gives us one other advantage, and that is we don't have to force things into a product model before they're ready. There'll always be another train coming along in a few minutes. We can keep those product designs coming to the market. Second, Snapshot. We're not going to talk a lot about that today, but I couldn't be happy with that.
The only thing we're finding with Snapshot, it's even more powerful as a rating variable than we originally thought, and we'll show you a little bit of that later. Clearly something that is starting to get traction in the market. Do I wish that this would be a great eureka moment and everything would sort of immediately come to it? Yes, but it's starting to get serious traction, and we're delighted to be in the position that we're in with Snapshot. The last point I'll make is the ability to be even more effective, taking rate to match trend. We always have to take rate. We're never going to not do that, so no apologies for that. If we can do it in a way that just matches the trend, we're so much better off.
You've seen me have to say occasionally we took a half a step back on retention because of prior rate changes, so on and so forth. That's awkward. Pat and I sort of a little bit jokingly, but philosophically agree that three 1% changes are better than one 3% change or any variant thereof. One of the things we have at Progressive is a machinery, if you like, to be able to take rates. A rate might change inside of a product design multiple times. You've got to be quick at that, be nimble at that, and try to do it in a way that matches the external trends. Those are three areas that we can do even better on with price to risk. This may be one of the only investor relations meetings where we're not going to talk specifically about claims.
Claims is actually in very good shape. We love to show you the graph of how to optimize our total loss and LE as a function of LE. We've been very open about where we stand on LE right now. We're very proud of that. We've found some continuous optimization points. Four big principles in claims that we're guiding principles that we care about a great deal. Obviously claims quality, cost, customer satisfaction in the work environment. We are really very solid on all those fronts, so I'm very happy about that. We can never take anything for granted. There's always some more movement in customer satisfaction that we can get, so we'll continue to keep our focus on that. Another area of claims that is really starting to emerge in a pretty big way is data and analytics inside of the claims environment.
Not just the simple analytics that you might have known, but the more complex analytics that allow us to look a little bit at sort of the underbelly of claims management and fraud. How can we detect fraud in ways that we could never see before through data analysis? You'll see just a little bit of that later on today. Then in a more positive sense, how can we use data analytics to help us determine things like the optimal use of medical examinations, independent medical examinations, or determine the potential for a very large loss, even if it's not so obvious when we first get the first notice of loss. Those ones always end up being train wrecks. If there are ways that we can do that, train wreck is not a good example right now. If we can do that would be terrific.
We're using those analytics and claims in a very smart way. Service centers, again, we're not going to talk about them today, but a big part of our strategy and a great part of our strategy for servicing the sort of the Robinsons type customers. Last year, we actually put a push and opened a good number of service centers. We're putting about 25% of our physical damage claims through our service centers around the country today. Hopefully you've all seen in one way, shape, or form what we think about relative to the customer satisfaction we generate. More importantly than that is sort of what I consider to be another one of these subtleties of how you get to win in this business. The service centers take out the friction costs in claims. Reducing the days rental by even basis points matters.
Reducing the rework situations that happen matters. The service center is our best response to that. Those sort of subtleties really come to bear there. Customer relationship management. Some of the changes are a little bit inside baseball. We don't need to go into those today. They're organizational and so on and so forth. We have always been a great transactional company, in my opinion, where we have demonstrated great costs. One of the things we're going to have to do is create this pivot to longer-term customer relationships. How do we treat customers who might be with us a decade, two decades? When do we talk to them? What do we say to them? How do we form those relationships? That's a different Progressive than historical Progressive.
It's a Progressive we love and something we can do a lot about. We've got the base there. We know what we need to get done, the close cousin of that is really going to be on our marketing efforts. We have, and you'll see this later today with Jeff, we're going to separate a little bit consumer marketing. What do we do with consumer marketing? We have a great presence that brings people into the company. Hopefully you got a chance to see some of the commercials running in the lobby. Some of those are recently released, some of them not yet released. What I hope you take from those is the extensibility of the network notion.
The network notion, Jeff will build on that a little bit later, of how we've used more characters to build around the Superstore to really make it come alive. The change we need to introduce is be as good at talking to the consumer population to come in through the front door. We need to be able to be just as good, if not better, to talk to our current customers. When, why, what message? We need to let them know because we have that very special relationship of having had that first relationship with them. Most people form insurance relationships through their auto insurance. We need to say to them, "We've got what you need when you need it. You never really need to move out." A lot to work on from our marketing opportunity, but it really is hand in glove with the customer relationship management.
I personally never found it very useful to find low points of aspiration. If I had to pick one in our industry, I think USAA is just an extraordinary example of the relationships that they form with their customers. Please do not misquote me. We are not USAA. On the other hand, when we see a mountain that we want to climb, and we resource it, and we have a sense of where we want to end up, for some, that gets a little scary. This is the area that has to go hand in glove with our Destination Era. Capital investments, absolutely discipline is really what we need to continue to focus on there. Yes, we could also say, "Gee, if we all knew when interest rates would rise, we would act differently." We don't.
To the extent, if they do, we will be an incredible beneficiary. We will be a beneficiary as soon as possible relative to our duration. We're going to stay disciplined. We're not going to do anything dramatically different than we've communicated to you for some time now. Yes, we'll look back and say, "Oh, that would have been an optimal point to have changed duration right back then, history will be able to allow us to do that." We're not even going to do that to ourselves. We know what we're doing, and more importantly, we're going to focus on what we can control. Right now, this isn't a controllable. Most of what you'll hear today is what we can control. It was great to be able to take an acquisition pretty much in stride.
We think our capital husbandry, as I mentioned earlier, is in good shape. Systems and technology don't generally put those front and center in a presentation like this, but it's critically important. Nothing else works unless we have the systems and technology to really support that. While internally, we all want better, faster, cheaper, overall, we're in a really strong position. Three vectors of change I would say that we need to focus on, and we are focusing on, making sure that the rapidly changing world of data and analytics is a world that we go along and make sure we know where we're positioned in that and that we never fall behind. We love that stuff, that's our sweet spot.
The second would be data or architectures in general for our systems, as opposed to sort of self-developed systems, which might've been 30 years ago. Now we need an architecture that embraces systems that we didn't develop ourselves, systems that some of our partners work with us on, and an architecture that pulls those together in a very meaningful way. The third point there would be making sure everything we do is really high quality. High quality, low cost tend to go together if you do it right. Lean development, lean design, agile development methodologies, those sorts of things are really becoming just a way of doing business at Progressive. That would be my very quick pegboard review of operations. Two reasons I think it's worthwhile going over that.
One is nothing stays static, you should recognize that we have the resources on each of these issues to make sure that we're always staying on top of it. That's our responsibility. The second, I think, is even more important. It's pretty cool to be talking about sort of a new strategy, or not so much a new strategy, but an expansion strategy and making sure all the foundation is in good shape. We are lucky enough to be in a position where I think the foundation is in great shape. I fully admit, after 16 years of doing this, that I haven't always felt that way. There's been times where we needed to focus more on our marketing, for example.
I would tell you that I think we're in really good shape and just in a great place to keep doing a lot of the things that you're going to see a little bit later. I have an assignment from my colleagues, which will help them out a great deal, to do a very quick recap of the last two years of these presentations. Not like when you come into the theater late, and you get sort of a quick recap, but we'll try to do something reasonably meaningful. Two years ago, I gave you a view like this and took you through what I call my blob. I think it's been colorized since then, and you'll see it slightly differently, but it's basically the same thing.
If you remember, down the bottom right-hand bottom corner is Sam, our relatively inconsistently insured or non-standard, whatever label you might want to place on them. We expanded on this, if any of you want those definitions, you can get them. Diane, our much more I'm in the workforce, I'm taking my job seriously, I'm taking life seriously. I'm actually sort of really a great citizen, my life will probably change a lot in the next whatever number of years. Over on the far left for you, the Wrights. Those are the self-bundlers, they've got all the products, but they don't necessarily have one agent or one company, there's a good number of those, we over-index there. The big section in the middle we call the Robinsons.
I think many of you now have become quite familiar with that term, that is whatever definition of a traditional American family you want to put around that, sort of parents, children, so on and so forth. We've been able, as basic as this is, it's extraordinarily powerful when you start to overlay your strategic actions, your market segments, the shopping, so on and so forth, we've done that for you. We did that a couple of years ago. What I want to focus on, and what we focused on as I bring you up to speed from last year, is the boundaries. This is static. Life is not. It's really the boundary conditions that are most important here, the notion of graduating from one state to another is a notion you'll hear about today.
In the white space, clearly, we've been operating, we think very successfully, but we've been really attacking some 60% of the marketplace. The white space for us is the 40% that we haven't really been a major player in. Fast-forward to last year, we decided to devote almost entirely the meeting to what we didn't have. It's always easy to talk about what you do have and what you might be doing there, I want to reinforce that by doing that, in no way does it signal less than our intensity around the places that we over-index. Those are absolutely paramount for us. What we're talking about is an extension strategy, not a replacement strategy.
I ended last year's discussion with you with a business model that says, great, to do that, yes, we have this asset of the first look, if you like, or the first relationship with a lot of people. We have a brand that allows us to do that. How can we then take that into a different place? We outlined some skills and some requirements. We're going to build on at least four of those elements today. You'll see that business model will actually pop. A lot of the work is, as you would expect, work in progress. There is really some significant opportunities to sort of report what we've done, directionally for the future, what we plan to do. I put the cover of the annual report there for those who have read that.
It's an opportunity for me to try to write down in some reasonable way, at least I try, what we're doing, both with our current business and the intensity we have around that, and also the extension into the Destination Era as we defined it last year for you. That clearly leaves open some questions. This format is so much better at addressing some of those questions, and that's what we're going to do today. Let me take you to some of those questions. Clearly, if there's an opportunity for us to be producing even better results who had been in our agency environment. Our independent agent results have not been as good as we would like. What are we doing about it? Pat Callahan is going to take us through an agenda of specific action items.
He's going to give us an overview of the agency world and some action items that we're actually addressing in that environment. To me, businesses atrophy if they don't sort of adapt to trends, and trends are certainly going to be around you. I can't think of a business that doesn't have environmental or technical or developmental or distribution trends. We're going to talk about some trends for us today. For us, we like to embrace trends. We like to sort of study the ones that we think are important and know them better than some of our competitors. We want to be able to leverage trends, not be scared of trends, but leverage them, be the short-term or long-term to our advantage, and build skill sets that are somewhat agnostic to whatever macro trend might or might not happen in our industry.
John Sauerland is going to take us through some of the trends that are on our mind and how we can effectively not only understand them a little better but leverage them to our advantage. The business model I just put up has multiple points in it. We're not going to get a chance to cover all of them. I'm going to ask Tricia to sort of give us an update on, hey, what have we done on some of those elements, skills, requirements? We've built out a Progressive Advantage Agency. How are we doing on that? We'll give you an update on the agenda for the foundation for continuing to attack the Robinsons. I mentioned already that Progressive has become a consumer brand of some note. We're quite proud of that, and it generates a great deal of new prospects, and it's working very well for us.
Jeff's going to give us an idea of just how to keep that alive, what we do to make sure it continually becomes or continually stays relevant. That's the key word is relevant. How do we make something that now has been accepted by the public always continuously relevant? Most importantly, he's going to pivot that a little bit to start saying, how do we use that same brand to talk to our current customers? There's a difference between consumer marketing and customer marketing. We want to make sure we're equally as good at both. Jeff will take us through that. I guess I have to be responsible to the quotes that I make, but I basically say that I'm not all that happy with retention. I have probably been saying that since 2000.
There are some in Progressive, I think, probably wish I wouldn't say it as often as I do, but I'm going to keep doing it. Retention is the Holy Grail for us, and certainly at the part of the business that we are most familiar with, it is sometimes two steps forward, half a step back. There's need to take rate. Consumers are very sensitive to rate at certain segments of the addressable population. Tricia's going to take us through what is really over a period of time, quite some interesting progress on retention, but more importantly, some of the focus that we've got now to make sure that we continually work on reasons for people to stay, eliminate the reasons that they choose to leave.
Clearly, our biggest opportunity, which is far from incremental in any sense of the word, is the opportunity bundled customers in our ASI relationship. That is huge, but we want to make sure it's not the only iron in the fire that we have. We're going to keep working on the rest of retention agenda, and also look forward to what could be a very dramatically different policy life expectancy in a class of customers. I would certainly forgive anybody if they said, "I never thought while you were CEO, we would have Progressive in the homeowners business, or for that matter, acquire much of anything because of my general favor for organic growth." No debate.
The debate really comes around the strategy that we've already demonstrated, that when we have that asset, that customer, if we can form the relationship with that customer, if we can continue to give them product and not have them go somewhere else, we can be that destination for them. We've done that very well with our partner companies. We've done it very well with ASI in the agency channel. What's changed? What's changed is when you get close enough to a company to see not only that it's working, but when there's a match of business models and cultures, you can't help but ask the question, what more could we do? In short, I changed my mind.
I got to know ASI, and I got to know John and Trevor, and I think we can do a lot more in the environment that we've just created, and it will be, as I said, far from incremental. We know we have to earn agents' place as a mainstream offering in their agency. We know we have to earn that. You're going to hear about that today, that we're not playing just to compete, we're playing to win in that part of the market as well. John Sauerland's going on the task of sort of giving you a little bit more of an overview of homeowners, a little bit about the market, a little about some of the shopping behaviors in the market.
Some of the reasons why we think we will get the results we want to get without the volatility that we don't want to get. Although it's going to be relatively soon, next week, he'll give you a first look at least the format that you can expect to see the results of property being incorporated in our monthly releases. Just to finish off my formal questions, I'm going to have Pat come back and talk about sort of, is it like now with big data tools and everybody's talking about it, is that just going to be homogeneity in our industry? Or is it really about how you use them, the insights you get?
While that might be a tricky question to answer, we're going to give you some insight, some of, not our deepest secrets, but a few that give you a sense of just what we focus on and almost the intensity we bring to trying to eke out a little more information from the data that we have, and hopefully reinforce that same point of it's the subtleties in this business that allow you to win, not just the mainstream type thing. Pat will end that. With that, I think we'll get started on the primary questions. I'm going to ask Pat to come up, take us through that agent agenda and why things are going to look a lot different in the future. Thanks.
I'm thrilled to be here today in a new role and to spend a little time talking with you about our agency business. Our distribution strategy is all about being broadly distributed where, when, and how consumers want to shop for insurance. To that end, our dual channel approach continues to serve us well. You'll see over the past 14 years, we've seen solid policy in force growth across both our agency and our direct businesses. I'll be the first to acknowledge that agency has been flattening over the past year and a half or 2. What I want to spend a little time today with you explaining, number 1, a decomp of what's driving that decline, but more importantly, what are we going to do about it?
To start with, I want to go up a level and talk a little bit about U.S. personal lines market and talk about the channel. 80% of personal lines in the U.S. is sold through agents or intermediaries. 35% of it sold through independent agents. This represents a $90 billion opportunity for Progressive in the independent agency channel. Independent agents control a large share of the channel, and it's been stable over time. With that said, our near-term results have not been great. In 2014, our new applications were down about 7%, driven by 3 primary areas. Number 1, we had active account restrictions at some national agencies. Number 2, we saw quoting declines across the network. Number 3, we did see conversion fall after we raised rates. I'll talk a little bit more about each of those in some more detail.
The independent agency channel is made up of a wide variety of business models. As the largest carrier in the channel, we're widely represented in many of those, provided they can meet two criteria. Number 1, they have to deliver profitable business for us. Number 2, they have to have a high likelihood of being a sustainable business model. National agencies aggregate consumer demand both online and offline through affiliates, banks, websites, and many other sources, and deliver a multi-carrier comparative rating experience. Consumers have been adopting or heading towards this market opportunity with reasonable speed of late. We've seen our share. Number 2, we restructured compensation, not only to take down the overall expense level in these accounts, but also to better balance between new and renewal compensation to provide incentives to these accounts to write better business with us.
Number 3, we implemented process changes such as restrictions on bill plans and electronic signature requirements to ensure the business we got from these accounts performed adequately. You'll notice on the screen that we saw an immediate improvement in profitability. However, we also saw an immediate decline in volume. To be crystal clear, we will take that trade-off all day long. At this point, with the account restrictions in place, we're very comfortable remaining open for business in these accounts that do make up a significant share of our agency business. In 2014, our volume was not just down at national agencies, it was down at traditional independent agents as well. The traditional independent agency channel is hypercompetitive, often with dozens of carriers competing for every volume about 25% of what we would on volume that comes from our proprietary quoting platform.
This gap in conversion is widening over time, and that's driven by really two things. Number 1, we're seeing more carriers participating in that auction. Number 2, we're seeing agents increasingly selling price. They're going to the lowest rate in that auction. Recognizing that they were introducing new product models and increasingly introducing product models that provide low new business rates with some subsidization across their renewal book of business. The combination of our rate increases, coupled with their hypercompetitive new business rates, led to almost immediate decline in our quote volume. Agent source is less competitive, not surprisingly, they didn't quote us first as often. We saw a drop in conversion. In that real-time rate auction, it's an efficient market, and we fell from the number 1 or 2 position, and as a result, saw a commensurate drop in conversion.
Carriers that play in this auction environment run that razor-thin line to risk better than any other carrier. The risk that you win in the auction, you're confident you can deliver your target profit margin on. We believe we are extremely well-positioned across all three of those, and I'll cover that a little more later. First, we want to take a look at the overall channel dynamics, and we'll start with a couple of structural advantages that we have in the independent agency channel. Number 1, we're more broadly distributed than any other carrier that writes personal lines through the channel. Number 2, we're the number 1 writer in the channel, which provides us not only scale advantage, but also the opportunity to leverage our fixed expenses across both our agency and our direct channel to provide hypercompetitive expense ratios.
We are the lowest expense ratio of any leading carrier within the channel. We can use that low expense ratio to either offer more competitive rates, potentially more value to consumers, or greater incentives to agents to place business with us, or some combination of all three. We feel very good about our positioning within the channel. Let's take a look at by line of business some trends that are taking place within the agency channel. Independent agents have been slowly giving up some share of the private passenger auto market to direct writers, and that's primarily been simple needs monoline customers. Given our dual channel approach packaged business in conjunction with the ASI property product, we're excited about the opportunity that this presents us to reach preferred customers through the independent agency channel.
We know that writing through independent agents requires a partnership between the carrier and the agent. Our positioning within agencies hasn't always been one of a preferred carrier. Those of us in this room should find these charts no surprise. We are extremely well-positioned on the less preferred end of the market within the independent agency channel. On the right side, while we've been improving our positioning, writing preferred more frequently over the past few years, we still have a long way to go. Today, six in 10 agents don't place their preferred auto customers with us. Those preferred auto customers come with lower pure premiums, lower expenses, and longer retention that we don't enjoy today in our auto book of business. Our challenge is to ask these agents what we need to do to close the gap.
When we did it, they told us three things that we're actively working on. Number one, we need to expand our property offering. Number two, we need to offer more competitive preferred rates. Number three, we need to offer incentives, SAMs, to make sure that we can continue to defend this core element that's made Progressive so successful. The histogram in the middle of the screen illustrates the rate distribution or the new business rate change distribution moving from our old product to our new product. The orange bars represent the more preferred or prior insurance business customers. You'll notice that the orange bars skew left of the zero. We are overall seeing lower rates for preferred customers in this product model. The blue bar is no prior insurance, the less preferred, you'll see both a tighter distribution and a distribution that skews slightly right of zero.
Slight increases for the less preferred. What you will notice is there is some dispersion. The blue bars do spread out, and there's several blue bars well to the left of the zero line, indicating that when we find segmentation opportunities to profitably after introduction of the new product. The chart on the left-hand side is real-time win rate data from these comparative rater auctions, and the blue line indicates Progressive's win rate over time. Shortly after introducing our new product, you can see we significantly increased the percentage of time that we were the lowest price or most competitive offering in the auction. On the right-hand side of the screen, you'll notice that we are writing a greater mix of new applications to more preferred customers. The specific cuts aren't necessarily all that important. It's homeowners, it's multi-car.
Any way you slice preferred business, we are writing more of it after introducing the new product. What gets us really excited, though, is the drivers of that increased mix is both quote mix, meaning agents are quoting us more often on their more preferred customers, and conversion, which we expect with lower rates for the more preferred customers. We do step back and ask ourselves, even with a highly competitive product, do we have evidence that agents will place their best customers with us? We didn't have to look beyond our special lines of recreational products to find that evidence. Our boat, motorcycle, and RV products, typically very preferred, written for the most preferred customers in the independent agency channel, requires two things. Number 1, we've got to increase the supply in the marketplace. Number 2, we have to change agent demand for this product.
I'll talk about supply first. At the end of 2014, we offered our Progressive Home Advantage product in 27 states. We're adding 10 more states this year, and those 37 states make up more than 90% of the U.S. population. We feel very good about our distribution footprint going into 2016. We also know that a third of U.S. households don't own their primary residence. That primary residence, when they rent, those who have renters insurance are far more preferred than the renters who don't. We get excited about renters insurance for a couple of reasons. Number 1, it is a more preferred renter who has renters insurance. It's an under-penetrated market, so close to two-thirds of renters don't have renters insurance today.
We find that those who buy renters insurance from us, 85% of them also buy auto insurance from us, which creates that renters and auto package that extends our policy life expectancy. We have evidence from the direct side of our business as well that when a customer has a renter's policy with us. We've talked about more competitive rates, we've talked about broadening our property offering. The remaining element is incentives for agents to place this business with us. I'll be the first to admit, for many years, we've tried to crack into preferred in the agency channel. We've tried contests, we've tried spiffs, we've tried annual policies, we've tried commission overrides. Package of both consumer and agent benefits that meet the needs of preferred customers that's never before been seen from Progressive. On the consumer side, we're bringing annual policies and more competitive rates.
On the agent side, we're bringing enhanced compensation, differentiated levels of service, ease of use through a combined bundled quoting application, and a path to partnership that enables those agents to envision us becoming, as Glenn put it, a mainstream carrier for preferred business in their agency. The path to partnership is what we depict here in the pyramid, and that the foundational element won't change from today. That will be broadly distributed vehicle and renters insurance available through tens of thousands of agencies. However, for a select group of those agencies who write preferred customers and package business as part of their workflow, there's now an opportunity lower rates for more competitive preferred pricing and ultimately become a partner over time with agents through this package. Accelerating growth within the agency channel is a top priority for me and my entire team. The channel's healthy.
We have structural advantages in broad distribution and our expenses. I'm absolutely confident that the combination of more competitive rates for preferred customers and expansion of our property offering and the incentives we're building to partner with agents over time will restore new business growth in the second half of 2015 and position us for longer term growth well beyond that. The agency channel challenges are absolutely near-term challenges. John Sauerland is going to join us next and talk a little bit about longer-term trends and some longer-term opportunities for Progressive.
Thanks, Pat. Good afternoon. Welcome. I'm going to talk about four trends that we think could have longer-term macro implications for our industry. We thrive on change here at Progressive and the opportunities it presents. It shouldn't surprise you to hear me say that while I talk about those trends, I'll also talk about how we are capitalizing on those trends in the near term and plan to over the long term. The first of those four trends is vehicle technology, sometimes referred to as the path to the autonomous vehicle. Media interest in this space is huge, of course. You've seen almost every day in the paper a new article. OEM activity is also picking up, and of course, there's at least one other very interesting party in this space now as well.
What does that mean for our industry in the near term and in the long term? Transportation network companies and car sharing, urbanization, millennials' attitudes towards vehicles. What do all those trends have to do with our industry going forward? Uber's growth is huge. They're reporting now 40,000 new drivers a month. Where does that go? We'll talk about that as well. I'm sure you've noticed gas prices have been somewhat volatile over time. $2 to $4 to $2 to $4 back to $2. Where is it headed now? More importantly probably, what are the implications long term of fuel efficiency standards that will certainly be entering the industry? Interest rates, very low. You all know this, I'm sure. I guess we can feel good that today they're at about 2X where they were two years ago.
I guess we could also feel good that the line isn't below the X-axis, which it is in some markets today. What does this mean for our industry, and how is Progressive playing? Two years ago, we shared with you some thoughts on some of these macro trends as well as some others, and we shared what we thought it meant for the future, but we also did so in the context of some longer-term history. We told you that the frequency of claims had been going down for quite some time. This is about a 30-year timeframe you're looking at. If you go back another couple of decades, you'll see the same thing. Frequency had been going down. For a number of reasons, we thought it would continue to go down, technology being a big driver of that.
Severity, this graph is in real terms, had been going up. It had flattened out a bit when we showed this to you 2 years ago, we again here thought it would continue to go up due to the continuing rise in medical costs, vehicle repair. We thought severity was going to continue to rise. The fleet growing a bit, certainly not what it had been growing or the rate it had been growing back in the '70s for sure, growing a bit, we expected further modest growth in fleet size and drivers. In net, we said that we thought the industry would continue to grow in real terms, this graph is real terms, over the longer term, albeit at modest levels. Of course, at that time, the industry had actually been shrinking in real terms over the past number of years, 2 years ago.
I'm going to go deeper into those four trends I discussed, then we'll revisit this view to update you on our thoughts. Vehicle technology. Certainly, we're seeing more frequency-reducing technology enter the fleet. There's a bunch of different technologies. I'm just going to drill down into two of them to give you a perspective on how we take advantage of this and think about it in the longer term. Electronic stability control has now been out in the marketplace about 20 years. It's up to about 50% of the fleet. Forward collision warning, a far newer technology, but newer meaning it's been out about a decade and is now up to about 5% of the fleet. Couple important points here I want you to take away. Number 1, it takes a long time for technology to flow into the fleet.
We do not believe that we will see any material level of technology flowing into the fleet on an aftermarket basis. When you think back on that frequency line, we do not see any step function there. We believe it continues to be a slow glide, if you will, on frequency. What do we do when we see new technology come into the marketplace? A big part of our competitive advantage is pricing segmentation. When we see something new that might have implications for price, we study it. We study it hard. Electronic stability control came in. We got data around it. We saw early on that it reduced the frequency of auto accidents, we had fairly rudimentary knowledge at that juncture. We direct price or one-way price or, in effect, offered a discount to all vehicles with electronic stability control.
We gathered more data over time. We figured out that SUVs were different than pickups, were different than sedans. We started pricing at that level. We get more data. We now price it down to the model level. We're now to the point of looking for interactions in the data. What do I mean by that? An example of an interaction is does the youthful driver benefit more from electronic stability control in an SUV than a mature driver does? I have two teenage sons at home driving an SUV, I can tell you, I think I've witnessed firsthand the benefit of electronic stability control in an SUV. Very happy for the technology. You get the point. We want to lead in terms of pricing segmentation. The faster and more accurately we can price, the quicker we win.
Forward collision warning, my apologies, is the other technology we want to talk about. Early results are very promising. We are definitely seeing frequency down in vehicles with forward collision warning. We are also seeing severity up. Why would severity go up? Well, in a lot of the vehicles, they're mounting the radar right behind the front grill. If they don't avoid the accident, little more damage than on the other vehicles. Net, however, we're seeing lower loss ratios. Here again, we're pricing this in, and we will continue to look for those opportunities so we get the best, the lowest frequency vehicles into our book at the right price. Transportation network companies and car sharing, while they're growing really fast, they're still really small. A fraction of a percentage of vehicles, of drivers, of the fleet, of miles driven. Are they having an impact? Probably.
We're not sure, honestly. We haven't yet seen it in the data. We have reason to believe that if there is an impact, we actually inordinately benefit. Our customer set is very young. The users of these services are very young as well. The most frequent time of day of use of these services is late in the evening. To the extent our customer is taking Uber home instead of driving their own vehicle home, we think we're benefiting. Of course, there's the potential scenario where our customer and the Uber driver driving home late at night is the same person. We take issue with that today, but maybe we shouldn't. Over the longer term, we want to be a solution for this marketplace, and we think we're uniquely situated to be a great solution.
We have more driving data with Snapshot than any other company in the country, and perhaps in the world. We're working very hard to get our mobile technology as accurate as that device that we send to consumers to plug into their vehicle. We expect to have that out in the not-too-distant future, and we can see a world where we can merge that app technology with a TNC app so that we can offer our customer the option of having coverage in both situations and accurately price it as we do. We also think it'd be a great service for a TNC to be able to allocate their costs more accurately than they do today. They incur significant insurance costs, and they spread it equally across all their drivers. Might be a tool to manage their driver behavior as well. Could be a different consumer proposition.
We're in conversations with the primary players here. Nothing to announce, for sure, but hopefully over the long term, we can be a solution for this marketplace. Moving on to gas prices. Gas demand is absolutely up. Vehicle miles traveled are also up. They're actually at the highest point we have seen in more than the past decade. What does this mean for accident frequency? If you sit where we sit, that's a really, really important question. Here again, with Snapshot data, we think we have tremendous insights. We know we have tremendous insights into the correlation between gas prices and accident frequency. We get this data real time, every day, every minute. We're downloading the data from the customers' cars, and we've seen over the time period where gas prices have dropped that we've seen a shift in the mix of trips.
You're going to drive to work, you're going to drive to the grocery store. You may or may not take the weekend getaway based upon the price of gas. We've seen the mix of longer trips go up as discretionary miles traveled, if you will, go up. I'm not going to tell you the exact formula here, but suffice it to say, if you look at the graph on the right, brakes per 1,000 miles are lower for longer trips. We use braking behavior in our Snapshot algorithm. You can put some of the pieces of the puzzle together. Most importantly, I want you to take away the fact that we are using this information and have been using this information to set price levels.
We understand gas prices, we understand what we think the implication on loss trend is, and we dial that in when we're changing rates. Interest rates are very low. We all understand that. This doesn't have direct correlation with industry size or profitability, but a lot of players, as I'm sure you know, vary their investment philosophy and their underwriting philosophy based on the environments in those respective marketplaces. Just to be very clear, we don't. We are very clear on our target margins on the operating side, and we view the investment side as supporting the operating side. As yields have dropped, we have dropped duration to protect that asset and, of course, to be better positioned for whatever the future holds here, but we don't know what the future holds. We haven't chased yield with duration. We have also not chased yield with quality.
Admittedly, we give you this metric in all of our releases, our average credit quality is now at A+. It was higher, but we're still a very high-quality portfolio. Looking back, we're pretty happy with our performance. I will immediately acknowledge the starting point on this chart. That said, comparing ourselves to a reasonable index, this is not how we benchmark ourselves internally. If you look at the Barclays Intermediate Government Credit Index relative to our performance, we feel pretty good. 5.4% annualized return here. If you throw our equities into the mix, you get to about 7.7% over this time frame. That's some discussion, deeper dive around the trends. Let's revisit the longer-term view we saw. Frequency is still down. It's down about the same amount as it was two years ago.
Severity, which you may recall had been flattening out, has actually picked up a bit. Again, real terms. The fleet continues to grow a bit more. Again, not huge growth in the fleet. We don't expect that. Net, where we saw the marketplace contracting a little bit two years ago, we're now seeing growth. We fully expect to see real modest growth over the longer term in what is approaching a $200 billion marketplace. I want to talk about one other trend that I don't consider a macro trend, and it's a very, very important trend in our market. That is the shift to web and, more importantly, mobile. About a third of our customers today in direct quote with us via a mobile device, a handheld or a tablet. Three and a half years ago or so, that was virtually nil.
We're now seeing that same experience trickle into our servicing platforms, not surprisingly. Pat said our approach is where, when, and how, that is absolutely what we are playing here. We want to have great service when it comes to quoting, buying, and servicing your policy and having a claim however you want to do it. It's really important that we've evolved our mobile technology as that trend in terms of mobile shift has accelerated. We've got to have as good an experience on a handheld as a desktop, or perhaps even better. We measure the yield, in this case, of a person quoting their auto insurance on a handheld versus the desktop internet. I'm happy to report that we are approaching what we internally call mobile parity.
That means we are getting as good a yield on a handheld as we are on a desktop. That allows us to continue to advertise aggressively in mass media and in mobile, which is certainly growing in terms of a category. Very important that we've had this success, I'm elated to tell you we are there, we won't stop after we hit 100, I assure you. Quoting in the property space is just starting to move to the web. I will first say, by far and away, as Pat showed you, complex needs households are going to continue to use agents. Okay? If you have a home, most likely you're going to have an agent. Increasingly with millennials, they are looking for their solution on the web and on a mobile device.
I personally see property online about where auto was back in the mid to late '90s. We have today a third of the share of online property quotes. We are number 1 by a factor of about 2. We are to some degree leading this trend, we certainly want to continue to lead it as we move forward. Property is a more complex product than auto. The quoting is more complex. We worked with our Progressive Home Advantage partners to get to better solutions online and mobile. This is actually a Homesite application that you arrive at naturally through Progressive. After we ask a few questions, the app goes out and pulls information from public sources such as tax records, pre-fills answers to questions like square footage, like numbers of bathrooms, like numbers of bedrooms. Then we have a number of extensive questions actually about your home.
Honestly, they're fairly odd questions for most people who haven't quoted their home. The first question is, what type of roof do you have? Virtually all homeowners carriers ask this question. What type of roof do you have? I admittedly failed that question. Provided with some pictures that I could scroll through and select, I then understood that my home has a hip roof. The next question is, what is your roof made of? Happy to say that I actually could get this 1 right, but not everybody can, right? I have an asphalt shingle roof. The thing I love most about this app is if you can't picture your roof, you hit the button on the bottom left, we give you the street view of your home. Pretty good, huh? You get the point.
Agents are primary distribution for complex households, hence ASI, and I'll talk more about that in a bit. Millennials are buying the renters. They're going to buy their condo. They might be buying their home down the road on the web, on a mobile device, and we're going to be there to lead. Talked about a lot of trends. We thrive on change. We thrive on taking advantage of trends. If you think back, the shift to direct, the shift to the web, the shift to credit and rating and underwriting, usage base, I'm very confident we're going to play these trends successfully as well. I'm also very confident that we will have a robust marketplace to do so. We think we got modest real growth in our marketplace that's approaching $200 billion, and we think we got plenty of room to grow at slightly under 9% share.
I'll now turn the floor over to Tricia Griffith to talk about Progress, Seeking the Robinsons.
Hopefully this chart is familiar to all of you. It's something that we talked about last year, and you'll probably continue to see it. What it represents is the overall market and Progressive share within our specific consumer segments. Picking up from last year, clearly, the Robinsons are big. They're about 40% of the market, and frankly, we're not that big yet. What I'm going to talk about today is really our approach from last year's business model, what we've worked on, and what we're going to continue to work on because this opportunity for us is really fantastic, and we feel like we're really well-positioned. The good news is we know that some of the things that we've been doing over the years are working.
As you can see In the last five years, we've grown our Robinson policies in force by about 24% on a combined annual growth rate. This is on a small base, so I don't want to exaggerate our growth rate, but we don't see this trajectory ending. In fact, we see it accelerating. That's really the good news. I'm going to go back to the business model and go through three things. One, we've already talked about, and I think I'll be the fourth person to talk about our obsession with segmentation. It is truly an important part of Progressive. What I'm going to talk about is our obsession now with customer segmentation. Our customers that we have in our books that we know are going to be our future Robinsons. Again, just a reminder, Robinsons are the auto-home bundled.
From our very beginnings, we were obsessed with ratings, we knew we could rate for any risk. We've been obsessed in the last 10 years for understanding our advertising and how we can reach certain consumers. We have to do that same thing in our customer base in order to have them understand at the exact right time that we have the right products and service for them. When we think about big opportunities like this, we usually have sort of concepts that we rally around. Think of in the past 10 years or so ago, when we really started talking about customer service. We related that to NPS. Think about our retention opportunities. We relate that to PLE. We're relating our opportunity to be more of a preferred carrier, and the notion around that is really around graduation.
We have millions of customers on our books. We need them to graduate to the Robinsons. We're going to grow some from the outside, we get that, but we know that's our biggest opportunity because they've come to us because they like our brand, we have good service, and now as their life evolves, they will graduate. This chart shows you the graduation rates from the end of 2013 to the end of 2014. You can see the biggest share of our improvement was through that graduation. We also got new business in right off the street that bundled right away, which is great for us because we weren't always seen as a company that had all the products and services to meet your needs.
The fact that people are seeing that right away is also very telling to us that our advertising is working, and our reputation's changing. Obviously, we've lost some people along the way from 2013 to 2014. That's extraordinarily disappointing to me. I care about every customer that leaves. Some of these customers unbundled, it may have been a situation in their life, and we got to figure that out. The ones that we've bundled and they leave are hugely disappointing. We continue to look at survey results, and in fact, a couple of weeks ago, John Murphy, who leads our CRM organization, met with some customers, some Robinsons. He met with some in-force Robinsons and some Robinsons who have defected just to figure out, "Why'd you leave?
What could we do?" Having those conversations, we'll continue to have those because we have to minimize this bar. What they told us was, "We felt like you sold us this great company. You came in here, you cranked up our rate a lot, and it didn't feel very fair, and we didn't feel protected." It might be the rate on their auto, or it might be the rate on their homeowners, but you didn't tell us. You didn't explain anything to us. They also like to be taken care of. They want to be nurtured once they're with us. They actually are okay with little rate changes here and there because they understand that. They want to be taken care of, whether it's in a claims experience or a call center experience.
We like to call the special things we can do service clues or kind of winks. We need to do more of that because they expect us to do more of that. This is really the primary way we believe going forward we will grow in that Robinsons book. It'll be graduation from our current book of business. Let me walk you through how we're thinking about segmentation in order to grow our Robinsons and graduate Robinsons. Let's take a generic look at the other segments. We know there's Sam, and he's not very likely to graduate. There are some, and we will figure out those Sams, and we'll make sure they graduate. Dianes are much more likely to graduate. That she goes through life events because right now she's not a homeowner, so we need to really stay with her.
Obviously, the Wrights are more likely to graduate. That's sort of the first order of our segmentation around graduation. Let's go to the next level. Let's overlay our financial responsibility model onto our Dianes. As you can see, average to poor credit Dianes are far less likely to graduate than our good credit Dianes. Again, we're kind of getting to that next order of segmentation, understanding the graduation opportunity. Now we know good credit Wrights, good credit Dianes are more likely to graduate. What time in their life? What age do they graduate? We know that good credit Dianes and Wrights are much more likely to graduate right in the area of about age 25 to 35.
We also know if we look at every interval in between, we know that a Diane that's 25 is a lot less likely to graduate in her first term, but a Wright who's 34 and comes to us is way more likely to graduate in that first term. This is really a complex problem. We need to understand and continue to narrow down and segment, but also understand that these needs happen at different times. We overlay some other opportunities that we see through segmentation, and the chart building behind me talks about those things that are knowable to us. We know under the age of 35, multiple car, high limits, full coverage, things like that. This information is also quite valuable to us. For me, it's a little bit after the fact.
You have this data, and you can kind of go back and say, "Oh, that makes sense." It continues to help us evolve. We really need to use this segmentation to create an action. That action has to be that that arrow continues to rise and are having what we call future Robinsons become Robinsons. We've been talking about that a lot as we've rolled out our business plan this year. One of the things that we are doing differently, we've reorganized our CRM organization, and we've created a role that is specific to understanding how to graduate Robinsons all along the way, our end-to-end customer service through cross-selling of policies, through having advanced analytics. This is a complex problem, we have to really deeply understand and have a relationship with each and every one of our customers.
We're really excited because he's the type of guy who's quite analytical and is going to figure out this really complex problem. By the way, we love problems like this. This is really in our wheelhouse. We're going to continue to understand how to improve those future Robinsons. Andrew Quigg is going to work in lockstep with Jeff Charney's group on understanding from end to end, how do we make sure we have just a great process for our customers, that we have really elegant marketing at the exact right time, that when we know she's going to buy a house, we're there for her.
To me, this is a really exciting opportunity. Let me move on to another topic in terms of additional life events portfolio. Last year, if you recall, I went through sort of a population of one, I walked you through what I called Imagine Diane, she was single, then she met a guy, they got married, all the things that came with it with her evolving insurance journey. At that point in time, a lot of the products, I admitted this last year, a lot of the products we didn't really have on our portfolio. As you can see, I'm extremely proud of the progress we have made of our events portfolio. Some of these are longer term products you would have. You have a classic car, that's one of them, mechanical breakdown, and some are more event-driven, travel insurance.
I'm really excited about where we're at here. We're going to continue to offer more and more products that our customers say they need, and we will continue to do that because we truly see that as our role in becoming a destination insurer. What this caused me to do, because people have actually been coming to Progressive through these products and not always through the auto product, it's caused me to just take a look at how we viewed our role now and going forward. Typically, because we were mostly monoline auto, we would take the view of the world that there's about 116 million households, about 105 million of those have at least one auto, we have about 9 million policies in force in the auto world. Makes sense. That's how we've grown. That's how we've looked at things.
Taking a little bit fresher look, we know that we actually have 3 million more incremental policyholders, people that have come in through other products. We call them advantage products. It could be a special lines product. We have the start of a relationship with 3 million other policyholders. You take a step back, you know that we've always tried to bundle. We know the more products you have, the longer you stay. We have about 1 million customers that have auto plus one or many other products. Again, just a fresher look of sort of our ability to have more and more relationships. As we go towards the future, we really see the opportunity to be how can we cross-sell 3 million customers our auto policies, how can we cross-sell 8 million auto customers our other policies?
We really have an opportunity, 11 million opportunities, to be able to, where we already have these relationships, continue to cross-sell and evolve these customers as their lives evolve. A much different way we're looking at the world. Because we're thinking about our world differently, we have to think about our metrics differently. We've created in this last year what we call a Destination Era scorecard. Metrics is foundational to everything we do. We like to measure things. We like some things to be instantaneous, so we know right away if something works, and we like other things to be longer term. These are very important to us, and this will continue to evolve. This is our Destination Era scorecard. I'm going to give you just a couple of things we look at, not necessarily the data, but the concept.
We're going to look at percentage of households with auto plus how many customers. We're going to look at products per household. We're going to look at and bifurcate future Robinsons and Robinsons based on new business app growth, on NPS, on PLEs, consideration, preference, all of those items. From an agency perspective, we're going to look at our number of PHA agencies, and of course, what Pat talked about, how many of those agency customers can we bundle with auto and home? I'm going to wrap up right now and talk about our in-house agency. Last year and even the prior year, we were building what we called our in-house agency.
We now have named it our Progressive Advantage Agency, we have really put a lot of committed resources into the agency, we've worked on many more partnerships with our PHA homeowners partnerships to be able to facilitate that bundling process on the direct side. We can take care of you. We call these our swim lanes. We can take care of you from your very high risk to the ultra preferred homes. More importantly, in that middle section, which takes up about 85% of the market, which is really important to us, we have multiple partners that can help meet those needs. To us, this is a really important part. Again, I'm very proud of all the work we've put into this because this is the key to the Destination Era. We have to be able to cross-sell our direct customers in this agency.
We're going to continue to put many more committed resources on this. As you can see, as we have grown our staff, we're going to continue to grow our staff, the quote starts go up, which makes sense. Because we have had many more partnerships, you can see that our DNQ rate goes down. DNQ is Does Not Qualify. It's basically when someone calls in to get a quote so we can bundle them, none of our partners were able to take that person. That's unfortunate. We want to minimize that. As we've added partners, our DNQ has gone down and our yield has gone up. Yield is defined as by finish over quote start. I'm going to wrap up my section and talk about something Glenn introduced last year, he called it our invisible balance sheet.
What he talked about was that our business strategy is only partially reflected in our current period GAAP reporting. We have a value to every customer that lasts a long time. We refer to that as Unearned Lifetime Earned Premium, ULEP. Just one more acronym for you. We know that that's been growing rapidly. One of the reasons is because we have been putting more Robinsons on the book. Robinsons clearly are more valuable to us on a long-term kind of book value because they have longer PLEs and higher LEV. The chart on your left shows that we've been growing ULEP at about 7.5% on a combined annual growth rate, and that's higher than our earned premium growth. The chart on the right shows that even though right now the Robinsons are only 1% of our new business, they're actually 5% of our ULEP.
Again, we understand and love every customer that we can rate for, we know that that bundle customer is one that's preferred and will last longer, and the value, of course, is much higher. To wrap up, significant progress on what we said we were going to work on last year. Very proud of that. Number two, we have even a deeper plan and the right talent to execute on that plan. Lastly, and probably most importantly, the energy and the excitement around the Robinsons and the Destination Era is the highest I've seen at Progressive since I've been here, and I'm halfway through my 28th year. That is really exciting. In fact, I was telling Glenn not too long ago, I had a woman from one of our CRM organizations come up to me and say, "Guess what?
I graduated two Robinsons right before lunch." People are getting it. They talk about it. What we know is when 26,000 people get it, we get it done. Now I'd like to introduce Jeff Charney to talk about how this really fits in with the marketing and our continuous brand.
Good afternoon. Welcome. The last time I spoke with you, I told you a little bit about the way our department is set up, and it really runs like a successful Hollywood network. I want you to think about networks like HBO. Think about ESPN. Think about upstarts like Showtime, or even think about innovators like Netflix, because these networks really operate on a simple principle. They want to reach the broadest audience as possible, target those audiences, and reach them in a very efficient manner. We do the same. Today, I'd like to take you behind the scenes of our network, show you how we are really winning with consumers. As Glenn says, we're winning with consumers. Over the past several years, we're winning.
How can we leverage those same network principles that we utilize with our consumers and apply them with our existing customer base as well? Just to get you in the swing of the network, I am going to ask you a very general question. Fox, The Simpsons, 26 years in, what is the biggest thing they have in common with Modern Family 6 years in? The answer is they are both anchor shows. They are the centerpiece of both networks. A lot of business drives off those shows. We have the same anchor network. We have the same anchor show, and that is Superstore. If you think about Superstore, it has a hero, you have foils, you have ensemble cast, you have set changes, you have spinoffs from that, and you take the character in different worlds.
It is the same format for any successful sitcom, it happens over and over and over again. Our hero is not Homer Simpson, it is Flo. She is real, she is relevant, she is authentic, I am telling you, even 8 years in, she is more relevant than ever. Over time, you put an ensemble cast, mostly an improv ensemble cast around her to really play off her a little bit. Even more, you add new characters every so often. Not just add new characters, but allow us to get different brand messages out there in a big way. Obviously, we look at all sorts of different characters, adding them to the mix all the time. Most of you have seen our cast of characters on one network, on one different channel, and that has been TV. We have been running for 8 years.
We have 114 different commercials, or we call them episodes, 30-second episodes, where we can look at all these episodes as a mosaic and say some are targeted to Sams, some are targeted to Dianes, some are targeted to the young Robinsons, some are even targeted to the Robinsons. For me as a CMO, it is a beautiful thing to have such a great network. If you think about it, all this content, it is a very complex media landscape. I do not have to tell you. You understand it. It is a very complex media landscape. I am not going to try to confuse you any more about this because it is very confusing out there. Norm, I am going to give you our chess pieces, what we are doing with our chess pieces. I am going to try to simplify it for you.
The way we simplify our approach to this network is through 6 Cs. The center of those 6 Cs are our consumers and our customers. There is a big difference between those two. The way to reach those consumers and the customers is through amazing content, great content, in the right context. Again, content at the right channel that consumers can digest all this content. If you do that correctly, from a network perspective, we have been trying to do it correctly, you get a connection. As Glenn said, we have got the connection. We are acquiring a lot of people. We are getting the quotes. You ultimately want to build from that connection into confidence with your customer base. I am going to take you through it one at a time. Let us go through content. We run an inverted pyramid strategy for the network.
It's inverted, at the top of the pyramid are our heavy hitters. Our heavy hitters more is our core Superstore asset, that anchor programming, which is Flo and Friends. Flo and Friends. From there, from a network, you can spin different characters off. You can spin off Flo's family. This is Janice, who's played by Flo. She plays all six characters. Flo's family is one spin-off from it. There's a young Flo spin-off, and there's even a Superworld spin-off that really doesn't have any. Flo's not even a part of it. We go from there to a direct response. If you want to sell hard, you can't sell too hard with Flo. She's been around eight years. You don't want to run her in the ground. You want The Box to help you do that. The Box has a little swagger, a little attitude.
He can sell a little harder than Flo and use The Box sparingly from time to time as part of the network strategy. If you want to sell a little softer, again, Tricia showed this last year, this is our apron platform. It doesn't run incredibly large amount, but when it does run, it makes you feel great about the quality, the quality behind the apron, which is Progressive. Softer sell. From there, you have miniseries, you might have seen them in the lobby today. These are Rate Suckers. This is a very disruptive series. It's a little bit in your face and really can carry a very different message about Snapshot. Hitting Snapshot in a very different way with Rate Suckers.
If you are under 35, a lot of you are under 35, you might be seeing a very targeted message, specifically on Facebook, with a character named Baby Man. Last but not least, any great network show, you look at all the network shows, we've studied a lot of them, they have cameo appearances. Cameo appearances. One, LeBron James stopped by the network earlier. I'll talk about that. Obviously, Marshawn Lynch also stopped by the network, I'll talk about those in just a second. That gives you a sense of the content. Look at context. TV still matters. Don't get me wrong. TV will always matter. As I told you several years ago, digital, social, and mobile, it's not emerging anymore. It's not emerging. It's not something people, oh my God, digital, social, mobile, it's emerging. It's ingrained in our DNA.
It's ingrained in our culture. We've been at it for several years. It is really part of us, if you think about the philosophy, if you look at our content strategy, I've told you before, our content strategy is to out-create, not outspend. We've talked about that. Our digital strategy is a very simple strategy, adopt and adapt. It's more than just fun words from a marketer talking about adopting and adapting. You have to be first. You have to be first. Be an early adopter. Be an early adopter, once you adopt, make sure you can adapt this into your network strategy. It's got to move fast. Speed, speed, be willing to really try things new. Just to give you a few examples, I mentioned Baby Man. A lot of you might not have seen it.
It's a little edgy content, if you think about it, Baby Man appeared on people's Facebook feeds. I'm a Facebook user, and I'm telling you, if you put something bad in my feed, I'm going to be pissed off. I'm not going to like it. Baby Man was endearing, it was engaging, and people really took to Baby Man. Again, trying to tell people, cut the cord and get off your parents' insurance company and come to Progressive. If you take it even further with event-related type things, and think about LeBron James. Think about nine months ago. LeBron James says he's coming back to Cleveland. Second game of the year against those dreaded Chicago Bulls. The second game of the year, we did a switch because that second game of the year was Halloween. Was Halloween a great holiday for marketers?
LeBron put on a wig, put on an apron, put on a name tag, and yes, he put on a different kind of headband and turned into Flobron. Flobron, for a digital audience only, reached hundreds of millions of people. Again, knowing your audience, knowing the channels you're going for, it worked great. It worked great. Think about the Super Bowl. Oh my gosh, the big story was the Deflategate. It's still going on. The biggest, most enigmatic character within the Super Bowl was Marshawn Lynch. Very controversial character. He turned out to be a big teddy bear, a very big teddy bear that really had not done any interviews until he talked to Kenny Mayne. Progressive interviewed him, released it a couple of hours before media day, and you can see he had a great fondness for Flo and for Progressive.
Again, very timely, very interesting content in the right context of the Super Bowl. Some of the less big name things, if you think about gamers, how many of you are gamers here? I'm definitely a gamer, and one of the biggest games really popular with our Robinsons is Words with Friends. What if you put your brand icon in the midst of that game and have Words with Flo? All these are very strategic. They seem fun. They're very strategic, and if you think about adopt and adapt, it doesn't only go through just the mediums, it goes through all the new technologies. Look at all the technologies from Periscope to Meerkat to Google Cardboard. We want to be first, and we want to be at it and start adapting them into our network.
Sometimes you really look at it and say, how can your content, how can your television content play really nicely with your digital content as well? One of those great examples, and again, playing in the right sandbox, is a spot we ran called After School Special. Really give you a little. I'm not going to show you the entire spot, but basically, it's Flo in an after-school special kind of way, consoling one of the ensemble cast, Jamie, and saying, "Jamie, look, I know you feel bad." He lost a big battle. He lost to our competitors. He didn't have the lowest price. Sometimes we win, sometimes we lose. That's what comparative rating is all about. Unfortunately, he's feeling down. Think about you as a parent, think about you with your kids. She's in the locker room. She's consoling him. He doesn't feel great.
This is the last 10 seconds of a spot, After School Special, in a very kind of syrupy sweet way.
You can't win them all. The important part is you help them save. Thanks, Flo. Let's go get you an ice cream cone, champ. With sprinkles? Sprinkles are for winners. I understand.
Sprinkles are for winners. I understand." We understood those four words. We understood how good, how big it could be. We really did. We thought it was going to be great. Said, "What do we do?" We said, "Look, how do we get that out in the zeitgeist in a very different way?" We just had launched that spot, and there was a thing going on, the National Championship of College Football. I'm an Ohio State grad. I'm an Ohio State graduate school grad. Ohio State versus Oregon. We decided whoever would win, thank goodness it was Ohio State that won. Whoever would win, right after the game, a few minutes right after the game, we would put something like this out.
Sprinkles are for Ohio." It could've been, "Sprinkles are for Ducks." Oh, my God, it would've been horrible if it would've been that. I'm offending people in Oregon now. It's just, "Sprinkles are for Ohio," those four words, and boom. Boom. It was like a social cannon went off. A social cannon went off and literally over and over again, thousands and thousands of people saying sprinkles are for winners. Something's very current, something's a little bit raunchy, but they just went out there. It's user-generated content. One of the best part of this user-generated content, it hit everywhere on social channels. It also hit mainstream channels. This is a nice little story that happened this weekend. It's a high school softball team. That was the Bad News Bears. They were not good last year.
The coach adopted a strategy, said, "Sprinkles are for winners." It's going to be our rallying cry. In every game, he would literally put sprinkles over the girls' heads. They won the state championship. It's a beautiful story from rags to riches, and just, again, sprinkles for all. This is the front page of the Tampa papers. Again, it's a big part of the zeitgeist. These things don't happen by accident. You know they don't happen by accident. We have to test and test again. It's part of our culture. It's part of our DNA. I'm just going to show you a couple of examples. We first launched it on the Rose Bowl. That was Florida State versus Oregon. We launched the Rose Bowl. We just launched the spot, ran it a little bit ahead of time before the game. No social boost.
You can see when the national championship at 12:06, when we put that "Sprinkles are for Ohio" out there, you can see what happened on the Twitter feeds, the Twitter sphere. People started talking about it. You can see we tried it with the Grammys. We put the spot, ran our spots, Sprinkles Winner spots before the Grammys, put some social boost on the content there, and you can see what happened here. Uh-oh, what happened at the Oscars? We didn't run the ad. We tried a different ad because we wanted to see, to make sure to test and test again. We know we have something with Sprinkles for Winners. You'll see it keep winning over time, and it's going to happen organically. So much to talk about from a literally a television and digital perspective.
Let me give you a less glamorous side of the business. How many of you have been to a trade show recently? A trade show. If you've been to a trade show recently or you've been in the past 20 years, it hasn't really changed. It's the same trade show. It's the same trade show guy, gal sitting there trying to get you to come to the booth, giving away tchotchkes. Say, "Come on, come, come." That's not a trade show. We approach it as entertainment. We approach it as a show. It's a trade show. It's a show. It's entertainment. From a motorcycle show, which again, you have energy in that show. You have things that families can do. If it's a boat show, we're the lead sponsor of our boat show, International Boat Show.
You have a boat where people can come on and you can go fishing and have a lot of time spent with it. If you go onto an RV show, again, we know that audience. We know how to reach that audience and provide the right content, right context to that audience. Last but not least, motorcycle rally shows. Motorcycle rally shows, there's a couple coming up in Sandusky next weekend. Again, I encourage you, go out there, feel the vibe. Because a lot of our competitors just put their name up there. They put their name. They put their name on an event. They put their name on a sporting thing. We don't. We're a title sponsor for a reason, because we want to own the show, and most importantly, own that customer experience. We learn a lot from these things.
I mean, most of the people spend upwards of almost 20 minutes at the booth. 20 minutes with our brand. We learn a lot. We don't have it cracked yet. Nobody has it cracked on how to measure trade shows, but we're getting closer. If you think about all of these things, talking about measurement, I mean, we wouldn't be doing these things if they weren't working. Superstore. Think about Superstore. Look at the runway of Superstore and look at, a lot of people, a lot of marketers use different dashboards. We look at it as an audience rating perspective. It's an audience ratings. Again, you can see here from Flo likability and likelihood to quote, we continue to go up. Again, just to remind you, eight years into the network. Eight years into the sitcom. A couple other dimensions.
We look at so many brand dimensions, so many metrics. If you think about makes insurance easier, think about your future Robinsons. This is very important to them. Think about, again, important to everybody is saving money. Everybody wants to save money. All segments want to save money. Again, we got a lot of messages out there, but we are right in the game on this message. Right in the game, and this is a very small subset of carriers that can really lead in this area. Tricia last year talked about quality, that journey to quality. Again, we're not all the way to Wrights yet. We're not there, but you can see we're in an upward trajectory on quality. Last but not least, again, this is really another wonderful thing about Progressive is that we buy our media inside the company.
The majority of our media is bought inside the company. We are all on the same floor. We're not in different zip codes, not in different area codes, not in different cities. We're on the same floor. We can buy it and be very efficient in our buy. You can see here our cost per prospect has remained constant despite having roughly a 30% growth in media spend. Again, just another example of network working with content and context working well together. I covered a little bit of context. Now comes the really fun part for me. If you think about it, consumers, we get it. We've been running a network for some time. My job as CMO is to really look at this and work with Tricia and John and all the executive team to really make sure we build on that connection. We have the connection.
Let's build on that connection. The easiest time to build on a connection, if you join a company, is to welcome you into the fold. If you join a place because you want to be a part of it, let's reach out and welcome our consumers in a wonderful way. Looking at the data, and we look at a lot of data, you can see that there's a lot of touch points. This is 81 different touch points. 81 different touch points at a potential consumer. This is not any one consumer, but a potential consumer could get hit 81 different times with different messaging from us. That's a lot. If you look at these things and reprogram those and really look at every one of them, and you could see what could happen here.
You reprogram these things, you have messages that would be more welcoming messages that'd be more affirmation, some to extend the brand and some to renew. We bucket these things. We bucket these things, and it's not easy to do this. We are very good at this. In a company like Progressive, we have very great people that really look at these things, and you have to look at these things one by one by one. It's not easy, and I'll give you just a couple of examples. This is a payment on schedule notification. Payment on schedule. This is basically telling people you're taking money out of their account. Would you want to get this kind of message? It's a little bit of a transactional message, but if you did something like this that says, again, "Payment, you're right on time." Your icon talking to you.
It's just talking to people in a backyard fence way, and it's really aligned to our brand. Just one more example, and I could go through many, many other of these, many examples here. This is an action needed reminder. Oh my gosh. The action needed reminder. Danger, Will Robinson, action needed reminder. It's intense. We want to be disruptive with our customer, but not this kind of disruption. If you did something like this, "Uh-oh, we missed a step. Together." It's just talking to people, as Lynn says, talk to people at the right time in the way they want to be talked to. It's not that we're making this stuff up. It's not rocket science. We know the future Robinsons. We know how they talk, and we want to talk to them in that same kind of way. Again, that's looking at the connection.
Let's look at the connection and bring this close to home with confidence, the journey to confidence. We've got the connection, let's build on it with confidence. The best way to do that is to hit people during life event stages. Tricia's talked about this. We know that you're a new dad. We can serve you a message that says, "Here's a little bit about being a new dad." If you are buying a new home, we know that about you. We can serve you a message like that. If you're a renter, we know about that, and we can get you the right content, right time, and also if you got married. These are just four of the many. Four of the many. Again, this is that kind of stuff.
Some of the stuff we do on the consumer side, that's the West Coast offense, that's great, but the stuff on the line of scrimmage is just as important to me and just as important to the company. All of these things. Some people just want to be communicated through the content hub. They want to read things. Others want to see videos. The Wrights might want to see a video about homeowners, about a PHA testimonial video. Some people like Sam and Diane, they want to talk directly to our icon. They really want to talk to Flo, be it on a loyalty program or be it on a motorcycle. On motorcycle, last but not least, our Robinsons.
They want to really understand what goes on in a service center and really go behind the doors of the service center and be served it in a very palatable way, in a way they really can view it in the best way possible. All of these things are just hitting people at the right time, right place, through life events. I told you how we measure on the consumer side. We're just as diligent on measuring on the customer side as well, from welcome emails to content hub language to having Flo or not having Flo in the piece of content, on and on, all the click-throughs on this stuff. We're pretty relentless on this, and we'll continue to be relentless on this.
We know that all of these things are adding up to a better experience from our customer, and hopefully, over time, that experience will really result in a longer policy life expectancy for us. There's the art and the science working very closely together. Just in summary, we've covered a lot of ground here. You guys have been great. We've covered a lot of ground. If you think about it from a consumer and customer perspective, we've been applying the network strategy for many years now. We know it works with consumers. This is very different than any other marketer will tell you. Nobody is applying this network strategy. Consumers get it. We hope the customers will as well. The network can work for both. It will not work unless we have the right content in the right context. It's easy to say this, guys.
It's easy to say it. You know it. It's very difficult to do it. Last but not least, we have the connection. We have the connection we need to win their confidence. I want you, people in this room and people listening out there, I want you to think about confidence and think about brands that you are confident in. I really want you to think about it. If you're confident in a brand, if you are really truly confident in a brand, you've reached your destination. If you've reached your destination, you will stay. Thank you. I'm going to turn it over back to Tricia, who will talk a little bit more about our retention efforts.
Glenn outlined at the beginning what's going on with retention, and the big question is, what are you doing about retention? Clearly, in the last decade, our PLEs have increased slowly. I'm not going to sugarcoat it because as you can see in the last quarter, we did one of those half-step backwards that Glenn talked about at the beginning. I'm not happy to see this. This is not a good place to be. We want to continue that trajectory, and we're going to continue to work towards that. Actually, even though I'm going to talk more specifically about some things we're doing around retention in each of the segments, it really has been somewhat of a rallying cry in the last year of the company. We have a lot of work done around this.
Even most of the talks today ultimately get to retention, whether it's growing Robinsons or fulfilling our bundle package in the agency business. It really is around having a reason for our customers to stay, whether it's through product or services. Although it's nice to see the continued trend and disappointing to see the recent downtick, we still see that as something of a goal, and we're very goal-oriented. We're numbers-oriented. Earlier this year, Glenn asked a group of us to work on something we call Path to 50% Increase in PLE. Stepping back and saying, okay, let's set a goal at some point in the future to say, how can we increase our PLEs by at least 50%? Of course, that will be the first of many as we achieve that.
This team has been working on that, and it's been great in my new role to be able to understand all the things that really affect retention. With Pat and I working on the product and pricing and the development, then with John Murphy, I talked about when we have that product, how do we best service and sell it? Ultimately with Mike Sieger, when the rubber meets the road, what happens in a claim? The four of us talk about retention all the time, talk about it in terms of the Destination Era, graduating Robinsons, growing both of our channels. If we can all get together and rally around that and have multiple strategic plans underneath all of these things, I'm confident that we'll get there.
While this is disappointing, it's just a rallying cry for us to say we're going to get better. Just like anything, we look at our retention in terms of our segments. As you can see clearly from SAM to Robinsons, the retention continues to increase. I'm going to take you through one little piece of each of our segments and some of the things we're looking at from a retention perspective. Let's go to SAM. SAM has been our traditional strength. Again, we love SAM. He's great. If we can price for him, that's perfect. We want to keep having SAMs come in the book. There's about 40% of our SAMs that are pretty early defectors. Again, if we can price for them, which we can almost all the time, that is great.
What we've learned from the new product rollout that we just recently released was that there's about 60% of SAMs, a different segment of SAMs that are actually way more likely to stay, and we're able to give them a decrease. In fact, their PLEs are 17 times longer than the SAM on the other side of the house. That's pretty significant. We know that their new business loss ratio is also significantly lower. Not all SAMs are created equal. We want both SAMs, but we're understanding them to a much different level. We talk about Diane a lot because Diane is so relevant in terms of our life events profile.
One of the things we need to understand, we're continuing to understand as we talk about the advanced analytics around the data within Progressive and the data that we can purchase and understand externally is, how and when does Diane shop? We know when Diane moves, that the frequency of her shopping goes up significantly before and after. Again, we need to understand, is it best to reach her two months before? Two weeks before? A week after? We need to continue to sub-segment that and understand when to really get the marketing that Jeff talked about to Diane in the way she wants it. We know if we get there, then her tendency is to stay.
You can see this stylized decay chart, and it really represents the difference between the blue line being Diane with an auto, and then the orange line being Diane comes to us, she gets renters in her first term and ultimately gets homeowners in her second term. The conditional PLE difference on those two lines are significant, and that's why this is so important. The next two slides that you're going to see are from the Wrights, and they're the U.S. population, so this is not Progressive data. We take those trends and we look at them and say we should take that and assume a lot of this is on our book. Over time, in the last 40 years, the definition of the traditional household has changed a bit. It used to be married couples with children.
That's about 40%, 40 years ago. Now that same definition, married couples with children, are actually under 20%. We know that changing demographic is different, and we need to understand and have products that fit non-traditional households, and that is really important. As we evolve and understand all the different non-traditional households that are there, we will continue to offer unique and different products. In addition to that, we know that at any given time, about 30% of the U.S. population has a child under the age of 18. Those children are going to get on the policies, and that adding that youthful driver is always a big experience from a rate shock, and we know that.
We have to figure out how to minimize that because we know the Robinsons and the Wrights are customers that we desire more of, are going to have children, and we want them to say, "Yes, I have Progressive because my parents always had Progressive." We need to really understand that rate shock and either help to minimize it, help to educate them on why it happens. More importantly, I showed you the fact that we have over a million customers with auto plus other products, and as that continues to grow, we need to look at things differently and more of a household underwriting and not just linear in terms of underwriting per product.
That's the exciting part, to be able to understand how we can do that and make sure that as people get added to the policies, it doesn't give them a reason to shop. The Robinsons, they like good rates too. They want to have stable rates. In our new product, we have even a subsegment of the Robinsons that have gotten, on average, about a 10% decrease, and we have seen the conversion go up significantly. Again, that's important for the Robinsons. They are inert normally when they get here, but they're always going to look for a good deal. That's really important for all of our customers. One of the things that the Robinsons like and what we learned when John Murphy went to the focus groups was that they want stable rates. It doesn't mean you can't tweak them a little bit.
They get that, they don't want to have that rate shock. This is an example of a normal distribution that happens after we roll out a new product. You can see the varying rates from plus 10, minus 10, and it really extends. Think of each one of those bars as a discrete rate. This is how normally it pans out, which makes sense based on our obsession with segmentation. When we overlap what we call rate stability, and all of you, we presented this to you several years ago, it really allows us to smooth out the range changes and get our customers to the indicated rates over time and kind of eliminate that rate shock. The Robinsons appreciate this as well. For all of our segments, low cost. It wouldn't be an IR conference without talking about our cost structure.
We continue to be very pleased with our progress on cost structure. On the left-hand side, you'll see our expense ratio has declined in the last five years by just shy of two points. What's incredible about that is this has happened in the same time where we've increased our advertising. We're going to continue to care about being lean and efficient. On the loss adjustment aside on the claim side, we've reduced LAE by over a point in the last five years. If you go back even another five years, that curve, that line is steeper, so it's been over two points in the last 10 years. Those are significant things that we have done in order to keep our rates competitive.
You have to care about being lean and efficient because that is the name of the game, to have rates continue to be competitive. We are not going to take our eyes off having a very competitive cost structure. I'm going to wrap up by showing you sort of what we think about when we think of the path to 50% improvement, it's really around taking segments of our population and other things that we know are important to our consumers and having very specific strategic plans underneath to get there, to be able to say, "If we do this, we believe it could be five months on our PLE to get to that." Again, we love those numbers, so we're going to continue to do that. This gives us an example.
Clearly, graduating Robinsons is a big part of it and will continue to be a big part of it. What do we think from a PLE perspective? We know people like competitive, stable rates. We know that people want elegant retention marketing. Get to me at the right time and help me make it easy to come to Progressive. We know that our current customer base wants to be nurtured. Make sure to take care of me once I've made the decision that I want to have a long-term relationship with Progressive. While I'm not at all satisfied where we're at with our PLE or our retention, I'm very satisfied with the plan we have in place, and I look forward to updating you the next time I see you. Now I'd like to bring John Sauerland back up to talk about homeowners.
Why and why now?
Thanks, Tricia. The core of this question was actually the introduction or the very beginning of the introduction that Glenn read in a conference call we had back in December announcing the acquisition of a majority interest in ARX Holding Corp. I encourage you, if you did not take part in that conference call, to listen to it. You can get there at our website, or of course, you can probably get a transcript from some of your other sources. I will attempt to do that approximately hour conference call just in the next a little less than 10 minutes. As Tricia pointed out, we have about 1% share of the Robinsons. We talk about a lot of white space here, 40% of the marketplace are Robinsons. We can grow a ton by growing more with Robinsons.
We break that down by channel, we see a pretty good penetration in the Sam, Diane, and Wrights segments through our agency channel. We see far less success in our agency channel in the Robinson segment. Multi-product households, I think we've told you this, but here's a little quantification, much more profitable. If you can get the auto and the home, and maybe motorcycle and boat as well, you're talking lifetime profit of three to 4x an auto-only household, and that's common across both channels. A ton of opportunity in terms of greater profitability when we penetrate multi-product households. Our progress in direct has been pretty good. We are approaching 20% of our auto customers now having more than one product through Progressive, and the bulk of that is that orange section you can see, which is property. We've done this through a third-party model.
Home Advantage third-party model, 10-12 home advantage carriers today in excess of about three-quarter million property policies in force direct. The third-party model has worked really well in our direct channel, and we'll continue to use that to penetrate this marketplace. The agency side, we've seen far less success. We had a partner in this channel, wasn't perfectly oriented to the channel. We now have a new partner who is outstanding, grew up with independent agents. ASI is an independent agent-focused company, outstanding operator. We're looking forward to getting the graph on the right side here looking more like the one on the left side, and maybe even steeper. Bundled auto and home in the independent agency channel is about a $50 billion opportunity. You take auto and home together, the entire marketplace is above a quarter trillion dollar marketplace today.
A little less than 20% of that is within that independent agent channel as bundles. More than half of the independent agent's business, we think, is bundled households. We wanna be consumers' number one choice. We wanna grow a lot with agents. We gotta play here to make that happen. You might be sitting there thinking, okay, I get it, that's not all new news. Why now? Two reasons. One, to some degree, we think the property marketplace has changed. We think the dynamics, to some degree, have changed. By far and away, more importantly, we think we found a great operator we're working really well with in ASI to penetrate those bundles in the agency channel. Over the past dozen years or so, average premiums in the property market have risen about 5% where we've seen virtually no growth in the auto marketplace in total.
If you go back about 20 years, the total homeowners marketplace was less than a quarter of the total auto marketplace. The homeowners marketplace today is approaching half the auto marketplace. Some fundamentals there we think have changed. Have the risk pools changed inordinately? I'm sure they have, but therein, we think lies some fundamental changes that are favorable. We also note that the best players in the property marketplace do really well. This is the past five years, so when you look at the total combined ratio at a 102, which isn't that dissimilar from auto, I get there haven't been a lot of storms hitting U.S. coasts in the past five years. What I want you to focus on is the differential between the best and the not so best, if you will. Right? The best 50 loss ratio, the worst quintile an 80 loss ratio.
Very happy to say ASI resides solidly in the best quintile, the best do really well. We shared ASI's results with you over the recent past in our press release back in December. You have seen these numbers before, but I share them again with you for reference. If you total up that combined ratio column, you come to about an 83 combined ratio. The other thing I want you to take note on this slide of is the difference between direct written premium and net written premium. The difference there is reinsurance. ASI uses reinsurance fairly extensively, and I personally believe it's a core part of their competitive advantage. We'll talk more about their reinsurance program in a moment. You can make money in homeowners. We understand that it requires a little more capital than auto.
This slide is simply to illustrate if you make great combined ratios, even though you have to have more capital, you can still achieve very robust returns on that capital. A number of assumptions that go into this chart here, but I think you get the point. If you make great combines, even though you're operating at a 2 to 1 or below, you can make great returns on that capital. ASI, as I mentioned, does use reinsurance extensively. They target less than 10% of their surplus at risk in any given occurrence. That will continue, so they can continue their reinsurance philosophy post the acquisition. They also do a lot to ensure that their exposure in the second catastrophe in any given year is not highly dissimilar than their first.
They prepay, if you will, some of their reinstatement premiums, which in the industry is referred to as reinsurance premium protection. Again, largely their exposure for the second occurrence, much similar to the first. Very diversified reinsurer set, virtually all A-rated or better, and they employ multi-year contracts to ensure they avoid some of the spikiness in premiums that happen after storm years. Over the longer term, we think we have a lot of opportunity to work with ASI to underwrite at the household level. We think we're starting to do that well across the products that we've historically written. We think we understand consumer behavior pretty well with vehicles. ASI is a great underwriter and a great pricer in the property segment. They're doing more going forward to understand behaviors, if you will, in the property itself.
We think over the long term, there's a lot of opportunities to be really good at the household, and we look forward to doing that with them. Beginning next week, we will release results inclusive of ARX. We will provide you segment-level view of ARX. This is, I hope of you're familiar with if you read our monthly releases. You can see property is a separate column. We'll also show you the policies in force count for ARX. I do want to mention that reporting for property at a monthly level is going to create some differences, if you will, than auto reporting. Reinsurance accounting as one. We will also show some amortization of acquisition costs, intangible assets within the expense ratio for the property segment here. You'll get to see obviously loss and LAE to understand the underlying performance of the business.
We've always been very transparent, I think, with our financials. We'll certainly continue to do so. We will update the guide to monthly reporting that we have on our investor website. If you have questions, check that out. If you have further questions, feel free to give us a call. We're obviously really excited to be working closely with ASI. They're a fantastic operator, and we look forward, and I'll steal a line from Glenn's annual report letter. We look forward to going forward with ASI with a far wider aperture for growth. I'll now turn the floor back over to Pat Callahan to talk about data and analytics.
Data-driven decision making has been in our DNA since the early days writing non-standard business, when we realized there's no such thing as a bad risk, only the wrong rate for that risk. At that point, we realized collecting and analyzing data was central to our business operation, and it remains so today. Across the world, we're seeing measurement and collection of data expanding at exponential rates. As more and more data becomes available to us about our customers, we relish the opportunity to leverage that data to generate insights and better meet consumers' needs.
Additionally, as advanced analytics tools become available to our team of analysts across the organization, they look forward to opportunities to spend less time collecting, curating, and cleansing data, and more time developing and testing hypotheses that have major impact on how we run the business and also ultimately how we change the business. Today, I want to share 4 examples from across the enterprise that demonstrate to you how we're leveraging better tools and additional data to improve our accuracy, increase our efficiency, and ultimately lower cost. I'll start with 2 examples from the pricing arena. Pricing for weather-related losses as part of comprehensive coverage is not all that challenging. It's something we do every day. What is challenging, though, is identifying and isolating abnormal weather's influence on our other core coverages, liability and collision.
Because a failure to be able to identify and isolate how abnormal weather influences underlying trends results in mispricing. I talked earlier about the auction environment, and imagine a world where benign weather takes place during the historical period, and that benign weather, unfortunately leads us to under-select our trend. We're underpriced in that auction environment, and we win tons of business that when weather reverts to the historical norm, we're underpriced on and lose money. The importance of isolating these one-time abnormal weather effects can't be overstated. As we do it, we can better uncover the underlying trends that enable us to price more accurately, deliver stable rates, which lead to ultimately longer retention and more consistent business growth. Now, pricing for weather-related events as part of our big data solution is nothing new. We've been doing it for several years.
What is new, though, is the scale and speed with which we can ingest larger and larger data sets and process them on a faster and faster rate. Our initial model took one data point a day on both temperature and precipitation from around 300 weather stations around the country. We solved this model once a month and applied the model output to about 300 territories around the country. This alone substantially improved our trend select accuracy. Our current model increases the sampling rate by 24 times. We take hourly data now from three times the number of stations, so close to 1,000 stations. We solve the model daily, and we apply the model output across 100 times greater territories than we did previously.
I'm not going to do the math for you, but you can imagine, number one, the size of the data set has expanded exponentially and resulting accuracy from the model has improved dramatically. This enormous data set is not only enabling us to model weather more accurately, but it's giving us insights as to how weather affects our consumers that we never knew previously. These data insights present the potential for frequency reduction through advanced early warning notification to our consumers of localized weather events. We're not there yet, but the potential is certainly there. The second pricing example that I want to share with you is within usage-based insurance or our Snapshot program. We've previously shared just how powerful and predictive our Snapshot rating element is in predicting insurance losses.
Today, as John mentioned, we have more data than any U.S., and probably any carrier worldwide, 13 billion miles of driving data. We have a team of data scientists and a state-of-the-art big data environment that's enabling us to continue to invest and advance this predictive power. The challenge that we face, though, is given the early adopter status that we had of this technology and our consumer success, we continue to add data at an incredibly rapid rate. The size of the data set at this point, it would be virtually impossible to analyze it with traditional tools. This would force us to do one of two things, either sample the data or truncate the data, both of which would significantly compromise our ability to leverage this first-mover and scale advantage to really advance the model.
Fortunately, the state-of-the-art big data environment that we've invested in, and continue to invest in, is enabling us to take advantage of the entire data set, and to draw new insights and rapidly test hypotheses that we have about that correlation between driving behavior and insurance losses better than any other company. We've previously shared that our most predictive rating variable is UBI or Snapshot. Greater than points, our second most predictive. I'm happy to share today that our latest algorithm in the usage-based insurance space is also more predictive than points. At this point, it's more predictive incrementally on top of the model that's in market today. You can imagine how much more powerful and how much more predictive this latest algorithm is.
The key takeaway for you about this is that while there's dozens of carriers who are offering usage-based insurance solutions in the marketplace, we are leveraging our big data environment to both preserve our scale and our first-mover advantage to ensure that not only can we defend our lead, but increasingly extend our lead in this space. I've talked about a couple of pricing examples, but the best pricing in the world doesn't matter if you can't efficiently acquire customers. I want to share one example from our customer acquisition space. This example, we had an employee who had a hypothesis, had an idea that there was potential correlation between one of dozens, tens of thousands rather, of federal government datasets and insurance shopping. Had an idea. Maybe there's correlation between these two datasets.
Our analytics framework enabled the individual to rapidly acquire, ingest, and test multiple hypotheses as to what correlation might exist between these datasets and insurance shopping. Quickly coming to a conclusion that there was high correlation, and the resulting action was fed to our internal media group, and they took action by, number one, increasing bids, and number two, timing them in known periods of increased insurance shopping. The outcome from this simple example is 10,000 incremental policies during these periods of increased shopping. A small example, but another example of how our analytical framework, and also our mindset, is enabling us to increasingly improve the efficiency of our business. Accurate pricing and efficient customer acquisition are two essential elements of our business model.
We all know that $0.75 on our premium dollar is handled through claims, returned to our customers through claims, and the accuracy in claims payments is an essential element of our business model. Reducing fraud within claims payments is key to maintaining a competitive price in our marketplace. We do a pretty good job with this today. We've got a league of trained investigators. We have some proactive modeling and tools in place, and we also have reactive on-demand capabilities to model some of these networks if an investigator or a claims rep has an intuition or sees something they want to investigate further. However, that wasn't good enough. The team recognized that with the advent of big data tools, they would potentially be able to draw more insights and greater connections between claims than we ever could through the manual processes we have today.
We've been investing significant dollars and assembling a massive data set that includes billions of claims data elements. These elements are things like phone numbers, names, addresses, dates of birth, email addresses, locations of claims, VINs, types of years, makes and models, body shop providers, medical providers, you name it. Billions of elements of historical data that we're building into this massive set. On top of the initial load, we're also loading on a daily basis, millions of additional data elements from claims that are open as we learn new things or new claims that come in the door and we find connections. This massive data set on its own creates no value. The value is created when we leverage advanced matching algorithms that enable us to find connections between these claims that relate elements to each other and potentially indicate suspicious activity.
These data elements were previously there all along. They were incredibly difficult or virtually impossible to find. Think of them as the proverbial needle in a haystack. They're there. They've been there all along, buried between billions and billions of irrelevant data. Now that we can find these connections and provide this information to our claims investigators in a timely fashion, they can focus on the right claim with the right information at the right time. What this does is dramatically increases the efficiency of our claims investigations. When fully deployed to our fraud investigations, will also improve the effectiveness of how we find and reduce fraud in the overall network. Improving efficiency improves our LAE and lowers cost. Reducing fraud in the network also reduces cost, enabling us to lower prices for customers who buy our products to use them as intended.
I could stand up here all afternoon and give you more and more examples of things that we do with our analytical data sets to improve how we run the business and increasingly change the business. I'm not going to do that because I'm probably giving you things, too many secrets already. What I will do, though, is I'll leave you with the notion that a sustainable competitive advantage can't be found buying servers and buying software. It can be found, however, in the DNA and culture of an organization whose primary source of historical competitive advantage has been acquiring data and analyzing it better than any other company. I'm confident that as new data sources emerge and new analytical tools emerge, we will leverage those data sources to develop insights for our business that are greater and greater value than any other company.
Our investment in the big data set tool area started long before it was a household word, and that was due to insight from Glenn, who recognized the trend and said we needed to get out in front of it. Next, I'd love to turn it back to Glenn for some closing comments.
Thanks. I actually know what's coming, and I still get a kick out of hearing the story. That's my privilege, I guess. I'm just going to wrap it up here a little bit. There's one element of our business model that just doesn't lend itself well to slides or numerical analysis, and that is the talent. The talent, a small piece of which you get to see, some others in the room, but many more that are not in the room. That's the only way that we can even have the thought process to envision some of the things we envision and to be able to get it done.
One of certainly the lasting legacies of Peter Lewis will be to Progressive, which is highly embraced by all of the management team, is to make sure that we try to attract the best possible talent and create a culture where we actually want to do our best work and have fun doing our best work. Sounds easy. There's probably variability in that. We're very proud of what we've achieved, and we can never ever lose that because that's the competitive advantage that really makes a lot of this possible. Our ability to have aspirations far greater than our current accomplishments only makes sense based on the perspective we have on talent. The talent, I'm happy to say, in my assessment inside the company is deep and strong.
We're taking a very active approach to term even more so than we have in prior years, trying to give some of our highest potential talent opportunities to see different parts of the company. Three of the four presenters today have been in their jobs about 40 days. We do it in a way where it's not a dramatic change, and we don't lose momentum. We're doing that consistently throughout the company. Quite an effort to really build that kind of a culture. That's the reason that we can do so much and so much of what we talk about today. I also think a lot and spend some of my personal time making sure that the pipeline is equally as exciting. Through our intern programs, through our relationships with academic institutions, the pipeline is not only just full, but it's exciting.
There are people who actually want to come and work at Progressive, and that I just want to make sure doesn't get overlooked because I don't have a slide that will show that. That doesn't make sense. It really is the heart and soul of everything we have today. Today, I hope we sort of gave you a very clear understanding of this was building on the strategic platform that we've put forward in the last couple of years. This was about execution. Pat took us through some very key points in our agency distribution. We've got to respond to the hyper-competitive environment with hyper-responsiveness. We'll do more. We've got to have the new product that we're rolling out, rolled out faster. We've got it. We've tested it. We're ready to go. We've got the next one coming along.
With ASI, we build a package that ultimately puts us in a mainstream position with agents. John gave us a sense of trends. Yeah, there are trends. Let's look at them. Let's embrace them. Let's find out ways to leverage them. Pat gave us a sense at the end with some data that hopefully reinforced a lot of those subtleties about what makes a difference in a company that really embraces their craft and tries to exploit it every day. Tricia and Jeff, sort of together, I think you start to see a whole new dimension to Progressive about how we're starting to treat that relationship with customers and not only the sell and the post-sale and the cross-sale, all those sorts of things. I think we have a very different perspective on that.
Tricia's results in other parts of the company make her especially suitable to take a look as a new role in Chief Operating Officer Personal Lines of the customer relationship management claims and personal lines and working in tandem with Jeff. Those are things that we really can see some great futures for ourselves. As they bring out the stools, I think that's my catchword to make sure they bring out the stools. I guess I have one opportunity to be a little bit I guess I have the right to be optimistic as a CEO. That seems fair. I am. I also am, hopefully, very objective. I actually don't remember a time in Progressive where I'm as optimistic as I am now. We have so much in the mix. We have a great relationship to attack a whole new part of the marketplace.
We keep building on the skills we have. We have great product, we have great skills, we have a great brand, and we're never satisfied. That's a good place to be. For me, it's a great time to be at Progressive. Let's talk again in a year, we'll give you an update. Cheers. For all the questions that we didn't get to that are on your mind. I think we need to do the wait until we get the microphones so we get on the webcast.
Are we on? Okay. Yeah. Cliff Gallant with Nomura. Just one, I wanted to clarify a comment you made about commissions to agents. Was there going to be an increase only on cross-sales? Secondary, just sort of bigger picture on the homeowners market. Could you talk a little bit about the competitive landscape and why you think that there is an opportunity for another major underwriter in the market?
Sure. John's very familiar with this stuff, maybe you can build on more of the second point. What Pat gave you was a sense of we are going to have, with ASIs, we currently have a wide distribution of both auto and home. No change. We're also constructing a program that we believe that there will be a smaller and relatively select number of agents who have a lot of preferred bundled business to give, obviously it has to be in concept, where we're going to do more for those agents, both in product design and compensation, that ultimately it gives them the reason to place that business with us. I'm going to extend that to something that may not be quite as obvious.
While we want to do that's how we meet their needs, we will be an incredible beneficiary of that because we get more data. In many ways, we want to get that data. We want to get it at the right price, of course. As we get more and more data and more and more of the interactions in the household, we absolutely sense that that's not only a lower pure premium business, but it's also something that we're more than willing to pay appropriately for. How about the marketplace?
To that I would add competitive rates, we got to have those. We know that. Easy use is also really, really important in the independent agency market. They have a lot of companies, so they got to have a way to manage all those within their agencies. We're working with ASI to put together a front end, so a bundled engine, if you will, similar to how we have on our direct website, where we go into the quoting flow and you get both at the outcome. We're working to do the similar approach in the agency channel. We think that competitive price and the easy use drives a lot of business. Yes, we will also be adjusting commission on bundled business and also dependent upon volume, I believe. Commission does matter for sure. Competitive rates, easy use are big, big drivers as well.
To the point is, does the market need another one? We have to earn that right, and we know that. We have a right and we have a positioning with agents, and to be fair, the positioning of Progressive auto is a little narrower than we might like. ASI has a position with agents that is very positive relative to the preferred. In concert with putting those two together, we think not only we can have great product, but you saw some of the things that aren't just sort of for graphs here. We are easy to do business with. The number of times I've heard agents say, "Gee, I wish my preferred companies did claims like you do," or, "They had technology like you do." I'm sure they're meant to be compliments.
Unfortunately, I don't always take them that way because it's like, well, you're not my preferred company. I know what they mean. Now we can solve that gap. Whether or not there's a need, there is a place for us to compete, and I think compete well. Yeah. Oh. We'll get you.
Okay.
Paul Newsome with Sandler O'Neill Partners. The Allstate and Chubbs of the world have made a lot of conversation or suggestion that the claims process has to change as you move up towards the preferred carriers. In some cases suggesting it's just more expensive to manage a claim, if it's preferred. Do you agree with that? Does the process have to change as you move up the scale? Is there a cost differentiation as you move up towards more preferred customers versus less preferred?
I'm going to say no as a shorthand answer to that, and then fill it in. Remember, our customers don't just hit other non-standard customers. The idea of dealing with preferred customers in claims is not something that we do any differently than Chubb. Chubb customers hit non-standard customers, so on and so forth. We don't make our claims experience functional on the class of customer. We try to make the claim stand on its own. Things like service center are absolutely advantage for the preferred customer, mainly because they value their time a great deal as well. Time management is a big function in the claim. In all seriousness, I am not concerned about a service issue or a change in claims protocol because of the different customers that we will service.
To the extent that the individual claim might be more expensive, functional cars and so on and so forth, that's priced in. Really not worried about that at all, and we have many, many years of experience dealing with preferred customers from the claims perspective. There was one here, so we'll just take that.
Thanks. Two questions, if I can. The first, broadly speaking, one of the big themes that, I guess John didn't touch on today, was Google Direct or the possibility of a comparative rater in the direct market. I was hoping we could get some thoughts on that. Secondly, early on, I think Tricia talked about the success of the introduction of the new product, and it's clearly pretty dramatic, but it does tail off right near the end of the chart as well. I was wondering if there's something there, maybe a competitor response that's driving that.
I'll start with the second one. Why don't you take Google? You're a little familiar with that. Tail off, recognize it's just a spread of rates. It's not saying that we don't want those customers, it's just that we may have to price them a little bit more. Any time we make a model change, there's going to be a distribution of rates. There's two actions that are important. One is if we keep the model the same and just take a rate change, you'd expect it to be reasonably normally distributed or uniformly distributed. When we do a model change, that's where we get more variability. That one was just showing you that there are some SAMs who absolutely we can still advantage more so than we even have before. We want those because they will actually produce more lifetime earned premium for us.
We want the others as well, but just at the right price. That's really all that was saying. You want to hit the Google?
Yeah, we go to Google. We always say where, when, and how consumers want to buy. When there's a distributor and a carrier relationship, that where, when, and how is dependent upon both of those parties being profitable. We will distribute, generally speaking, where, when, and how we can make money and the distributor can make money. We take an account by account discretion and may take long-term strategic implications into account in considering who we work with. We were an early player in the aggregator marketplace. We actually were what I would call an angel investor, and I think the first that is now an insurance.com. We've worked with a lot of aggregators over time, and I will be frank in saying it's been a struggle I think for both carrier and distributor in that model.
Pat showed you results from some of those call center and online aggregators as well. We're getting better there. If you noted, there was a target combined ratio line on the slide he showed, and those bars are coming down, but they're not yet at that target. Where we can make money and where we think it's strategically the good thing to do long term, we'll play.
Greg Peters, Raymond James. Two questions. Just circling back, conceptually, I thought you did a good job talking about frequency. One of the comments or topic areas you didn't really touch upon in that secular trend of improvement was distracted driver issues. I thought maybe you'd follow up with some comments on that. I know you said, Glenn, at the beginning that you didn't really want to touch upon commercial auto, but I feel like I'm obliged to circle back because your performance in commercial auto has been even better in a GAAP performance basis than your personal auto relative to industry. Can you tell us a little bit about your business mix in commercial auto?
Can you also talk to us a little bit about the use of technology, event recorders, et cetera, that's helping to drive a better result, why we can't expect even a better growth rate out of that business?
Sure. Why don't I take the commercial one, distracted driver I'm not sure that we have a lot to say because we just don't get that data. I wish we did. We'll talk about it a little bit more. Obviously, I don't have great recall of what I said. It wasn't that I don't want to talk about commercial. It's just I only have two hours, actually I'd love to talk about commercial. The results are stellar right now. What's going on? Actually, John Barbagallo presented to us. He's here today. He presented to us last year really did a little bit of, hey, we missed a couple of things, we're the first to admit when we miss some things. What we really got is that pendulum, instead of swinging now, has really come in a nice center position.
Our business is still, in terms of units, primarily still local contractors, a lot of places where we're actually getting some real traction is the for-hire segment, especially sort of trucks and short haul or even longer haul trucking. Those segments we haven't always had huge amounts of data. We're growing our data there, as we're growing our data, we're becoming more confident. What we went through probably in the last couple of years was using the data we have trying to get to the right price point. I think we're much more confident now. We'll never be overconfident, we're much more confident we have that. What you're seeing in the marketplace is multiple things. There's no way to sort of quickly just dissect it, we have a rate that we're confident with, we're open.
There are others, specifically in the for-hire sector, that are not as open for business right now. It's a great place to be, where you actually have product people need product, others are not offering it at the same level. What you're seeing with our commercial is really pretty phenomenal sort of combination of growth profitability. No one, neither John nor I, are going to sit here say, expect those kind of results on a continuing basis. We've priced to a different target outcome, we've priced to a target outcome on profitability we would really love, we hope we can keep that growth going for quite some time. Market conditions are very favorable for us right now, both the economy a little bit of a pullback from some of our competition.
To follow up on the distracted driving, we don't do a lot of studies on that internally. We do work at times with some academic institutions who are at the forefront of analyzing technology in vehicles and transportation networks in general. I love a line from a very astute researcher in one of those institutions who effectively said, yeah, a lot more technology is out there, a lot more opportunities to distract, but at the same time, the desire to live has not diminished. Therefore people are still fairly attentive.
Actually, I think you're onto something. I wish, as our Snapshot work tells us, causal results are always better than correlated results. They just are. I'd love to have more insight into drivers who are putting on makeup as they're driving, or literally coming to work today, there was someone way across the center line, the eyes were down. I'm kind of like It's only that much, and it could end up with a very different result. Distracted driving is a big deal, and I'm not forecasting anything here. I hope that doesn't go up as cars start to do more for us. That might be an interesting interaction variable to study over time. I wish we had access to more data.
The only country I'm aware of that allowed any meaningful study was Australia, where they actually were able to match cell phone records to the best time of the accident and see if, in fact, there were a correlation between people being on the phone and the time of the accident, and the result was, yes, there is distracted driving, and yes, it causes frequency.
We've got one down here.
Yeah.
Oh, sure.
We'll get you a mic.
You should take a lap.
Oh, sure. I can read those.
Thanks. Brian Peery, Sansum Partners. Question for you, Glenn, and maybe for John Auer as well on ASI, which I understand to have historically targeted mispriced pockets of business in homeowners and pockets of inefficiency and what they are going to do, I think, as the exclusive partner in the agent channel is to quote any homeowner that comes into that channel. Those seem like different challenges to me. I'm wondering if you could talk a little bit about how different those challenges are and if you see that as a challenge to evolve what they are quoting on.
Okay. John, I'll get you to join in with the answer here. I'll give you my assessment of, certainly an assessment based on our taking a long look at ASI. Certainly when I joined, agents used to tell me what my target was. I was kind of like, "Well, actually, my target is really a lot broader than that. That just happens to be where we're competitive." Don't mistake the fact that ASI necessarily has a particularly strong or sweet spot in certain areas, let's say new construction. Yeah, it's a sweet spot. Best known in the marketplace. It doesn't mean to the exclusion of all else. We will not, Progressive, be trying to get ASI to do things that they don't necessarily think are the right things to do.
Their philosophy has always been very similar to ours, price every risk, but at the right rate. It's more an issue of competitiveness relative to market versus trying to be too niche player. John?
Well, actually, I mean, Glenn said exactly what I would say, the notion that we've tried to price segments is not true. We've tried to match the rate with the risk. Actually, when we started the company, we set an initial booklet that I wrote. In our vision statement, we were going to be the Progressive of residential property insurance. By that, it was primarily pricing. We felt that Progressive had made themselves progressive by pricing better than their competitors, and we wanted to do that in property. Back in 1997 when we were formed, started writing in 1998, property pricing was at the dark ages compared to auto. It's come a long ways in the 17 or so years that we've been in business.
I think we're a leader in pricing. We continue to be. We do end up being competitive in segments. I think that's where the market was overpriced. We tend to not be competitive in segments where the market was underpriced then. At this point, we just try to work hard to maintain our advantage in pricing. So far, we've done pretty good.
Hi, Ian Gutterman with BAM. Two, if I may. First, on the agency competition, where you're talking about as you raised prices a year ago, making it less competitive. On the same token, you guys obviously have the lowest expense ratio in the agency channel, right? If you're targeting 96 and you have an expense ratio advantage, someone's coming with a lower price, that means they're at a 99 or 102 or whatever it is. Is your sense that those competitors are happy being at 99 or 102, or they don't know it and we're going to see a problem down the road and that might be an opportunity? I guess that on the agency side. Frequency, since no one's brought it up yet.
Obviously, there's a lot of concern about two of your big competitors have reported a spike in frequency, seemingly blaming it on the economy as much as weather. A lot of the smaller regionals seem to have a bump for seemingly similar reasons. I understand your explanation about the uptick in miles driven being lower frequency miles. That makes sense to me. We're not seeing that at your competitors. Do you have any thoughts on why? I think in the Q you said frequency is, maybe it was 1% is pretty close to zero. Why you think you're not seeing what your competitors are seeing on frequency lately?
Yeah. I'll start with the first one. I don't always know what other people's targets are. I don't mean that as a dismissive answer. Yes, you will see occasionally that someone's going to write at a 98, 99. I don't know how happy they are with that or not. That's not something I can only make a statement about what Progressive will do for shareholders, go about doing it. My expectation is if we don't, we fix it. You know what our goals are. I will tell you something that sort of is not always so obvious, that if we have sort of a beta error, it's that we will take rate a little earlier than others. All right?
I think Pat even acknowledged that we took rate last year, and we saw what the results are, maybe it was a little more. Our error, if we take rate a little early and it doesn't pan out, is we'll be more profitable. The alternative You can fill in the blanks for the alternative, is you wait too late, and when you're at 97, 98, and I'm using absolute terms, and you've got a book of unearned premium, there is nothing that is going to make that unearned premium run off at some rate that's attractive. You go through this cycle of sort of up and down growth. We don't want to do that, what we talked about today is even though we've tried to make those rate changes small, we've got our own challenge now to make them even smaller and even faster.
The more that we can make price be a continuous variable, the better off we are.
Just to clarify one thing on that.
Yeah.
A few years back, there was an issue where, I can't remember how many years back now it was, where some of the state managers were getting maybe a little conservative about the 96, right? skewing towards one side of it.
Being more of a 94, you made an effort to fix that. Is that fixed or has that maybe become part of the issue again, is that that's maybe crept back in?
No. I think we will always sort of make trend assessments sometimes we'll be right on the money, not very often. Sometimes we'll be a little over, a little under. If we're a little under, we can take more bites at the apple. Product managers, by definition, are always going to be close or under their goal. Guess what? That's what we actually ask them to do. The aggregate works out quite nicely because for every one that's slightly under, we'll have one or two states that are not in that, and it makes the proposition to shareholders a lot more manageable when we say 96 aggregate. Your other issue was-
Frequency
the frequency. Yeah, that one is tough, I think it's tough for you to sort of sort through. We see what we see, our competitors see what they see.
Right.
I'll talk a little differently about severity, frequency, yes, you're right. In terms of bodily injury, UMBI or sort of physical or, excuse me, bodily damage type coverages, we're looking at a percent or two. That's great. That's low. I would say that's historically low. We will be careful not to bank on that, but that's what we're seeing. We're seeing specials, which are sort of the actual medical costs, actually come down. Now, part of what we do in claims adjustment is make sure we're not paying for specials that don't exist. There are lots of ICD9 codes and so on and so forth that don't apply to auto accidents that sometimes show up on the bills. You got to be able to make sure you're not paying those. Our special trend is actually in a very good place.
Generals tends to be somewhat of a function of the specials. Therefore, we're overall in a pretty good place. Severity, we're actually seeing severity also sit pretty low. I would say that's partly something we can be pretty proud of. We've worked on severity, but my best advice to you is just watch every month. Because these things, they almost defy logic at times. You sort of go from one quarter and then the next quarter it's slightly different than you expected. While I don't think in my career I'll actually figure out exactly why you can see different outcomes and different data sets, the one thing that we will do is we'll respond to our data as quickly as possible and not try to overstep. If we got slightly understepped, we'll have another step possible.
Do you have any other comments on frequency or severity?
The only comment I would make on frequency, the long-term trends are very, very clear, frequency and severity. Weather, we have come a long way in understanding, and Pat gave you a little more insight into how deep we now go to understand weather and its impact on our experience and using that information to price forward. While frequency was higher in first quarter than a normal first quarter, it was actually lower weather adjusted, we think, than 2014. Having that understanding, that level of understanding helped us avoid pricing to what I'll call noise or short-term things that shouldn't be priced in over the long term. I think that's an advantage for us.
I'm going to follow up. Bob Glasspiegel from Janney. I'm going to follow up a little bit on Ian's last question and Cesar on your closing comments that you said you've never been more optimistic about the prospects to drill down a little bit further on that. The question is, with commercial auto growing double digit, you say your key competitors, some of them are pulling back. With homeowners getting rolled out, which has faster organic growth, it seems like you're setting up to be able to be a double-digit grower again on the top line, which we haven't seen in quite some time. Maybe you could expand on the roadmap of why you're so incredibly optimistic right now, and whether double-digit growth is a theoretical possibility over some intermediate term outlook.
I'm looking to see how many outs you're giving me, Bob. Theoretical possibility. No, I appreciate your commentary. You're right. Commercial's in a great spot. I'm serious about my comment about optimism. That part of the business is working really well. Agent is not working really well. Do we have a plan? Yeah, we have a plan. Is it a sustainable plan? Yeah, it's a sustainable plan. We'll have new product, new rating, and we'll just keep that going. We won't fix it with one punch. This will not be a one punch and fix. The ASI opportunity is just when I refer to the white space, it's big white space. We keep acknowledging, we did it at least three or four times. We're pretty small, so you can't help but be optimistic.
If we do it half as well as we think we can do it's going to be a big growth opportunity. With the agency piece sort of holding us back I know we can fix that. That one, frankly, we've been and seen more times than we've seen anything else. We have a brand that's working really well for us. We don't have any operational holes. Whether it produces double digit, that's guidance that you know you're not going to get from me, right? We came awfully close in the first quarter if you give us a little discount for agency not working as well as we'd like. Frankly, our retention, those would be the two places that we're saying not good enough. They're not getting three checks on my physics teacher's little system. We'll fix those. We will fix them.
I mean, the auto CPI is bounced up about 150 basis points. It's not a perfect measure of price. With The Allstate Corporation and GEICO saying pricing is going up, how much price do you think will be roughly in your near-term element of revenues?
Is that what they're saying? I didn't know that. Okay. You know what? They have to do what they have to do. I put in my first quarter letter, frankly, any extended guidance, I know what you're looking for, and I want to be responsive. I'm just saying right now, with the frequencies we released, we've got the opportunity to keep a reasonable rate level and recognize it's 51 jurisdictions, when we start making summary statements, but a reasonable rate level, I'd certainly like to keep the momentum that we see in the direct channel. That'd be great to keep it going. It'd be great to give us a great start as we roll out the product in the agency arena and certainly with the ASI. If we get the benefit of that'll be terrific.
My but is if we see some trends that force us to take rates up, we'll do that, too. At least for the far forward as I can see, say two or three months, it looks pretty good for us.
Up here.
Yes, we'll take one from here. I'd like to follow up on a discussion about the many new cars having additional frequency-reducing technologies. Is Progressive finding that customers driving these newer vehicles are materially more profitable? If yes, is this because Progressive has been conservative when estimating the benefits of such technologies? John was the one that did this. I think the real answer is we try to price in the advantages to get more of that type of business. It's not a question of profitability. We try to hold the profitability somewhat constant and create a price differential that makes us more attractive.
You said it well. Do it as fast as possible.
Okay. More important. Let's take another one from here while we're reading. Progressive's first Q personal auto combined ratio is meaningfully better than Allstate brand and GEICO. What do you attribute this to? Is the gap sustainable? I have to be serious in my answer, right? I've never found it to be particularly useful to talk about how we contrast with someone else. We see our data, we report it to you as we see it. That's the best answer I can give you. I just answered Bob's question that we think we can maintain a reasonably low inflation rate on our rates for at least as far forward as we see. If they're seeing an opportunity that forces them to take rates up, maybe they're doing something not too dissimilar to what we did.
It's a timing difference, I'm glad that we will always have that bias towards a slightly early timing in the marketplace. It does come back to pay us off. Maybe this is the time. Stay tuned. You see it every month, same as we do.
Scott.
Glenn, John, if you allow me to take a look at the restaurant industry relative to the insurance industry, compare what the restaurants do relative to what you're attempting to do. McDonald's, when it started out, they were doing very well with a simple hamburger, cheeseburger, a milkshake, and fries. Growth slowed. They introduced the Big Mac. Growth slowed. They introduced breakfast, so on and so forth. Can you help us understand what you're doing means to 96 and grow as fast as you can?
Sure. It means literally that will be preserved at all costs. I'm not sure I can continue the analogy with McDonald's, also I want to reference something that was said a couple of times. We're not taking our eye off any part of the business that we over-index in. The skills and talents that we have are very transferable to another part of the segment. It's a little bit less about a taste or a preference item. As far as the 96 goes for consumers or for shareholders, that is an aggregate goal that I can sit here and say, I will be responsive to any situations that differ from that. That is the goal. There is no change.
To perhaps to Bob's question of a little bit my own personal excitement, I want to be careful of how I state this because we're not doing anything yet. If we have an ASI that produces over any reasonable period of time, a very attractive combined ratio, you have Progressive with a very attractive combined ratio. As we start to look at the interactions in the household, we start to put data sets together that we've never had before and find some of those pure premium kind of opportunities. Frankly, I think not only can we produce a nice margin, but we can continue to produce an even better and more attractively priced margin. 96 is sacrosanct.
Glenn.
Sorry.
Gary Ransom from Dowling. I had a question on
Thinking about growth, profitability, and then throwing in PLE into the equation. If you succeed in getting more Robinsons and you suddenly have an increase in loss costs that requires rate, and maybe it's a disruptive rate increase. You had that as recently as 2012. There's suddenly a different equation because now maybe with this Robinson group, you can decide to either have a 97 for one year and then five more years of 96, or raise rates today to keep your 96 but lose half the customers. There's a different dynamic that's involved there, and I guess the question is whether you are willing to tolerate that as you build your book of Robinsons.
We have job openings. You're exactly right. To Harry's question, the 96, and I would ask you to reread the definition of 96 in the annual report that we put out. It is the amalgam of all parts of our business. Our special lines and our commercial business don't all run at exactly the same targets. Our new business doesn't run at the same target as our renewal business. That will be another piece in this sort of amalgam, and there may be times where one sector, which won't be reportable and visible to you, where we say 96 and a half is more than acceptable relative to our outlook. We won't do it on the comm, however.
We will have to sort of prove to ourselves that policy life expectancy is real, that it is absolutely real, and that it doesn't take us in aggregate out of whack relative to our 96%. The notion of a household is a very important notion, and as I said, decades. It is reasonable for us to expect that we can have more customers, we have some, who have been with us decades. Does that require a slightly different calculus in the understanding of how to change the rate and so on and so forth? Yeah. Our business is getting more complicated. Our goal is not changing. By the way, just since I saw you take the microphone, we're not doing a lot right now, but we're working with Rand McNally on measured usage inside of trucks. We actually have a Snapshot semi-equivalent in the commercial mode.
I remember that was part of your question.
John, you talked about the homeowners' business getting to be better in the aggregate. What gives you the confidence that it is getting better on a secular basis, and you're not just buying into peak profitability on a cyclical basis?
I would remind that by far more important in our decision to take a majority interest in ARX was ARX in and of itself. We think the fundamentals are somewhat different, and that, we think, is evidenced by the premium trends that I mentioned. I acknowledge fully that the underlying loss costs in both those segments have differed over time as well. We have seen one industry grow materially in terms of premium and the other stay relatively flat. We think there's great opportunity with ARX. They've performed extremely well throughout their history. Recognize the core objective there is that auto PLE and getting more of those Robinsons because we think of it not only in terms of the outcomes at ARX, we think of it as the outcome of the whole.
We know those auto PLEs are going to go up and fill that white space of the Robinsons we don't have today. I'd also suggest to you that I cannot prove this per se, but certainly some discussions would bear it out, but I think regulators become more interested in having homeowners priced appropriately. They saw some of the negative effects of having it sort of underpriced and ultimately catch up, and obviously, there are some states more notable than others. I think regulators are much more open to getting the right rate for homeowners than they have perhaps than 25 years ago or even 10 years ago.
Hi, James McLickie with Citi. I have two questions. My first is on the agency side. The policies that you're losing, is it a case where the business isn't as profitable as the business that you're retaining, and you actually don't mind to see that business go, but you try to raise price, and they go to a competitor? Is there something else at work there? My second question is on the development. In 2014 was negative. This year to date, it's positive. What are you seeing there that's driving that difference?
You want to split the question and I'll take the agency piece that you're developing? We price every customer at a level that we feel comfortable with. If we need to adjust that, there's a reason for it. No, I don't like losing a customer at any level in the company. On the other hand, I'm not happy to retain a customer at the wrong price. There are really two dynamics going on. One is we make an adjustment to a price, and the customer may or may not like it, but the second dynamic is there is somewhere else for them to go. If in fact, that's the case and we're right with our assessment of price, then it's sort of an adverse selection for others, and ultimately that pendulum swings back as slowly as it might swing.
We're not happy to see people leave, but the first priority is to make sure we got the right rate for all customers, as best as we can determine that right rate. When you saw Tricia talk about rate stability, and it's fair to say that whenever you have two models or two algorithms for pricing, this is an inexact science. It might produce one answer and another answer later as you change your model. We will try to find ways to sort of.
Transition from one rate to another. Our point estimates, neither might be right, but we think we should be somewhere in the middle. That's really the essence of rate stability. We'll try to do that as much as we can for customers as well. If it truly is at the outer end of that distribution and they have an alternative, we understand that. We don't like it, we also wouldn't be prepared to write it at a lower price.
Is that the majority of what's going on with the losses? Those are on the farther out of the spectrum?
Well, by definition, if we're taking the prices up on that sector, yeah, we're uncomfortable with the loss ratio that it's been producing.
Got you.
On the development side, we have had favorable development through first quarter, and we had unfavorable last quarter, or 2014. We always try to have our loss reserves as accurate as possible, as soon as possible. Our actuarial group is almost like our pricing group, and they're always looking for ways to get better, more accurate estimates of those claims as soon as possible. I'd also tell you that we work very hard to be transparent in our methodologies, if you haven't read it, we put out a summary of our methodologies, I think around the summer. June, July, or August. Take a look at that. Through the first quarter, it's driven predominantly by severity being lower than we expected in our bodily injury coverages. We report incurred severity for BI was down 1% for the first quarter.
We obviously look at paid severity as well. That was up about one point, but we priced in severity expectations that it were higher than that. Which I think is a very reasonable expectation if you look at the long-term trends, and I think of it as a good problem to have. If we are outperforming our severity expectations, we'll look to adjust over time. Minus one or plus one is probably not a reasonable long-term expectation for BI severity. That's the underlying story. It's unfortunately almost time, and I want to make sure that we don't miss a question that has come through on the web. From David Small. The real question, I'll summarize, it's quite long, is really sort of you talked about comparative raters, so on and so forth, but what about Google? Bottom line is it's Google. Which is different.
We've seen comparative raters in all different domains. We talked about insurance.com and so on and so forth. While we believe that comparative raters, including Google, will attract certain clients or carriers that perhaps are not as comfortable with their own marketing, they will do that. You'll get the results of an auction environment very clearly. We're very conscious of Google. It's only in California. We have really no great update on their results. We're not seeing it in our results, in search or anything like that at this point in time. It's a fair question. The fact is, it is Google, which makes it different. How different? I don't know. Google obviously has to wrestle with a lot of problems. One is we're a big buyer of search and so on and so forth. It's hard to be sort of both a competitor and a friend.
There's lots of things going on there, at least to my knowledge, I have not seen any major movement by the top 50% or so of the market share. In some ways, it's a little bit of the best of the rest type of comparison, I'm not being awkward with that statement. I'm just saying that's sort of what we're seeing right now. I hope the consumers will understand, while it might be a very different opportunity to get one more look in, that they don't overlook what is clearly a very viable marketplace in other places through agents and through Progressive direct and GEICO and so on and so forth. I just can't imagine that consumers could sort of overlook those. That's probably a better question for next year when we actually see it play out a little bit more.
I've been given this sort of cut sign, so in any language, I know what that means. I really appreciate you showing interest in our company. I mean that sincerely, and as I said earlier, let's talk again next year. Thanks.