Glenn M. Renwick - Chief Executive Officer, President, Director and Member of Executive Committee
Susan Patricia Griffith - President of Claims Group
M. Jeffrey Charney - Chief Marketing Officer
John A. Barbagallo - President of Commercial Lines Group
Brian C. Domeck - Chief Financial Officer and Vice President
John P. Sauerland - President of Personal Lines Group
William M. Cody - Chief Investment Officer
Progressive Corporation (PGR) Investor Relations Meeting June 14, 2012 9:00 AM ET
Glenn M. Renwick
Good morning, everybody. It's actually great to be here again. Thanks for coming. Before we get started, the Safe Harbor, most of you are familiar with it. It's either in your books, or maybe on the screen right now. So let's respect that during the course of the day.
This year, we're going to change the format up just a little bit, partly to -- because we always want to change the format just a little bit. So we're going to actually start asking the questions. And Brian and I get the chance to talk to you during conference calls, during the course of the year, so we have a pretty good idea of what's on your mind. And frankly, we reached out to a good number of you to sort of give us some questions, give us a sense of what you really want to hear about in this session.
So a combination of those things. What I've done is asked one of my Senior Managers to lead 4 sections of real interest and use some colleagues to actually help them in that area. The bios of those colleagues are in your book. I'm going to refer to them by name. So certainly feel free to look towards the end of the book, and you'll get the sense of what they do in their regular day job.
But using a question-and-answer format, we hope -- we clearly won't get to everything that's going on in the company, but we hope it'll give a little more insight. As I've often said on conference calls, "That's a great question, I wish I had more time to develop the question, to give you a little bit more of a data insight to it." This is the format that we choose to use to do that rather than reformatting information that we've already given you. So that's the opportunity. That's what we're going to do today, and hopefully you'll get a lot of your questions resolved.
I'm going to take you through sort of the agenda, more importantly, why we chose these particular topics. And then Brian and I will come back for a Q&A session of your choosing at the end.
But let's start a little bit with the agenda, and the first element is Snapshot. I've heard a few of you were interested in that, so a good reason to start with that. I've used some pretty bold language to describe my feelings about what Snapshot will mean to the industry. And while I have no doubt or concern that alternative views exists, I consider this to be one of the most important things I've personally seen in my carrier. So today, we're going to sort of pose some questions.
What do we know now that we didn't know this time last year when we talked to you? How does the economics of deeper discounting really work? And perhaps the most important question that's on my mind is how do we take something very powerful, this very powerful underwriting notion and extend the addressable audience? So it's not just shoppers but an audience that's larger than shoppers that have some exposure to this concept and to what we can do for them. John Sauerland and Tom Hollyer, both with us last year, and their team have done some terrific work this year. So they're actually going to post those questions and answer them.
So that'll be at least a good introduction and a good update on what's happening with Snapshot. Now as I have a slide going behind my head, I am sure that there's more than one of you are pondering just what is that meaningless mess behind my head? Well, frankly, it's not a meaningless mess. And I admit that this is a little more staged just simply for this presentation, for privacy and some other things, but think of this as real time. I've invited a few thousand -- tens of thousands of our customers to join us today. And these are our customers. Arrayed by speed, and now we're slowing it down a little bit. These are customers that are actually traveling at about 25 miles an hour right now. And if I come in just a little closer, I can see that we got one vehicle here, happens to be a Ford Fusion, happens to be in New Orleans. Today, I happen to know, this vehicle has taken 2 trips: one, highly likely, a commute to work; the second, very short trip, coffee and beignets, perhaps. Two insurable events. Interesting. On track for about a 7% discount. This week, the individual is doing about 14% less miles than their weekly norm. Maybe they took a day off. It doesn't mean much on an individual basis, but pretty interesting when you can extrapolate those statistics. Happens to be one of our promoters on the last Net Promoter Score, 4 or 5 in this case, right?
Here's what I want you to take away from that data set, interesting, little fun. But just think for a moment, at the vastness, the volume and, frankly, the velocity of the data that we now have available to us. This is a long way from age, sex, marital status and points. Most of us by now have heard the term "big data," and new professions, data scientists, unstructured data, [indiscernible] data sets, where algorithms run on top of the data, looking for and testing hypotheses.
This is the only real way to manage these massive, massive, massive data sets. And frankly, this is a new skill, a new skill that we must be extremely confident in. And while it's not important to me how you see this, I see it as a real skill that insurance companies must have to sort of be in the future of auto rating. This doesn't lend itself to basic drop it down to a spreadsheet and do some relatively simple or rudimentary or even advanced statistics. This is very, very big stuff.
The second thing to take away from this is the relatively low technology barrier to getting the data. We happen to use a device today. We've got other means of achieving the same ends. But ultimately, the car is going to talk to us. We know that. And anyone who wants to listen will have access to this data. So it's all about how we use the data. So what I want to set up with this commentary is just a sense of the realism of the world that we're living in now from a data perspective versus what we've seen in the past. And the real question is what can we use that for, for market advantage. I think you're going to see a good first cut at that today.
Retention. I have consistently used this format and many inside Progressive to keep focused on something that I consider to be the most important element that stands between Progressive being all that I can envision it being and the task that we need to do to get there. And retention is top of mind for that and always will be top of mind for that. We've had the opportunity to take you through some of the process improvements that we've made. We've taken you through some of the measurements that we've made. We come into this meeting and are fully well aware that some of our -- specifically our Direct policy life expectancy, somewhat disappointing negative trends. You could reasonably imagine we've done a lot of diagnostics on that. A couple of corrections that we've had to take in a lot of states with relatively large rate revisions, which I don't like, I like smaller bites of the apple, have absolutely been the primary contributor to that. The time lag is not so easy to sort of place on that, but I think we're in a better position now and hopefully you'll stay tuned to those results, but I think you'll see that those will likely improve during the course of the year as we put those rates -- necessary rate changes behind us.
Tricia Griffith and colleagues Pat Callahan and Mike Sieger are going to continue their story of retention today. And more importantly, they're going to focus on current customers and things that we can do to current customers, whether through experience, or multiple products, and what their experiences in multiple products will do for them in the future and perhaps, give you a sense of what I'll call a reason to believe why Progressive is positioned today to do things that we weren't positioned to do only a few years ago. So a very powerful session.
I was looking back at my notes to the first one of these that I got the opportunity to do in 2000. And one of the topics in 2000 was, will the Internet really be important in insurance? And so generally, it was really an open question at that time for some. I know it seems comical now, but we were 5 years post our first website, 2.5 years post our first sale on the Internet, but still, we were asking some of the right questions and some maybe not so right. The fact is we never hedged our bet on the Internet. Now we have a very second opportunity with mobile, and we're not going to hedge our bet on mobile either. We can all debate exactly what the usage and adoption rate will be and exactly what impact it might have, but we are going to try to take you through today, Jeff Charney and Matt Lehman, are going to take you through our assessment of mobile, its usage inside and outside insurance and how we're adapting to use mobile to our advantage. I would tell you that in combination with the powerful underwriting and segmentation scheme, mobile can be very powerful. What I've asked Jeff to do is try to make this a little more tangible. So we're going to show you a little bit of a demonstration of something I've already written about in how we use the camera to affect a quote, and what the future of quoting can be when we've this pocket technology that all of us have become so dependent on. So I think you'll get a real sense of just our investment in mobile and where this could lead for us.
The fourth section is going to be led by John Barbagallo and Karen Barone. And what I've asked John to do is something that's a little different than we normally would do in a meeting like this. This meeting sets itself up very nicely to talk about things like Snapshot and retention and mobile. But frankly, none of those things matter unless the factory is working really well. So while I'm kindly referring to the back office as our factory, there's lots and lots of things that have to work to make any of these plans viable and be able to keep up with the kind of growth that we expect. So John's going to take you through some investments we're making in the factory that actually allow us to maintain our brand quality, 25,000, 26,000 people, maintaining your brand experience is critical. How can we maintain? Frankly, improved quality. How can we make speed a key -- continue to be a key element and a hallmark at Progressive, especially when we have clear rate opportunities to do rate revisions and needs for those? How do we give them to market very quickly? Hopefully, you've noticed that our Agency business has been very strong this year, very proud of that, very pleased with that. What Karen is going to take us through is an investment and our agents to try to help them maintain relevancy in the same world that we're all living in, primarily by exposing them to tools that leverage our experience in the digital world. So things we've learned from marketing in the digital world, how can we package that up to make it relevant for our agents?
One initiative that John is not going to talk about that I'll take 2 seconds on is the replacement of our core processing system. Not a particularly big topic that we normally talk about in a meeting like this, but I guess there’s a certain amount of passion for me, having sort of been both the CIO and the CEO, and replacing a core processing system. It's huge. I can say today that we have our core processing replacement system up, operating, processing policies in 8 states, more by the end of the year, it is meeting user functionality, it is using all expectations, it has better embedded quality measures and, probably, systems we've ever put forth before. And frankly, I think a lot of people at Progressive would be very happy if I just stopped there. The fact is it's been unbelievably consumptive. Doing things at the heart of the business and changing sort of the rails as the train is moving is very hard. I now have a very much a personal appreciation for just how good companies and good people sometimes fail to change those most fundamental systems, even if they're good at building around the outside. The exposure has been sort of changing for me, and I have to tell you, it is unbelievably delightful to know that Progressive is being set up that we won't be talking about a legacy system that's holding us back 5 years from now. So we're very happy with our progress there, more to come.
So I hope that, that certainly will give us a flavor on some of the key issues. Now for those of you who are a little bit more meat-and-potatoes types, we're going to actually start the question-and-answer session with a little more formal questions that we posed to ourselves, but they really should be your questions. We derived them from you. So I've asked Brian Domeck to come up and do a lightning round, so give us a quick reconciliation between accident year and calendar year on this year, last year. What's going on with development? Give us some insight there. What's happening in our top states? Give us some color on the top states, and a few other questions in that mode. Now before I give up the stage, I'd like to at least take us through -- I think I got behind a little on my clicker here -- just a few things that, again, don't necessarily hit the radar screen all the time. But I'll just give you my impression of what's on my top list of things to sort of worry about or to be happy with. And on the positive side, organization. We can have all the best ideas in the world. The question is, do we have the organization in place to be able to implement them?
And I would tell you that -- of course, I'm biased, but we have a very, very strong organization. And I see no holes whatsoever in our ability to implement everything we're going to talk about today. That's not just from people at my direct report group. I think they would say that all the way down the organizations, we feel very good about our organization, and that is a major issue. If you don't have the organization, you really don't have the ability to implement.
Product rollout. Well, Snapshot might be the first thing that comes to your mind. Hopefully, over the years, you've seen us with our genie curves, always telling you, "We're pushing out, we're pushing out the science by continuing to evolve our product." We don't tend to come to the market with major product releases and names and numbers. We're more like painting the Golden Gate Bridge. Start at one in keep going, keep going, keep going. We're finding right now a very receptive environment to be able to roll out all our new ideas, and our product rollouts have really been very successful. Of course, Snapshot is primary amongst them. You'll hear more about that.
Claims performance. I've had the opportunity to say, on numerous occasions our claims performance is at all-time highs, knowing full well I would have to stop saying that at some point when marginal improvements flattened out. This is not the year. Not only is our claims organization in as good a shape as it's been in, certainly, recent history, our performance is extremely strong. So again, it's no point talking about some of these things if some of the back office is not working or claims is not working. Claims is working beautifully right now. You heard last year from Tricia that several elements of our claims organization are at all-time highs. Most of those numbers are now higher than they were at that point.
Premium growth. This shouldn't come as a surprise to anyone. Grow as fast as we can at a 96. That has not changed in many, many years. We're starting to see some premium growth. I'd love to see a few double-digits, just a personal sort of desire for me to see that again. But high single-digits is actually working out very nicely, so I like the growth we're seeing, clearly. Won't be any surprise, certainly, after yesterday's number to recognize that we've got a little bit more premium growth to be put on top of that. So average premium starting to go up, something we haven't seen for a long time. So I like the growth that we're starting to see, both, obviously, in units, which that I care most about, but also in some rate.
Capital position, Brian's got that on his lighting round. So I'll let him go into more detail on that. Just suffice to say, a very solid capital position. No constraints with anything that we want to do.
Progressive Home Advantage. I think the key point I'd like to make here is this is a viable consumer proposition. Consumers are very comfortable coming to Progressive, attracted by the brand of Progressive and buying more of their product needs from Progressive. More importantly, those customers who come to us and find their needs changing over time feel much more comfortable coming back and saying, "Now my needs have changed. I need a homeowner's." Is this a viable business model? Is it a viable consumer model? My answer to both is yes.
I actually have a couple of announcements that I'll make at the beginning of the Q&A session, but -- and one of them involves some strengthening of relationships in PHA. So I couldn't be more happy with those elements. Obviously, there's always the counter side. Managing severity. I mean, severity is clearly going up. We're not the only ones noticing that. We're seeing some variability and frequency. I said -- apparently it caught some attention -- these are dangerous times. I'll say it again. When things change from sort of negative to positive or start to get a greater acceleration, it is difficult to get the timing right. That's our job. We know what our job is. But managing changing environments with frequency and severity, it makes it that much more challenging to walk along our 96 goal and be there all the time. We know how to do that. I don't like missing any months, any years, any period at all, but we do know how to do that. So comfortable with that, but actually there's some positives with the increasing premiums.
Challenges to income. This is no news here at all to anyone. It's pretty hard right now to sort of see the classic insurance model of making an underwriting profit of X and adding to it with an investment income of Y. Y is depressed right now. Understand that, wish it wasn't, it is. We're going to stay the course. We know where our franchise really is. We're going to continue to stay the course and create a great, great book of underwriting and so that you'll all know as soon as we know when investment returns, perhaps, get a little more healthy.
I wouldn't normally put this on the list, but since I mentioned PolicyPro, couple of years, we'll still have a lot of resource consumption in getting PolicyPro over the goal line. The key here is we're not going to lose our discipline. We're going to keep the discipline to make sure that we don't get 90% of the way complete and then sort of always have that as a lagging issue. We will take this all the way to the goal line and be very proud that that's something that Progressive can build on for a long time to come.
And you'll hear a little bit about this today, the concierge claims service. Again, last year, you heard that our delivery of claims through our concierge centers is absolutely the best on all 4 measures that we care about. My frustration, and it probably has become a frustration, is not enough of our consumers get to use it. So we've got to double down on our ability to market this and make that proposition something that people absolutely see and want to use because our capacity is more than available. Even though we're meeting break-even targets or better, we have capacity to be able to provide this. And it's not just my assessment that consumers want this. This is what the consumers tell us they want out of the claims experience. So we've got some real challenges to turn something that's working very well into something that's working extremely well, and that's a challenge that we intend to step up to.
So let's get started with the first meeting topic, Snapshot. I'll turn it over to John Sauerland and Tom Hollyer.
Susan Patricia Griffith
Good morning. It's good to see all of you again. Last year, we met. As you recall, we did a decade in review. And when we met, Glenn said if you had to have a mantra for the last decade, it would have to be retention. In fact, his best guess would be that it would continue to be retention. So Pat and Mike and I are going to talk today about retention. But we're not just going to give an update. You see those in our monthly reporting, in the conference calls. We're going to talk about the next level of science around retention. In fact, I think what you'll hear today will have been through almost every one of the presentations is the science of retention.
So John and Tom just talked about Snapshot. Snapshot gives our customers this incredible ability to say, "How's my driving? I want you to know how my driving is." And when it's good, and they get a discount, they stay. I personally am very excited about our ability to now have consumers test drive Snapshot. Glenn opened by talking about an addressable audience. In claims, we have an incredible set of addressable people that love our service and say, "I'd love to come to Progressive. How do I get there?" And a lot of them convert immediately. But some, I think, along the way, maybe forget about that emotional connection. And when their renewal comes, they don't necessarily think about changing carriers. This gives us a perfect time to say, "Try us out right now." They already know the outcome of what they're buying, because they've had the service. What a great opportunity. So I'm very excited about our ability to test drive Snapshot.
So the question that the 3 of us asked ourselves was this: What are you doing to make Progressive a destination insurer for more and more customers? And the answer is simply a lot. In fact, we continue to focus on the nature of our customer, so the mix of our business, and Pat is going to go into more detail around that, and the nurture of our business, caring for our customers, what can we do to continue to give them a reason to stay? And Mike is going to talk about that.
While what we've done in the past we would consider sort of our undergraduate work, so today you're going to see our next level of study. What you're seeing behind me build is our retention curve for the last 12 years. We've made a lot of progress with retention. We know we have a lot of work to do. Above the curve line, you'll see a lot of the investments that we've made. I'm not going to go into a lot of them. You'll hear some of them today. A lot of the stuff isn't necessarily new to you, but it's critically important to get a firm understanding of how embedded retention is in our culture.
Let me go into one example we call CMT, it's Consumer Marketing Tiers. In 2008, we rolled out some consumer marketing tiers and we introduced you to the Robinsons [ph]. They were in our advance segment, that's an internal word we use, bundled, continuous insured, bundled home and auto, continuous insured, they have some other products, so they have a boat and an umbrella policy. We know them well, we continue to grow this book of business.
Now we continue to have to evolve with them as well. They are perfect Snapshot customers. They tried it out 6 months ago. Yes, they got a discount, and they're staying with us. We love that. We're going to continue to evolve as the Robinsons [ph] evolve. And in fact, their son, David, at the time was 19. Four years later, he's 23, he's graduated from college, and he's now in what we call our Choice segment. That Choice segment is very influenced by what their parent says. So he's always known Progressive. So he gets Progressive. He has an apartment, he is a renter, and we're going to continue to evolve with him. He just recently bought a motorcycle to tool around with on the weekends, and we're going to be there for that additional product as well.
What Dave loves is that he can handle all his insurance needs on his smartphone. Different than what his parents love. So this is really important for us to understand and look at our tiers, put them into smaller segments, and when we find meaningful differences, that will guide our marketing insight.
Let's go below the retention line, and you'll see some of the measurements that we've been using at Progressive. PLE I think you're all very familiar with, we talked to you about this over a decade ago, policy life expectancy. This is just the number of months on average that we expect our policyholders to remain a customer with us. And we slice this into many different segments, by the channel with which you purchase, your product, your tier. We use this very powerful metric, we'll continue to use this. In 2006, as a company, we rolled out NPS, or Net Promoter Score. While not a Progressive measure, very important measure, because these are people who have had an experience with us and they stay with us. It's been a positive experience. In fact, our promoters recommend Progressive to friends or colleagues. So having that NPS measure with the customer satisfaction which is very correlated to retention as well as that PLE measure has been a nice complement to each other.
Today, I'd like to introduce you to something we've been using internally that we call future PLE. It's taking that original PLE and then given actions or experiences our customers have, we understand a little bit more about them. For you stats people, think of it as conditional probabilities. Let me walk you through a very simple example. So this is an auto-only decay curve. These are auto-only customers that we don't necessarily know a lot about when they first come on the books. We know a little bit about their past driving history, a little bit about their prior carrier and their prior coverage. But just like any relationship, as that evolves, we get to know more about them so we can serve their needs better.
So let's take a subset of these customers that have been here 24 months. As you can see, their PLE is longer than the average auto PLE. But in order to get that example and kind of understand the conditions that we're trying to get at, let's take that auto-only subset and start that clock ticking at that 24-month period. This is where we're going to get the insights. In this orange area under the curve is where we'll start to understand the conditions that have happened with both of those cohorts to move them both up. So was there an accident involved? Did they change their policy? What happened so we can continue to move both of those lines up and out. Let's take it one step further. Customers given the fact that they've been here 24 months and sometime within that 24 months purchased another policy. You guessed it. Their future PLE is longer than the other 2 cohorts. Again, our focus is going to be on that orange area under the curve to continue to understand what makes those curves shift upward.
So what we're going to do is we're going to take a immense amount of data, observe those conditions and then react. React with better product offerings and better service. So Pat, we know multi-product penetration increases retention. So what are you doing to get more customers and more policies together to increase that stickiness?
So Tricia introduced some aggregate policy life expectancy data, but we know all customers aren't created equal, especially from a product needs perspective. So what I want to demonstrate to you this morning is how we're leveraging our expertise in customer segmentation and applying that to building out our product portfolio.
If we start with the basic framework that highlights 2 of the key retention metrics that Tricia mentioned, policy life expectancy and Net Promoter Score, and apply that framework to a simple segmentation of our auto book of business based on the products owned and some other demographic data. We'll start with a cut of our auto-only customers. We call this monoline auto customers, segmented by whether or not these customers own a home. We immediately see that there's a difference, and those that own their home have a higher policy life expectancy and Net Promoter Score. Not surprising in a least bit. If we take a next look at additional customer segments who not only insure autos with us but also have one additional Progressive product. It might be a motorcycle, it might be a snowmobile or ATV, it could be a renter's policy. What you're seeing on the screen is, not surprisingly, higher product ownership, greater products owned drives up policy life expectancy and Net Promoter Score. What we'll look at next, though, is a slice of our customers segmented by those that not only have auto with us but also insure their property through our Progressive Home Advantage program and have at least 1 additional product plus some demographic slices.
Three things that I want you to note from this overall segmentation. First, solid correlation between policy life expectancy and Net Promoter Score. So while we monitor, measure and act on Net Promoter Score and make investments to move that, we're investing to improve our policy life expectancy. Second, you're going to see that meaningful separation between those customers on the upper right who have multiple products and bundled property and auto with us versus even those that have 2 products with us and obviously those monoline auto customers.
The third piece that's important is the bubble size on this chart represents the size of that segment within our current auto book of business. So 2 things jump out right away. number one, obviously, we're highly-concentrated in the lower left with monoline auto customers, not surprising given our heritage as an auto insurer. But secondarily, when you see in those segments that retain a long time, we don't have a great share of our auto book of business in those segments today. But what's important to note is that those segments represent a significant portion of the U.S. private passenger auto market, and that's the opportunity that we're excited about, finding out how we're going to grow those segments within our current book of business.
We'll come back to those long-retaining customers in a minute. But first, I want to talk a little bit about what we're doing today for those large segments of customers who we know only have one product with us, an auto product, and what we're doing to move them up the product ownership hierarchy. So we know that virtually all of this monoline auto customers either have or will have a need for property insurance at some point in their lives. And we also know from our experience to date in the PHA program that when they bundle that property with their auto, retention goes up dramatically. So first, we look at non-homeowners. And you'll see that those who bundle the renter's with auto through our Progressive Home Advantage program retain significantly longer. You'll see it on the Y axis. You see this with homeowners as well. While homeowners in general, as we discussed, retain longer than non-homeowners, when they bundle that homeowners with us, retention goes up a lot.
So recognizing that we have large segments of our customers who have a need for property insurance and we have experienced bundling their property insurance with their auto, driving up auto policy life expectancy, it's imperative that, as Glenn mentioned, we continue to invest in having a viable property solution to meet these customers' needs.
A viable property solution has 2 key components. Number one, supply our capacity to meet customer demand; and number two, a demonstrated consumer appeal, the ability to get people to be aware of the program and to take action once the program exists.
We'll start with supply. John showed you this graph last year, which is highlighting our PHA underwriting partners by state. This map has evolved in the last 12 months. In fact, we've added 2 additional carriers to the program, bringing to a total of 6 the number of partners underwriting property insurance for our auto customers. These carriers, in aggregate, beat property industry loss ratios last year, so we feel pretty good about the reliability of their rating models. And what we're seeing from a country-wide perspective is that they're asking for more of our business. All 3 of these give a solid indication that the supply or capacity within our property partners will remain strong going forward.
Now all the supply of property insurance in the world will not drive auto retention unless our customers know about it and take action on it. And that's why we continue to invest in marketing campaigns that are driving not only awareness of the economic and ease-of-use benefits of the Progressive Home Advantage Property program, but also highlighting that through all of our customer communication. So what you see on your screen is direct mail, you'll see box slips that we do in customer communication, newsletters. We do audio messaging on all of our customer service inbound lines, where Flo talks to customers about the benefit of the Progressive Home Advantage program. And additionally, we're dedicating some of our mass media television advertising that we know reaches both current customers and future Progressive customers with the message that we can be their destination insurer and the benefits of bundling property and auto.
Now our entire investment in the property insurance space is intended to drive top line premium growth by increasing auto retention. And the key success measure of that is are we penetrating our auto book with more property insurance? The answer to that is yes. Over the past couple of years, we've increased penetration of our auto book of business with property insurance threefold. The majority of the property insurance we're selling through the PHA program is to current customers. And at present, about 12% of our auto customers have an additional product insured with Progressive. Now what we're equally excited about is what we're seeing in the direct side of our auto business. We're seeing significant growth in direct auto renter bundles taking place. And we're excited about that for a couple of reasons. First, you saw the renter versus non-renter bundler graph a couple of slides ago, and that's showing you significant separation between those non-homeowners who bundle with us and those that don't. The second thing that gets us really excited here is that we see the renter/auto bundlers converting to home/auto bundlers at twice the rate of those that don't have renter's bundled with us.
So when Glenn mentioned kind of being there for the life events that are our customers experience, this is a classic example of that. Now as we see that growth, what we're doing is continuing to invest in building that renter/auto bundle within our direct book of business today, because we believe that's building a bundled home/auto pipeline of future customers for tomorrow. We're excited about that, we'll continue to invest there, and you'll hear more about that from us throughout the year.
Now let's shift gears to take just a minute to talk about those long-retaining guys in the upper right-hand side, the 3-plus product customers, and what we're doing to attract and grow those organically within our book of business. We know these customers aren't easy to move from one carrier to another. They stay with us a long time when they come to us, but we also know that, that means they're sticky with their current carrier. They don't shop a lot, they don't go through the life events that drive insurance shopping that frequently, so they're tough to move. But we also know that the combination of the products they own and their long PLEs make them the highest customer lifetime value segment of anything we're going after. And the combination of that customer lifetime value plus the fact that we're underrepresented in those segments, makes it imperative that we continue to go after growing our share of those segments. We're attacking it really in 2 areas. First, we're working to attract more of the multi-vehicle owning households to Progressive by leveraging our full suite of leading products.
And second, as America's leading insurer of motorcycles, #1 insurer of motorcycles, and a leading writer of boats and RVs, we're uniquely positioned to tap into the millions of current Progressive customers who trust us to protect those recreational products but just have not yet tried our auto or our property solution. We fully recognize that, historically, we've been in a subordinate relationship with these customers relative to their bundled home and auto carrier. But now we believe the combination of a viable property solution in our PHA program plus some of the unique auto-rating elements that Tom and John mentioned with Snapshot and Test Drive that provide phenomenal potential savings for these multi-vehicle households, we're at a tipping point in acquiring that bundled home and auto from our existing customers that we know drives policy life expectancy and top line premium growth.
Susan Patricia Griffith
Pat's uniquely qualified to really care about this complex customer, not just because he's a special lines GM, he actually has 10 vehicles with insured with Progressive, 7 of which are toys. So he gets it. He gets it from the business side, what we need to offer, but from the customer side, he's constantly looking at that. So that's a really great marriage. So multi-product penetration is one part of the formula, but the other part has to be servicing, policy servicing, claim servicing. Think of it like this: our customers have their own retention line, their policy life expectancy. Each one of those lines have events or touch points. So the time you purchase insurance, you have a loss, if you add a vehicle, if you have a loyalty program that you just got into, all those things, all those touch points give us the ability to get to know our customers better and then improve that experience and ultimately extend retention.
That's our goal. So Mike, what are you doing in terms of a policy and claim servicing to take those curves and tilt them all up and to the right?
I'm going to share 2 agenda items with you this morning. The first is a high-level overview of our servicing-related retention initiatives over the last decade or so and how they have evolved. And the second is a bit of a deeper dive into what we're doing from a claims perspective to improve retention.
We can segment our retention efforts over the last 10 years or so into 4 major categories. First category is blocking and tackling. And the initiatives in this category are really focused on eliminating friendly fire and making sure that everyday transactions with our insureds are as easy as possible. Our second category has been focused on creating a deeper emotional engagement with our customers. We know that as our customers' lives change, their needs evolve, and we need to continue to design, develop and deliver unique and personalized services to meet those needs. Pat talked to you just a few minutes ago about the importance of and our focus on multi-product policies. And as we continue to write more of that business, we need to make sure that we have the tools and the skills required to serve an increasing mix of preferred customers.
And finally, and maybe most importantly, we need to continue to invest in building our science around retention. Think of this as leveraging the skills we've built over the years in data analysis and segmentation as it relates to risk pricing and applying those same skills to our growing body of customer behavior data. This evolution over the last decade or so has really been a shift in focus from just reducing detractors to truly creating Progressive promoters and an environment for the 20-year customer. Now all of these initiatives are really about giving Progressive an opportunity to engage with our customers, because when we can engage with them, we can improve their retention profile. Our claims experience is certainly one of our more significant opportunities to interact with customers, and we may have no greater claims lever to improve retention than our service centers. So allow me to bring the Robinsons [ph] back into the story for a minute to discuss how we're going to use our service centers to improve retention.
Tricia told you that Dave Robinson [ph] is about ready to move out of the house. He's 23, it's probably about time. The Robinsons [ph] are very excited about that. They're going to take a an opportunity to downsize and move downtown. Mrs. Robinson [ph], however, is still a little bit unfamiliar with her new surroundings, and unfortunately, she gets involved in an accident. Now the good news for the Robinsons [ph] is that they're only about 10 miles from one of our Progressive service centers. And Mrs. Robinsons [ph] elects to use the service centers for both the inspection and the repair management of her vehicle. We know that customers who choose our service centers as their repair method rate at the highest NPS result of any of our repair options.
Now this NPS result doesn't give you a really true sense of how it feels to be a service center customer. So take a look at some of these recent customer comments from NPS surveys, and you'll see words and phrases like "fan for life", "best experience ever," "I'll never leave," "I'm going to share it with all my friends and family." We hear this all the time from our service center customers.
I've been involved in service centers from almost the beginning, over a decade ago. And I recently just returned from a trip where I visited 6 of our service centers, one right after the other. And my objective or my intent was to get a fresh look, almost an outsider's perspective, if you will. So I sat in the lobbies and I just watched. I observed our claim reps interacting with our customers. And when you do that, you totally get every one of these comments. In fact, these comments almost come to life right in front of you in the lobby. I returned from this trip very energized and excited about our ability to offer concierge level of service across all our service centers. And in fact, I truly believe that the customer service experience at our service centers is unmatched in the industry.
Let me try and give you a taste of what it's like to be a service center customer. In fact, take a walk with me through one of our newest facilities in Columbus, Ohio. It may feel like you're actually entering a Superstore from one of our commercials. Imagine how it feels to be greeted by name as you walk into the facility or having a place for your kids to play or maybe just watch TV. Or maybe you want to plug in and get some work done while you wait for the estimate to be completed. We have rental cars on-site, so we can get you on your way as quickly as possible. And you can rest assured, because we perform a quality inspection of every repair vehicle and provide a lifetime guarantee on those repairs. That's a little bit of what it's like to be a service center customer.
But as I mentioned earlier, we've got some work to do. We need to get more customers to choose the service center as their preferred option. The economics around just one more customer, one more car in our current service center infrastructure are obviously very attractive. As Tricia told this audience last year, the service center was not only the best option for our customers, it produces the best outcomes for Progressive. A vehicle through one of our service centers delivers our highest customer satisfaction, our most accurate estimate, our most efficient workflow and our best work environment. It's a win across the board for Progressive and our customers.
Now we've made a significant investment in service centers. We wouldn't have made that investment nor would we be making additional investments if we weren't highly confident about the benefits and the returns. So over the next year, we're going to be engaged in a number of initiatives to make sure that we achieve those benefits. We will be reenergizing the company around our service centers. We'll be launching an internal awareness campaign to make sure every employee understands the benefits of service centers and can communicate those to our customers. We'll continue to work relentlessly to make sure the customer experience remains best in class. And we'll open up several more to get closer to our customers.
We will be particularly focused on 2 things. One, making sure that our customers are aware of their local service center before they have a claim. We have lots of opportunities to this. We have many touch points and interactions with our customers. One interesting option is to pre-register our customers for their service center preference during their new customer on-boarding so that in the event of a claim, we can get them into the service center as quickly as possible and get their car in the repair process right away.
Number two, we need to improve the ability of our claim reps to educate and inform both our insureds and our claimants about the service center when they have a claim. Now number two becomes a lot easier if we can execute successfully on number one. But number two will always be critical. It's the moment of truth for us. It's a conversation with the customer about what they want to do with their vehicle at the time of loss. We have thousands of claim reps involved in thousands of these conversations every day. And frankly, we need to do a better job managing both the consistency around the content and the outcomes of these conversations. We have learned a lot about this over the last year and I'm very confident that we're going to make great progress in the next year.
Now Glenn mentioned earlier that we have capacity in our concierge claims centers. I feel very good that the initiatives I've discussed today and the current excitement level we have in our organization and the energy we have around service centers will go a long way in taking care of that problem. And at this time next year, we'll have significantly more customers choosing our service centers as their preferred option.
Susan Patricia Griffith
I'm thrilled that Mike's going to be spearheading what we're calling internally as our service center relaunch. And I look forward to updating you on that progress in the next year or so. So really, the ultimate question I think you want to hear is this: does better retention science drive top line growth? And the answer is, unequivocally, yes. In fact, if we add one month to all of our policies, that equates to $1.4 billion in additional lifetime premiums. While that's significant and we'll always look at month after month, we want to look at it a little bit differently. We want to be able to understand who are the consumers we need to build around for to be the 5, 10, 15-year customer . Why wouldn't the Robinsons [ph] stay with us as we continued to evolve with them and give them product offerings that fit their needs? Why wouldn't we continue to grow with Dave Robinson [ph] as he goes from an apartment to a home, he moves, all the events that happened in his life? We need to be there to give these customers a reason to stay.
These are the curves, our retention curves for the last 10 years. As you'll see, we have 10, 5 and current year. On average, we have increased our PLEs over this 10 years more than one month and at a much greater rate in the last 5 years.
In the 2011 -- oops, sorry, this is basically the future, I don’t want to give any guidance, but I want to be able to tell you that I think that with our science and the investments that we've made, what we've talked about today and what you've heard others talk about today, I feel great about our progress in retention. Glenn a couple of times in the annual report talked about acquisition. And in that same sentence, he talked about the importance or even more important at how retention is. So we'd love to have a vibrant book of business coming in the door, but we absolutely have to focus on retention. Retention is a low-cost acquisition model. It's good use of our dollars, and it's great use of our segmentation expertise at Progressive. So we're going to continue to evolve with our next level of studies, evolve with this science to get closer to our goal of becoming consumers' #1 choice.
So now I'd like to hand it over to Jeff Charney and Matt Lehman. They're going to give you an update on mobile, which I think you're going to find very exciting.
M. Jeffrey Charney
Good morning. Very different -- I'll show you a picture here. And Matt Lehman and I, we have -- we do have the same hairstyle, but in some ways we're very different people. Matt is a technology guy, I'm a marketing guy. But one thing we are firmly in agreement on, firmly in agreement, and that is the power of this mobile device. This mobile device I have here. And realistically, we have only scratched the surface, only scratched the surface of where this device could go for marketing as well as business in general. And just a little real-time example here, all of you brought your phones? If there was a table here, if there was a table, you would look at that phone an average of 12x, an average of 12x during the next hour, especially if there was a table here. Some of you would just unlock your phones. Some of you would do the, if you're a Mac person, do the push and swipe. Other people would just glance at their phones and some of you would just put the phone underneath the table and look around to make sure nobody's watching and just go after it. And that's cool, that's cool. We understand that. We understand that behavior, because we are glued to these phones. We are glued to the phones. And realistically, these phones have become the modern-day Linus's security blanket, where we feel really comfortable when we know exactly where the phone is and we feel even more secure when we have these phones directly in our hand.
So today, we're going to look at the 3 Ws of mobile and realistically look at what's really happening out there, why it's so, so important to us and where are we going with mobile. I'll start with the first W.
What is really happening? What is really happening inside and outside of Progressive? And I'm going to focus a little more outside of Progressive. So if you think about it, it is definitely a 3-screen world. 3-screen world, you have your tablet, you have the desktop and laptop and you have your mobile device. So it's a 3-screen world, and I think we all are very well aware of it. And we're not only building a plan that complement -- uses mobile as a compliment. We're building a plan that is built specifically around, around the mobile device. And if you look at it, last year, John Sauerland showed a [indiscernible] that mobile was booming. He said it was booming, and I'm telling you, the data is definitely, definitely supports that. And if you take a look even deeper into the data and you look at the 3 screens, you can see -- this is a weekly distribution searches by device, look at the weekly distribution, and you can see you here with the first 2 devices, with tablet, with tablet and desktop, you can see that people are really glued to their desktop and the laptops during the day, and on weekends, nights, they're going to the tablets, but when you really overlay mobile, you can see it is a constant, constant go-to device that people are really, really glued to. And if you take it -- take that weekly data and go deeper into the daily data, you can see throughout the day, laptops, desktops, and at night, you turn to your mobile device, but you could turn -- I mean, sorry, the mobile devices, literally tablet and mobile is really a constant -- it's a constant companion. It is with us always, and we are turning to it always.
That was a 3.5-minute overview of what is going on outside, outside the industry. Now Matt Lehman will take us through what's going inside the doors of Progressive.
So last year, John Sauerland shared some early data from Progressive, and he said we're starting to see those early signs of growth in mobile activity. So whether it's looking at mobile payments or mobile quoting, we're starting to see those early signs. And if we bring that data current, what we'll see, not unexpectedly, is that not only has that growth continued, but, in fact, it's accelerated. So to Jeff's points about growing adoption of mobile devices, at the macro level, we're seeing that clearly play out with adoption at Progressive. Now one nice thing about this growth is we're able to begin to see data patterns or anomalies. And we saw one in December of 2010, around the holiday season, we saw another one, we saw it repeat itself in December of 2011, and if we take a closer look, what we'll see is, obviously, a jump in mobile quoting comparing the week leading up to the December holiday to the week after the December holiday. Now obviously, these are lots of people getting new smartphones, new tablets around the December holidays and using them for insurance. And while we may or may not react to this kind of data pattern or anomaly, it's helpful because it allows us to better plan for things like mobile demand generation as well as mobile road maps.
Now speaking of road maps, last year, John shared with you our mobile quoting road map, and here's an update. You will hear more from Jeff in a few minutes about the importance of speed, how speed plays into the mobile context. But we think speed plays here as well. When you think about the aggressiveness of how we've been able to roll out mobile quoting, over the last several years, it's really helping to grow that business. And again, at the end of 2011, in 42 states, you could quote and buy what we call a 1x1 policy, so again, single vehicle, single driver. I want to stress that this is a full quote experience. This is an accurate rate based upon your characteristics, and you can further personalize that rate by naming your price, by changing coverages, by adding Snapshot, and then you can buy it right there on the device.
Earlier this year, we launched what we call 3x3. And again, this is the ability to get that full quote-and-buy experience for 3 drivers in 3 vehicles. Now this is important in a couple of ways. The first is that it allows mobile quoting to move upmarket into a more preferred household, that multi-vehicle household. But second, as of today, it's unique to Progressive within our industry. And then finally, we do have teams today back in Mayfield Village working on our end-state mobile quoting experience. This is what we call end-by-end. So again, it's that same full quote-and-buy experience on a mobile device for a number of vehicles and drivers sufficient to meet our total quoting demand.
Another element of mobile that we spoke about last year was mobile paid search. And again, paid search is an important part of our referral source. These are customers that show a high intent to quote. And last year, we showed some Google data that said about 10% of paid clicks were coming on mobile devices. Again, fast-forward a year later, that number has doubled. It's over 20%, again, this is Google data, for our category. But what I can assure you is that our data is tracking at or above that level. So when we think about the top of the funnel, people coming to us in mobile devices, they're doing so with paid search, we're meeting them with that full mobile quote-and-buy experience.
An area that we really didn't talk too much about last year was this integration of mobile and social media. Up there, you see Flo, our flagship social media icon, and she has over 4 million fans or likes on Facebook. The other data point we notice that as of today, most people are using and interacting with social media, Facebook included, on their mobile devices. So we know we're seeing that intersection of mobile and social media happening, and we've been aggressive trying to activate that customer base. We've been rolling out engaging activities and games through social media specifically geared towards mobile. And so while that prospective customer when they're using social media may not be thinking about getting an insurance quote today, we're creating that level of brand awareness so when they are, we're top of mind for brand consideration, and on top of that, we're there to meet them with that mobile quote-and-buy experience.
So hopefully I've updated you, given you a flavor of how that growth is happening externally, that huge adoption in mobile phones and tablets is playing out here in insurance. Jeff?
M. Jeffrey Charney
So why is mobile so important to us? Why is mobile so important to us? And it's realistically because people consume information in a very, very different way. They want it lightning speed, they want it ease of use and they want access to it all times. This is one tweet we got from one our customers, an actual tweet that said giving a shout-out to Progressive. She was in bed, got a quote that way. That's not necessarily how all of us would do it, but this is what she does. And she is not alone in this. There are many other customers we hear about that want the phone to be their direct point of contact with Progressive, and they do it at all hours and all times. And realistically, if you think about the way people adopt, the way people adopt the phones and the speed at which they are adopting the phone -- adoption speed, it's phenomenal.
I'll give you a little case study from Progressive. And you got to rewind 15 years ago, to 1997. It's the first time we launched the online buy. We got our first quote in 6 weeks, 6 weeks. Fast forward a little bit to 2011, we launched mobile buy. Our first quote in 60 minutes. And just this year, we launched online, we have -- basically you have your online optimize buy for special lines? We got our first quote in 6 minutes. Six minutes. Again, if you build it, if you build it right, they will come and they will find you.
You have to build it right. Taking that speed analogy one step further, there was talk many years ago about 15 minutes to get a quote, and that was more on the PC on the Internet, 15 minutes you can get a quote. Later on, we talked about 10 minutes, 10 minutes you could get a quote. Just last year with the mobile device, we're at 7 minutes. We broke the 7-minute barrier, and this year, this year alone, we have broken into the 6-minute barrier, and in fact, it's 5 minutes and 50 seconds on average, on average for people to get a quote on a mobile device. And let me tell you here right now, we will never, ever sacrifice quality and ease for speed. People want all 3: great quality experience, ease of use, all access, and they want it fast. And it's not just me coming up with this stuff. It's not just Matt and I saying, "This is what we think they want." We're listening to our consumers. We're gauging our consumers day in, day out. They are not just whispering what they want, they're shouting what they want, and these insights are coming at us every day, and we're applying these insights to everything we're doing. Matt's going to take it a little bit deeper into how we slice and dice the consumer data and show you how we segment our consumer.
So you've heard us talk a couple of times today about what we call our Choice segment. Again, it's our term for it. It's basically a 20-something consumer, non-homeowner, single car, possibly a renter. And there's a couple of things about this segment that are important. One is we do believe that many of the folks in this segment will mature into that multi-product, multi-policy, long-term PLE customer. The second thing is they love their mobile phones. And so when we think about are we getting traction through our mobile offerings with the segment of where we want to grow? The answer is yes. Again, 2 simple data points behind me. One is policies enforced. The second one are users of our electronic bill payment service. And in both cases, we're seeing a higher percentage of that Choice segment using mobile than a non-mobile.
And to put a face on this Choice segment, I promise you this is the last time you'll see Dave Robinson [ph], but we'll go to Dave Robinson [ph]. And again, Dave [ph], he's 23, he's a renter. When he needed to buy that first auto policy, he bought it on his phone. When he's adding that motorcycle and he's looking to quote-and-buy that policy, he's going to do that on his phone. And as his needs mature, his going to continue to look to his phone or perhaps his tablet as that method of choice. And Progressive will be there with mobile services to meet him. And if we can lock him in today with mobile, we'd have a great opportunity to grow with him as his needs grow as well.
However, it's not just about consumers like Dave [ph]. It's about forward-looking agents as well. And so earlier this year, we released a mobile-friendly version of our foragentsonly.com website. Again, that's the site where an agent would quote or bind a policy for a new customer, service a policy for an existing customer. And to be honest, the traction here has not been as strong as it has been on the consumer side, but this does a couple of things for us. First, it continues to enhance our position as a technology leader within the agent space, and frankly, we are starting to see agents adopt and think that adoption will continue to grow.
So when you put it all together, whether it's locking in Choice customers like Dave Robinson [ph], whether it's helping forward-looking agents build their business. And lastly, when you think about customer reaction to mobile, we're seeing [ph] again the data is thin, but we are seeing NPS scores that are directionally positive for mobile. So this is an important part of our strategy for acquisition retention and servicing, an area that we think is consistent with our brand of using technology to make insurance easier. Jeff?
M. Jeffrey Charney
So where are we going? Where we going with mobile? I'm going to take you through a somewhat simplistic 4-pronged strategy. It starts first with products, product development. Do we have the best products in the market? Second of all, we are very fortunate to have this store, this retail space where we can take those products and get them in the market.
Third, do we have the best products? Sure. We have to make sure we have the best products. But what is that next product that maybe even the consumer doesn't even know they need? How do you really take those products to product innovation? And finally, going directly into market in a big, bold, integrated, integrated way.
So I'm going to break this down a little bit further for you. If you -- first of all, start with product development. And if you think about it right now, Tom talked about -- let me back it up one second. Tom talked about his money chart. In some ways, this is our money chart. It's 13 categories. 13 categories here where we feel we stack up very, very well against our competition. Very, very well. You have to start with this. It's great to start with this.
Next thing you can do is go from there into the store. We are very fortunate to have the store. And this year, we've gone into the store when one of the commercials is a commercial called Gear Up, and all we did was have flow, just leave our phone behind. And again, it was a wink to the customer that we have this technology that they made.
We're going deeper into the store this year with a new commercial. And again, it's not ready for prime time quite yet, but it will be coming out very, very soon. It's a commercial called Pocket Superstore. And what it is, going back to what a lot of things we do in the Superstore is, is just what human nature. The human nature right now, a lot of people have these phones, and a lot of people have some very, very bad ringtones, very bad and very embarrassing ringtones. And it's just human nature, you see these things, you're trying to impress somebody, your ringtone goes off, not so impressed. It's features on -- focuses on 2 guys waiting outside a club, and one person's ringtone keeps going off. So I will show you that. I'll show you that right now. It's called Pocket Superstore.
It is an embarrassing [indiscernible] but I'm going to show you the most important thing is this phone is with you every step of the way, every step of the way.
Let's go deeper into product innovation. And Matt and his team have been working very, very hard to develop our product based on some of the research we got from our consumers. And the research showed that people use these cameras constantly. The point-and-shoot cameras they used to carry, the used to carry a point-and-shoot. They carry a SLR. They're really only carrying their phone now. They're carrying their phone, and they love, love, love to use their cameras. And we thought then and we talked to consumers and said, "Hey, if you could get a quote by pointing your camera at your license, by pointing your camera at your insurance card or by pointing your camera at your VIN ID, would that work for you?" Consumer said yes. A lot of work later, we came up with a demo of how this would work, and it's called Camera/VIN Capture.
It looks cool in a demo phase. But once you get it demoed and once you get all the kinks out, which we did, you go to market. And you go right back to the Superstore. Again, leverage that great asset we have in the Superstore.
And we developed a program, a new commercial called Bad License. And again, another human truth. Another human truth that basically says people aren't so happy with their license picture, and Flo specifically. And this is the commercial, taken to demo, going to product innovation, the commercial called Bad License.
And we've all been there, we've all been there. And then the one thing I would say to you -- and there's somebody's mobile phone going off right now, that's cool. It's not such a embarrassing ringtone, not that bad. But if you think about it, I'd like to challenge you. I kindly challenge you to go download the app and see how this process works. It might not be for you. But I think if you do this, if you take the challenge, download and take a picture of your license, see how easy it is to get a quote, I think you'll agree that this mobile device, it's just one of the many ways this mobile device will change the way people are quoting in the future.
So you look at this all together, it's a very simple yet complex system. And it's all based on always having the best product, taking it into the Superstore, looking to make the product better and then taking it in a bigger, more integrated way and all the way throughout the process listening, listening, engaging the pulse of your consumer.
And I mentioned integration. It's about integrated campaigns. And some of the things I couldn't talk about today are everything from mobile. We know where advertising is going in mobile, and we are very aware of it and that is included in our plans. We go from there to apps. There's so many apps that are going on in the world today, and people are addicted to apps. We have our own apps and are very targeted to the consumers we're looking for in our app world.
Gaming. How many of us are gamers? Again, 72% of all households are gamers, 72%. The average age of a gamer is 37. We recognize the importance of games, and we're in great spaces in the gaming technology with mobile.
And finally, just text. Just the basic blocking and tackling text. We're there as well, we're there as well.
So all of this stuff, all of the plan, all of the strategy I'm talking to you, how does it all add up? How does it add up?
This was not a great chart. This was last year at this time. Last year at this time. And we weren't happy with this. We weren't happy. We went deeper. And this year, this is what the chart is saying. This is Millward Brown data. It just came out. And it says that we are taking that rightful place as a leader in this space, in this space.
So just to summarize, first of all, we talked about what is going on both inside and outside Progressive, and as we answered here, it's radical growth really mixed with the capabilities inside to meet that growth.
Second question we asked is, "Why is mobile so important?" We are looking at the consumer, not only today, but what consumer will need tomorrow. Consumers are just looking at mobile as the device to get into Progressive, and we know that.
And finally, where are we going? We have an integrated strategy that could only touch on today, a robust, integrated strategy built around mobile, built around mobile, not just to complement mobile to meet this intense demand.
And I've been up here, Matt and I have been up here around -- roughly around 20 minutes, roughly around 20 minutes. What has happened in the mobile world during those 20 minutes? First of all, 12,000, 12,000 Android. In that 20 minutes, 12,000 Android devices have been activated. During that 20 minutes, 4,000 iPods, iPhones have been activated. And last but not least, 1.1 million apps have been downloaded during this 20 minutes. During the next 20 minutes, it will happen again. In the next 20 minutes, it will happen again. In the next 20 minutes, it will happen again. This is a brave new world.
And I know a lot of you are saying, look, Jeff, where are the sales? Where are the sales coming from? Where are the sales?
And I'm going to show you a map. And this is a data visualization map. This is not cartoons. This is not animation. This is based on real data and real mobile quotes. These are quotes. Quote starts from 1 week, I select a target from 1 week of a month. And if you look at these, and you look at every one of the dots on the map. And you look at this as June 14, 2012, I want you to think about June 14, 2013, June 14, 2014 and June 14, 2025. It just shows you there are people behind every one of these quote starts and people that want to do and want to have a very different Progressive experience with their mobile device. And if you look at that and really visualize it, you can see why Matt, myself and all the management team at Progressive are so, so excited about this third stream.
Thanks very much. I will turn over to John Barbagallo and Karen Barone, who will take you in deeper inside the factory.
John A. Barbagallo
Good morning. Tom Hollyer and I are very excited to be here today, talk to you about Snapshot. We're going to talk a little bit about where we've been over the past year, and then we'll give you a little insight into where we're going.
Tom Hollyer, you may recall, was here last year talking about Snapshot. Our Snapshot development reports into Tom, amongst a number of other things, so Tom is our resident expert today.
When Glenn delivered his broader comments at this meeting last year, he was giving a bit of a perspective on the decade. And when he talked about usage-based rating or Snapshot, he talked a bit prospectively and said in effect, a decade from now when we look back at our offering around usage-based rating, we will consider it very elementary. Not surprising. But frankly, when I look back just 5 years, I have the same impression around the magnitude of change. We have come very far with our usage-based rating. And honestly, we feel we are just starting to take off.
Today we're going to focus on our operations with Snapshot. Some of you probably know, in fact, I got some questions out in the breakfast time there around our intellectual property. We have had success in that space. We defended 1 patent over the course of probably the last year, or more, actually. And we were granted 2 incremental patents this year. So great progress there.
But we fully understand that winning in this space is about leadership and execution. So we're going to focus our comments today in this section on operations around Snapshot.
And to start with that, we're going to have Tom Hollyer give you a little insight into the past year around UBI. So Tom, what is a year in UBI worth?
Thank you, and good morning. I'm going to give you an update of our progress on UBI over the last 12 months. And bottom line, we are very much on the trajectory that we outlined at last year's investors conference. In fact, we just passed an important milestone where we have now written $1 billion of UBI premium in the trailing 12 months. So that was a personal goal of mine. I'm very happy that we got there.
We've also made a lot of progress expanding distribution. This map shows where we have UBI. We are filling in the map. Indiana and Washington are in process. We are also optimistic about Illinois.
And one of the good things is that we've had good working relationships with the regulators, and we've been able to achieve this rollout while maintaining trade secret protection for our UBI algorithms. So, again, we have the broadest distribution of any of our competitors.
So the question is, "How much is a year of UBI experience worth?" We think it's worth a lot. Last year when I was here, I told you that we were approaching 2.5 billion miles of cumulative driving experience in UBI. We are now approaching 5 billion miles. So that is very important because we have a huge scale advantage in terms of the amount of data that we have for analysis. We have detailed driving behavior data and loss cost for each of these policies. It is a very rich data set. And I can tell you that we have, absolutely, our best people working on it. So again, we have doubled our cumulative experience over the past year.
Now also last year when I was here, I told you that driving behavior data is our most predictive rating variable. I was right about that. What I was wrong about is the magnitude. I mentioned we have been mining our data, and we have identified opportunities to improve the accuracy of our UBI algorithms by an amount equal to points. So if you look on this chart, points is the #2 most powerful rating variable. We've identified opportunities to improve UBI accuracy by an amount equal to points. So it is already our most powerful variable. It is getting even more powerful.
Last year, I also told you that it was our intent to achieve a leadership position in the UBI brand space. We have done that. On the left, you can see awareness of Snapshot. We're really starting to climb following introduction of our national media campaign, and it's continued to climb. And the same is true of attribution.
So in summary, what is a year worth? We think it's a lot. We've gotten very good, very broad distribution. We have doubled our cumulative experience, we've identified opportunities to greatly improve the accuracy of our UBI algorithms, and we have achieved a very strong brand position.
Now let me turn it back over to John, who's going to discuss how Snapshot can increase demand and long-term growth.
John A. Barbagallo
I think I'm good. I think I'm good with the lapel. It sounds great. A couple of things I hope you took away from Tom's comments. We have a huge lead [Audio Gap] UBI space. And the rating segmentation here is by far and away the greatest we have seen with any variable preceding this. I think you can intuitively understand that how an individual drives is way more predictive of their future loss cost than are broad-brush things, as Glenn mentioned, like age, marital status or gender. So we have a huge lead here. I personally believe that usage-based rating or underwriting will be a required competency to play in this industry sometime in the future. You pick the date, but I personally believe that it will be a required competency to play in the industry in the future, usage-based rating.
So that's a little longer-term commentary. What's the nearer-term question? And that is around growth. Over the course of the past a little over a year now since we launched Snapshot, we have been including about 20% of our creative rotation on television with Snapshot ads. We've gotten better and better over the course of this time in explaining Snapshot to consumers, and Snapshot is not a very simple thing at the outset to understand, so we've gotten better and better. And Tom showed you that awareness of Snapshot has increased. He showed you that attribution has increased. What we find as well is when we talk to people about Snapshot, the consideration for Progressive overall goes up. You may say, "Well, that seems like what you intended," and that is absolutely the case. But consider that over the course of the past 5 years, we worked through a number of different models to get UBI to consumers. And while actuarially sound and accurate, not all of those models were highly appealing to consumers. So we had a surcharge model. We had a model where we charged a technology fee to consumers. We now have a model that when we talk to U.S. households about Snapshot, consideration for Progressive goes up. We think that's significant.
So that's a consideration. Do they call or click? What this graph shows you is the effectiveness, if you will, of our best advertisement across different categories of ads we run in our Superstore campaign. And what you can see here is Snapshot is above average. On this slide, the index is averaged to 1. Snapshot is above average. And I should also mention that while Name your Price continues to be by far and away our most effective marketing message, and we're going to continue to leverage that, a lot of these things are not mutually exclusive. So you can Name your Price and enjoy Snapshot. But what I'm trying to show here is that Snapshot, when we use it in our Superstore campaign, over-indexes in terms of response.
So people call or click. Do they take Snapshot? The short answer here is yes. This is our Direct business. We told you previously that around 1 in 3 consumers in our direct channel select Snapshot. You can see at our national launch, we took a jump up. And you can see that we continued to grow the take rate, if you will, of Snapshot -- this is actually new business mix, but I might refer to that as take rate -- over the time since we have launched the product.
Another important thing you see here is that the phone channel within our Direct business has a take rate higher than Internet.
What we conclude from this is that when we are able to talk to consumers more and more about Snapshot and explain it, the propensity to opt in is higher. So we're going to continue to use the learnings around how we talk to consumers about Snapshot to improve our Internet take rate as well. And we think we can get both of these lines continuing to rise and perhaps converge.
So we've got awareness, we've got consideration, we've got demand, they're taking it. The end question obviously is, "Do we sell more business because of Snapshot?" The short answer here is yes. I don't have the complete data set for you here. But I can tell you we did a very controlled test within our online quoting. We're in a handful of states. We randomly removed the Snapshot offer from the quote flow. So we were able to understand the full funnel economics, if you will, from start to sale. And I can tell you in aggregate, sales went up. And just to put that order of magnitude, I can tell you lifetime premium annually where we would have Snapshot everywhere goes up in excess of $100 million annually. So lifetime premium per annum, up over $100 million as a result of offering Snapshot everywhere.
So I hope you've concluded Snapshot is helping us grow new business in the direct channel. Our agency channel, also very, very important, our -- actually, our larger channel at this point. We told you previously the take rate here has been below the direct channel. We haven't actually quantified that for you, but it continues to be low. That said, our agents are beginning to adopt Snapshot. We require with our current version of Snapshot agents to what we call certify into offering Snapshot, which simply means they have to complete some basic training around the program. And we now have almost 25,000 of our agents certified to offer Snapshot. You may recall we have over 35,000 agents and brokers countrywide, so 25,000 is a huge chunk, especially in the states in which we offer Snapshot.
We also see the take rate line here. It also jumped at national launch. That's great. Recently, maybe it's risen some. It has not risen to the degree that we think it should or could.
And recently, we did a what we call a Snapshot revolution. So think of a sales blitz in our home state of Ohio around Snapshot. Took an integrated approach, had our entire field sales force focusing on Snapshot for a period. You can see on the bottom left chart here that since that blitz, take rate in Ohio Agency business is up about 20% since we rolled that out.
Actually starting this week, countrywide, we're employing the same blitz approach for all of our agents who have Snapshot. So as Glenn mentioned in the annual, the quote is up here, we fully expect increased consideration and use amongst our agents soon.
Another very important thing to consider when you're thinking about agents and Snapshot is that to date, we have seen a very good mix of driver scores coming out of our agents. So you might interpret that as an agent having some ability to know based on knowing their customer whom to offer Snapshot to. So we're seeing a better mix of driver scores coming out of our agents to date than we are in our Direct business.
So I've talked a lot about new business so far. Retention, as you know and as we know, drives growth. Can we retain customers longer with Snapshot? The chart on the left here shows you our NPS score for our current version of Snapshot relative to non-Snapshot users. And this is a cut by tenure, but I can tell you pretty much however we slice this data, net promoter scores for our current version of Snapshot are higher than non-Snapshot users. This highly correlated and translates into better retention.
The graph on the right here is what we call a decay curve. And since we'll be using these throughout the presentation, let me pause and make sure you understand how we -- how to read these. We start with 100% of our customers in any cohort, think of all customers that incept within a month or year, at the top left, and then we measure the percentage of those that are still with us over time. And as you can see, the current Snapshot version is well above the non-Snapshot group. And I think you can see that gap is growing over time.
You may be wondering what this means to top line premium, and you probably have already read the bottom of the graph, but let me walk you through that -- or the chart, let me walk you through that, nevertheless. So retention is far better. We are giving discounts to most of these customers. What does that mean in terms of premium? The net of those, unadjusted, is a lifetime premium increase for Snapshot customers of about 25%. Snapshot customers overall skew more preferred. Preferred customers, all else equal, will stay with us longer just because of who they are. When we do some adjustment for several variables that are reflective of that preferred nature, we still conclude that lifetime premium for Snapshot customers is 13% higher than non-Snapshot customers.
I'll also tell you that I think that 13% is conservative. So we arrived at that by taking the current gap in that retention rate and projecting it out as equal into eternity, if you will. I think you can see that gap has been growing. If that gap continues to grow, we will see an even higher increase in lifetime premium for Snapshot customers.
So hope that gives you good perspective around Snapshot as a growth vehicle. I know I've heard a lot of questions around the profit side of Snapshot. We're just giving discounts in this program. So a great question that Tom Hollyer is going to answer for us is, "How do the economics work?"
Thank you. Can you hear me? Okay. So we have received a number of questions about how Snapshot economics work, particularly since we are not surcharging drivers that have worse-than-expected driving behavior. So to explain that, I'm going to start with more of a conceptual model. And then after that, I'll get into some of our actual results.
So let's start with the conceptual model. Across the X axis, we have total cost. Across the Y axis, we have price. And think of each of those dots as a customer segment. So we have an array of customers.
Now if these customers were perfectly priced, they'd be right on the diagonal. They're not perfectly priced because we don't have perfect information.
Now let's introduce Snapshot. The orange customers are the ones that sign up for Snapshot. Now with Snapshot, we have a much better understanding of their actual loss costs. And for those people that have better-than-expected driving behavior, we can lower their price by up to 30%, and they turn green. Now on average, these customers still remain above the line.
Now we do have to take a modest base rate increase because better drivers tend to self-select into Snapshot. And so the remaining drivers are slightly worse than average.
So we now have 2 sets of Snapshot customers. First are the drivers with better-than-expected driving behavior. They were able to achieve discounts. Again -- but the way we have our pricing set up, on average, they remain above the line. So we are making better margins than our 96 target. We then have a second set of customers whose driving is worse than expected. They are below the line.
Now the mix of these 2 customer sets is very important. It really is all about the mix, because the excess margins or the good margins from the good drivers need to offset the worst margins from the poor drivers. And we can influence that mix by who we attract and who we retain.
So for the good drivers, if we can have -- if through our marketing, we can attract people with good driving behavior, that helps our mix. If we can also retain those customers by offering them discounts that our competitors cannot match, that also helps our mix.
Conversely, for customers who have worse driving than expected, perhaps they're not very pleased about the fact that they did not receive a Snapshot Discount. And perhaps some of them choose to defect to the competition. That actually helps our mix, and it hurts our competitors' mix, because our competitors likely do not have the benefit of understanding driving behavior. So it is a classic example of adverse selection.
So that's the conceptual model. Now I'm going to turn to some of our actual results. First, loss experience. Very important. This shows our adjusted loss ratio relativity by UBI tier group with the best drivers on the left and the worst drivers on the right. And our adjusted loss ratio means we took out the effect of the UBI discount.
There's 3 lines on there. One is our predictive loss ratio and then our new -- actual new business and our actual renewal business. And what you see is they line up very closely. So the -- our predictions coming out of R&D are very -- are proven to be very accurate, and I personally feel very good about that. And in fact, you could argue that the new and renewal actual experience has been slightly better than forecast.
Now this chart I refer to as the money chart. In blue, we have lifetime loss experience, loss ratio. In orange, think of it as 1 year retention. And then going across the X axis on the far left, we have non-Snapshot customers. And then moving towards the right, we have Snapshot customers grouped based on the discount that they receive. So remember from the 2 previous slide, I talked about there's 2 customer sets and how their mix is very important.
First, let's turn to the drivers who are worse than expected. What you see is their loss ratio is higher. That's not a surprise. But what you also see is that their retention is lower. So in fact, they are defecting to the competition at a higher rate. And again, that helps our mix.
Now turning to the customers who have better-than-expected behavior. Two things to note here. One is their loss ratios are very good. And actually, the better drivers they are, the lower their loss ratio is. So we are making good margins on these customers. The other thing to note, that the better drivers they are, the higher their retention. That is also very important, because it helps our mix. And in fact, the retention from those customers is higher than we would expect, given the discount that we are granting to them. So they like the fact they have a personalized rate. They like the fact that it's their policy, and they stay.
I should also tell you that the actual mix that we are achieving between these 2 groups is meeting or exceeding our pricing assumptions.
Now the final piece of the economics is our cost structure. This shows our Snapshot technology cost over time. We have been driving down our costs, lowering the price of devices, lowering the price of telecommunications. And in fact, we are -- our lifetime tech expense is approaching 1% of premium. So we really are leveraging our scale. We take this into account in our pricing.
So a quick summary. Particularly those drivers that are better than expected, we are achieving very good margins and we are achieving very good retention.
Now let me turn it back over to John, who's going to discuss how Snapshot can appeal to a broader audience.
John A. Barbagallo
Hope you have a little better understanding around the economics there. The economics are working. And as that mix shifts and those folks retain longer, they will work even better.
As I pointed out earlier, around 1 in 3, but an increasing percentage of direct customers opt in to Snapshot today. A huge opportunity for us is to increase or broaden the appeal of Snapshot.
Last year, we showed you the following pie chart, which was our best estimate from a number of sources around what portion of the world is open to usage-based rating and what portion will opt in. You can see we thought about 40% liked it, 3 out of 4 of those were willing to enroll; 25%, unsure; with the balance disliked it.
But to date, we've been talking to consumers about a model that we think of as a buy-then-try model.
So just to review how our current version of Snapshot works. Today at day 0, which is the day you buy the policy, you opt in to Snapshot. We send you the device, you plug it in, you drive for 30 days. And after that period, we give you an initial discount. That applies to your policy premium. You keep the device in your car for the balance of the 6-month term. And then after 6 months, we use that information to provide you what your final discount is, which you carry forward.
We think and have been piloting that another option here can significantly broaden the appeal of Snapshot. We think of this as the try-then-buy version.
So today, actually in a lot of states, you can test drive Snapshot. So if you go to our website, you can sign up, we send you the device, you plug it in, you drive for 30 days. At which point, we give you your initial discount. Of course, not everybody gets a discount in that first 30 days, but the majority of customers do. They then take that discount into the quoting process, and that applies to the new business premium. So to the extent they are getting a discount, that makes the premium more competitive than it does in the above model. Similar to the buy-then-try version, you keep the device in your car for the balance of that 6-month term, and we give you, again, your final discount.
When we talk to consumers about this model, if you think of that pie chart, we get some different perspectives. So that unsure crowd, they say, "I might try that." It's a no-risk, free trial offer, if you will. The dislike crowd, they might not say that they like it, but they say, "I might try it." So we think this significantly expands the appeal.
And we're now thinking about the available universe, if you will, for Snapshot customers, as per the chart on the right side of this chart. So this shows the spectrum of people that are in market shopping today. A chunk of those opt in to Snapshot already. We think the try-then-buy version will appeal to even more of those customers that are actively shopping. And then we think -- we know there's a huge segment of the population that is relatively inert. They have no intention of shopping their auto insurance anytime soon. But to some of those customers, the opportunity to show that they are a better driver and maybe get a rate we think might bring them into the market.
Let me share with you one customer's experience from our pilot to give you a little more perspective.
So a data point of one, but hopefully that gives you a little flavor of what consumers will think about as we offer test drive -- the opportunity to test drive Snapshot.
To date, in the pilot, we have had somewhat of a basic, a rudimentary experience for those consumers. As we built it, we didn't want to build it out too robustly until we saw exactly how to build it best. And actually last week, we elevated a far more engaging experience for test driving Snapshot.
If you think about the mix that Tom talked a lot about, our goal is to get the best drivers really engaged and ensure that they come and get a quote from Progressive. To ensure the mix is optimal, we have varying experiences across the spectrum of what you are showing us your indicated discount will be. So this experience is just elevated. We don't have, obviously, any data from that experience yet, but we think it will ensure the mix is really advantageous to us.
When you think about the try-then-buy model in the context of Tom's economic model, if you will, what I'm showing here is the same loss ratio curve that Tom showed you for our current Snapshot version. I collapsed the buckets a bit, because the pilot is not a huge data set in any stretch of the imagination. It's fairly a small pilot, but representative, we think. So we've used the loss data here from the broader current Snapshot version to show you across those discount buckets what our loss experience is. We're also showing you actual pilot results around the conversion for customers in the pilot.
Not surprisingly, folks that get 21% to 30% discounts convert at a rate 3x higher than those who get no discounts.
So if you think of this model, one way to think about it is that for our lowest price point, the conversion of customers is the highest, and our margin is also the highest. Say that again. So from our lowest price point, where the propensity to buy Progressive is highest, our margin is also the highest. Pretty interesting model in auto insurance and I think in any industry.
We're very excited to see how we can engage U.S. households to get them to test drive Snapshot. We have a fully integrated media campaign we'll be launching here within the next several weeks. And just to give you a little flavor around that campaign, I'll show you our kick-off television ad.
So again, we're very excited to roll this opportunity to test drive Snapshot out to U.S. households. That will start in a few weeks.
But in aggregate, we feel great about our Snapshot program, in excess of $1 billion, a leadership position in what I believe will be a required competency going into the future with a lot of opportunities to deploy to a broader audience with a very interesting economic model.
That's it for Snapshot. We're now going to turn the floor over to Tricia Griffith, who will talk about retention with Pat Callahan and Mike Sieger.
Thanks, Jeff. That is some great stuff. Very exciting things happening in the mobile space. Glenn in his opening comments talked about -- touched on the importance of investing in our infrastructure, or as we call it here, our factory. He spoke specifically about PolicyPro, which is our replacement core processing system, which clearly is a large-scale multiyear investment. And there have been others, our ongoing investment in concierge claim centers, for instance. We've been able to make those investments and increase our advertising spend while maintaining very competitive expense and LAE ratios.
Now we're going to touch on a few other investments this morning that, while not on the scale of something like PolicyPro, are still important and get to kind of core competencies around analytics and big data that are becoming increasingly important for companies wanting to compete. Now these investments will not only provide for kind of the more efficient operation of the business going forward, but will position us to absorb future growth in a way that will not compromise, but will actually improve the quality and consistency of our customer experiences, our position and performance in the independent agent channel and our responsiveness in a rapidly changing cost and rate environment.
I'm going to start by addressing this question, which is, "As you grow, how will you manage quality and consistency of your customer interaction experiences?" Now, customers interact with us millions of times a year in many different ways, be it voice, online, on mobile or in person. And it's the very nature of those millions of interactions that will ultimately define the Progressive brand. So it's important that we have each of them be of the highest quality and as brand-consistent as we can possibly make them.
Now even with the rapid development and adoption of online and mobile consumer applications, some of which Matt and Jeff just shared with you, phone calls continue to be a very important vehicle of customer interaction, especially on more complex or sensitive matters where oftentimes, a phone call is the best and fastest way to service the customer.
As you see on the screen, we handle 47 million phone calls a year across Personal Lines, Commercial Lines and Claims. We do that in 9 different call center locations, and we have also people working at home. Now managing the quality and consistency of all those interactions is made challenging just by the vast number of interactions, variance in the skill development of our more than 6,300 customer service representatives and, frankly, just the very nature of the calls themselves.
Now traditional approaches to managing call quality and mining call data for insights on the customer experience can be very labor-intensive and almost impossible to scale. Call quality was assessed through a random sample audit of individual calls and the review of customer survey data, which together provided a rather incomplete view of the actual experience. What we were missing was the key issue of call content, and we lacked the ability to zero in on the most relevant moments of the experience.
What you see on the screen now is an animation of an actual phone call.
In the next image, you'll see is what we call -- is a view of what we call a painted call. Now we are working with an external vendor to capture the critical elements of all our phone calls. Things like customer distress levels, which would be indicated here by the red dots; something called non-interaction time shown here by the white portion of the call bar, between minutes 6 and 7. And occasionally, we look for keywords or phrases, which would be indicated by a little black arrow there.
Now we're capturing and compiling this information over millions of phone calls, creating many, many terabytes of data. And we are assessing them through a series of linguistic algorithms to develop new insights into the actual experience. And this is beginning to help us in a number of different ways.
One way it's helping us is to improve the quality and consistency of our phone calls by providing our managers with timely and detailed reporting on those critical aspects of the call.
In this very simple example, we are able to take a group of our customer service representatives and segment them into quartiles based on the average amount of non-interaction phone time they were experiencing on their calls. Now non-interaction phone time is the portion of the phone call where neither party is saying anything, and that can happen for a variety of reasons. But it's a known correlative with less satisfying experiences, and it also tends to extend the length of the call. Now if you use a rough dollar-per-minute variable cost of phone calls, you can see that nonproductive time on the phone not only lessens the experience, but it can be very expensive. By simply segmenting this group based on that one variable, our managers were able to make a slight reallocation of our coaching resources. Coaches spending more time with those representatives operating in the lower quartiles. And more importantly, the coaches were able to focus in specifically on those phone calls that were driving the high average.
By doing this, we were very quickly able to improve the overall quality of our phone experiences by taking out non-interaction time, and we are able to start closing the gap between the top performers, who are also improving, and those in the bottom quartile, making the experience more consistent. This also made the coaches and the managers more efficient and more productive, which makes the whole model scalable. That's just one simple example of how call analytics is starting to help us.
Another way it's helping us is in the early identification of product and process improvement opportunities by quickly revealing unexpected changes in call performance. So again, another simple example here. But in the State of Colorado, we saw an uptick in non-interaction time and an attendant increase in average call length in the State of Colorado just simply going from November to December. With this information in hand, managers were able to drill down on these Colorado phone calls by call type and by call pattern. And it revealed that a new discount introduction in Colorado in November was creating a fairly high degree of confusion for certain customers. Knowing this, they were able to redesign the process in the communication around that discount and improve the experience.
Now under the old methodology with these Colorado phone calls being spread over literally hundreds of customer service representatives in 3, maybe even 4, different locations, this kind of issue would have gone undetected for a very long time, if it was uncovered at all.
So here, similar to how Snapshot is beginning to drive and unlock new insights into auto insurance pricing by mining a massive data set of actual driving experience, call analytics hold similar promise for improving the quality and the cost-effectiveness of our customer experiences by revealing the most relevant moments of those experiences and allowing us to process call data at an entirely new scale and speed.
Now Karen is going to share with us some investments we're making in our agency distribution system. And she'll do that by addressing the question, "How can you leverage scale, increase engagement and drive performance in the independent agency channel?"
Early this morning, Glenn shared that our Agency business is doing well, and there are a number of things contributing to that: product advancements, working with comparative raters to improve our rate presentation. Certainly competitive rates helped. But there's also the work going on to engage our agents to drive results.
And I would like to share with you today a couple of tangible examples of how we are engaging our agents and also at the same time how we're adding value beyond the commissions we pay our agents by leveraging our scale and our digital marketing expertise to help our agents improve their online marketing efforts.
The first example is the creation of a diagnostic tool and assessment process that we refer to internally as Scout. This enables an account sales representative to assess an agent's online presence, consult on where they have areas of opportunity and then guide them to the resources they need. And we review 3 areas of an agent's online presence, local search, social media and then website optimization.
I want to give you a quick demonstration of the tool just so you have a feel for how the process works. But keep in mind, I'm only reviewing with you 1 of over 20 checkpoints that we review in assessing an agent's online presence.
In this example, I'm reviewing local search and specifically whether or not the agent has claimed their business listing on the main search engines, Google, Bing and Yahoo!. We use a web service to check their listing. And here we find in this example that the agency has not claimed their business listing on the main search engines. That's a clear opportunity for our agent.
We go back to the tool, enter our findings. And then after we would complete all of the checkpoints in the local search assessment, we would produce what we refer to as a scouting report. This scouting report becomes the basis of our consulting and engagement process with the agency.
You might be asking yourselves, "Why not just host a workshop and cover this with your agents?" We've tried that. And we find that a workshop isn't as effective as an individual assessment that is customized for an agent's specific areas of opportunity, especially when that assessment comes from someone who understands the agency business through an extended working relationship. The assessment process positions our account sales representative as a digital marketing adviser, taking the best of what we've learned in guiding the agency on the right investments to make based on their unique opportunities, their goals and their budgets.
There's a few things to remember about this process. First is we focus our efforts in this area on agents who are promoters and advocates of Progressive so that we're maximizing the return on our time investment in that agency. Second is that this the engagement point with the agent on helping them improve their online marketing efforts. And then finally, the quality and the value of this assessment matches comparable services that are found in the industry at a much higher price point.
The next example of how we're engaging agents and how we're leveraging our online marketing expertise to help our agents is focused on local search specifically. And this is a visualization of the local search ecosystem.
A business owner needs to accomplish 2 things in this space in order to gain visibility. First is to claim and verify their business listing on the main search engines. Again, that's Google, Bing, Yahoo!. The second is to get that listing into the various online directories, represented here in this visual. Examples would be Yelp, Yellow Book or Superpages.
This can be a pretty challenging environment to operate in for a small business owner, and we recognize this as an opportunity for our agents and created an online visibility program called ListAgent.
Let's go back to the scouting example that we reviewed a few minutes ago. We learned that the Walker Auto Insurance Agency had an opportunity to claim and verify their business listing. The connection here is that ListAgent is a service and an easy solution that we would recommend to this agency to claim, verify and manage their listing.
There are 2 options for an agency to choose from with participating in ListAgents, but the core listing management service is $59 a year, and comparable services retail in the marketplace for about $300.
So for an agency, ListAgent is an inexpensive and easy solution to gain consistent exposure in the online ecosystem. It's really a "do it for me" solution that improves an agency's local search visibility. Early results are telling us we are seeing improvements in local search visibility for the agents that are participating, and the feedback from agents is very positive.
I want to share with you a few of our agents and what they're telling us in terms of the results they're receiving.
So I've given you 2 examples of how we're creating noncommissioned value for our agents that leverages our online marketing expertise and is scalable through our existing agency sales force. We're creating engagement to drive results, and this will enable us to sustain our growth and maximize our share in the channel.
I'm going to turn it over to John who is going to answer our final question, which is, "How will you remain responsive to changing cost and rate environment while ensuring quality control?"
John A. Barbagallo
Yes, I'm going to touch on 1 last factory investment, and it has to do with our rate-making and product-delivery system. Admittedly, this is kind of inside baseball [ph] stuff. And as Karen said, I'll be responsive to the question, "How do we remain responsive in this environment and not compromise the quality of our products?"
We believe we create competitive advantage at Progressive by getting our rate change -- rate and product changes to market quickly and being nimble. To that end, we do about 300 rate revisions a year across the product portfolio of varying degrees of complexity. We do this to stay ahead of the competition and to track as closely as possible with changes in loss cost and expense trends.
Now for some, there's a degree of rate stickiness in auto insurance pricing, due in part to the time, cost and complexity of making changes. Now we recognize that, but we accept the challenge. And we prefer to make more frequent small adjustments to rate as opposed to waiting longer and making more fundamental changes.
Now rate revisions are complex, albeit repeatable, processes. And this is actually a simplified version of the process.
At Progressive, a typical rate revision will involve more than 200 programmable tables, dozens and dozens of edits and lots of other logic. Conservatively, there are 500,000 changeable data elements in a single rate revision, or for us that's about 150 million changeable data elements each year that need to be reviewed and verified for accuracy. In addition to just the sheer amount of data, a typical rate revision, depending on complexity, may include 10 or more different business and IT functional areas, touching the process multiple times at various points, introducing the opportunity for error and requiring that we have a high degree of consistency across all the functional areas.
Now to manage this high degree of change and collaboration, we have created a tool we call Intellicheck, and we apply that repeatedly throughout the process. Intellicheck allows us to manage the quality risk associated with rate revisions without adding additional complexity or cumbersome manual processes. Essentially what Intellicheck does is it provides our analysts and our programmers with rich visual comparisons and trend analyses and allow them to quickly review large amounts of data for accuracy against what is intended and for compliance with established business rules.
Now I'm going to show you a quick outtake from an internal Intellicheck demo. And I want to manage expectations a little bit here. This is not a highly produced professional training video. In fact, it was created by one of our young analysts on his laptop in an afternoon. And I think you'll quickly see you get the feeling that it's something only a hard-core data analyst could truly love.
But we won't. That demo actually runs for 15 minutes. And believe it or not, it enjoys wide viewing at Progressive. But let me just get to the bottom line. We're geeks. This tool was developed in our Commercial Auto area, and that's where it's been in use the longest, although we now have it deployed across the enterprise for all rate revisions.
In Commercial Auto, since we've been using Intellicheck, we have seen a more than 30% reduction in the amount of time we spend auditing our programming while increasing the scope of what gets audited by more than 100%. So a huge gain in productivity. We have also seen a more than 50% reduction in the time it takes us to identify potential errors that also shortens our cycle time and eliminate potentially costly rework further down the process. And we're very confident that Intellicheck has been a key contributor to reducing our actual realized defect rate, already a small number, but we've reduced that by 48%, and we continue to see improvement there.
So going back to the original question, we are very confident we can ramp up the level of rate revision activity as necessary without compromising the quality of our products at all.
So real quickly, we have covered a few other factory investments, investments we're making in our call centers, investments with our independent agents and investment in our rate-making and product delivery system that will allow us to absorb growth, lower our costs and continue to improve the quality of our customer experiences.
Now Brian Domeck is going to join us on stage to open up the Q&A by leading a quick lightning round.
Brian C. Domeck
Well, we're headed down the home stretch here now, in a few minutes, Glenn will join me on stage for the more formal Q&A. But before that, I will host what we'll call the lightning round Q&A. This is a series of questions, Glenn referred to them as meat-and-potatoes questions. I think of them as some of the most provocative questions that we ask ourselves. I will answer a few of the questions. I'll also offer a few of those to my colleagues to answer. And we think they're a response to the questions we have heard from you in the past, either with our monthly news Releases or in anticipation of today's meeting. So if we're ready, we're going to begin lightning round.
Question #1. Profit margins are narrowing. Can you provide a little bit of commentary on the major drivers?
I will answer this question, and I will give you a couple of different viewpoints: one from a calendar year perspective, and secondly, from our perspective a more important perspective, from an accident year perspective. So we released May's results yesterday, reported a 97.5 combined ratio, higher than desired, bringing our year-to-date combined ratio to a 94.9. It's about 3.8 points higher than the similar calendar year combined ratio as of last May.
What are some of the major drivers? I've identified 5 here major drivers. First, for Personal Auto, our 3 largest states, Florida, Texas and California, all have a calendar year combined ratio several points above similar combined ratio as of May last year. In fact, those 3 states contribute 2.4 points higher to the calendar year combined ratio.
Commercial Auto so far this year has a combined ratio about 4 points higher than a year ago, that contributes 4/10 of a point to the delta; and special lines has a combined ratio of about 4.5 points higher than the same period a year ago. That contributes about 3/10 of a point to the calendar year combined ratio. Those component pieces are about 3.1 points in total. Glenn in his first quarter letter mentioned the mild and warm weather in the spring contributing to increased frequency in our motorcycle product line. That is the major driver of the special lines higher combined ratio so far this year.
Now let's go to a different perspective and look at an accident year perspective. So far, through May, on an accident year perspective, our combined ratio is 93.9, which is 7/10 of a point higher than the year ago. So what's the difference? Last year at this time in May, we had over $130 million of favorable loss reserve development, which is contributing 2.1 points to the combined ratio. This year, so far this year, we've had unfavorable development of $66 million, which is about adding a point to our calendar year combined ratio. So this delta difference going from favorable development to unfavorable development of over $200 million is basically contributing about 3 points of the delta calendar year difference. And in fact, that change from favorable development last year to unfavorable development this year explains most of the calendar year difference in Florida Auto, Commercial Auto and, to a little bit lesser extent, the California Personal Auto combined ratio differences.
Okay, that begs lightning round question 2. What's driving the year-to-date unfavorable loss reserve development?
For this, I'll start with our objective function with loss reserves. First, we want them to be adequate with minimal variation. By adequate, it means we do not want unfavorable loss reserve development, and, in fact, we're disappointed that so far this year we have unfavorable development.
Additionally, we want minimal variation. So we do not want large amounts of favorable reserve development.
If you remember, in both 2010 and 2011, we had significant amounts of favorable reserve development. In 2010, it was over $300 million, and last year, it was about $240 million. We actually do not want large amounts of favorable loss reserve development either because it creates problems in our pricing. So what's happened so far this year in the drivers of that 2012 unfavorable development? In the previous years, severities across most coverages have been pretty moderate, and -- including in the bodily injury coverage, which is where most of the loss reserves are. And as a result of those moderate and low severities, over the last couple of years, we've actually lowered reserves over $400 million to try to get closer to that objective of having minimal variation. So we would expect this year to have less favorable reserve development. We didn't expect unfavorable development, but we would have expected a lower absolute variation.
So what has changed? Severity trends have changed. I mentioned in the past few years, severities have been pretty moderate. But so far this year, through the first 5 months this year, severity trends have increased across almost all coverages. For bodily injury, property damage, personal injury protection and also collision, our severity trends are now in that 5% to 7% ranges, which are significantly different than the 1s and 2s that we've had in past years.
And in particular, our estimates of severity trends for bodily injury have increased, and, in fact, the 2011 accident year for bodily injury, our estimated severity is now about 2% higher than it was as of December last year. The unfavorable development in BI severity is what's driving a lot of the unfavorable development in both Personal Auto and Commercial Auto and is contributing to those calendar year results. And the final factor I would say in terms of the unfavorable development is in Florida for the Personal Injury Protection coverage. So far this year, we've had a significant amount of unfavorable development. I will say it relates to litigation environment in the state of Florida, which has reopened some of what we thought were previously closed claims. I would say PIP development has been pretty moderate in other states, but in Florida, it's been pretty significant. And that is not unique to us. It is an industry-specific environment in Florida.
So those are the major drivers of what are driving the unfavorable development so far this year. Again, something that is not desired. We remain vigilant. We're looking at our reserves each and every month, and we hope that it will not continue. Now, to provide a little bit more color in terms of Personal Auto, the next question is John Sauerland, can you provide a little bit more color on some of our large Personal Auto states and their results so far this year?
John P. Sauerland
Sure. Before I go into the state specifics, I wanted to address sort of rate level in aggregate, so Brian told you severities have been going up. In 2011, for our Personal Auto programs, we increased rates slightly less than one point. Year-to-date, we've raised rates in excess of 2 points, and we have plans to raise our rates 2 to 3 more points between now and the fall. So think of 4 to 5 rates, and 4% to 5% rate increase for Personal Auto year-to-date through fall of 2012. So we are taking actions to address the severity trends at a higher level, so aggregate rate level. We're also taking actions specifically at more of an underwriting level, and I'll cover some of those topics as we talk about specific states.
So first state, Florida, our largest state, about $1.5 billion state for us. We have great share there. Brian told you about the dynamic PIP environment. So there's been some court rulings that have reopened some claims from previous periods, and we're working through those. Florida will remain a dynamic PIP environment. As most of you probably know, there will be a law change in Florida that will take effect at the beginning of next year. It calls for a different means of handling PIP claims or different limits when you are in different situations. It also calls for rate filings before those changes take effect. We can't tell you exactly what all that means. We'll certainly play out over the course of next year. That said, we are fairly confident that in the course of handling claims well and pricing well, we can still hit our targets in Florida and grow.
As you can see here, while Brian told you the combined ratio is up in Florida, on a trailing-12-month basis, it's actually still slightly below 96. We have been taking rates up there. If we need to take more rate, we will. We're also doing selective verification at point of sale. So in some environments, in larger PIP states, normally urban environments, there are efforts to employ insurance for reasons other than insurance, and so we work hard to ensure the people that we are writing actually intend to insure.
The next state, our second-biggest state, is Texas, and hopefully, the picture is worth a thousand words. Unfortunately, a large hailstorm hit Dallas just yesterday again. The weather has been a challenge for us in Texas so far. We've been raising rates in Texas, and I'll tell you here again, I'm fairly confident that we can manage in this environment. It's somewhat uncertain with the weather patterns. With that said, we think we can continue to grow in Texas and continue to hit our targets.
Another state, not on Brian's list, but certainly a larger market, a little lower-market-share state for Progressive, New York. You can see from the graph that we have had some historical profitability challenges in New York. We are currently and have been beating our targets for quite some time now. We've done a great job, I think, in figuring out more of that point-of-sale underwriting that's necessary in New York. We have raised rates. We are growing in both our agency business, very nicely, as well as our direct business. And here's another market that we think we can grow a lot in going forward and hit our margins.
Next state, another PIP state, Michigan. Very challenging to handle unlimited PIP in Michigan. It's difficult to reserve for, handle the claims for and consequently to price for. We have seen marked improvement in our combined ratio in Michigan. You can see about a 10-point improvement. That said, trailing 12, we're still a little over 100. We will tell you that trailing 6 combined ratio in Michigan is significantly better than 100. So we think we are managing our way out of this. We will remain very diligent, insuring that we have good processes in place and good pricing in place for unlimited PIP in Michigan. In the short-term, we've been shrinking, which is the right thing to do when you're expressing profitability difficulties. But in the longer term here, we want to get back to growth and, obviously, profitability.
Final state, Massachusetts, not a top 10 state but a key one for us. I didn't put numbers on there, but I think the chart tells the story pretty well. We are now hitting our target margins in Massachusetts. That was certainly not the case for too long a period, honestly. That said, we feel like we have the product structured well now, and we're actually starting to grow premium in Massachusetts. So we think Massachusetts can be a profitable and growth environment for us going forward.
Then I'll hand it to Brian for question #4.
Brian C. Domeck
Next question. This relates to Commercial Auto. Written premium growth is pretty significant so far this year, so what are the major drivers? And John Barbagallo will answer this one.
John A. Barbagallo
Well, let me start by saying we are pleased that the return of profitable top line growth we saw in 2011 has, in fact, carried over into 2012. Now what's behind the growth is actually rather straightforward. The principal driver to the top line growth is we are seeing a very healthy increase in our average written premium per policy, about 13% year-over-year. The main contributor to that is the fact that we have been very consistently addressing rate need, really since the beginning, January 2011, and our aggregate rate take over that time has been 10%. The good news is that rate take is actually showing up. So it's sticking on renewals. Our renewal rates have held up very well through all of that, and it's showing up in new business as well.
Another contributor is the simple fact that general improvement in the small business economy. We're starting to see positive premium trend overall. We've actually had 10 consecutive months of positive premium trend. So our customers are beginning to buy more coverage, and we've seen a very nice rebound in the number of vehicles insured per policy back to prerecession levels.
A third and final contributor which really is driving new business average premium is we're seeing a slight mix shift toward our for-hire transportation sector. That is the segment we serve that carries the highest average premium per policy. That segment of the business, for-hire transportation, is the one that's definitely shown the fastest recovery overall. And in terms of the insurance market, it's the one segment that is really showing signs of firming up. So we're actually seeing some capacity lead that, and we're definitely seeing rates go up for for-hire transportation. So most of what we're seeing on the top line is this very nice improvement in average written premium per policy.
A secondary contributor, also important, though, is we all are seeing some nice year-over-year growth in new business applications, and that is being particularly influenced by the state of Florida. Florida is our largest state for Commercial Auto as it is for Personal Auto, and it's a state where we actually have been in decline for a while as we were addressing some profitability problems. We're profitable in Florida now, and we're growing, and overall, we're growing new business applications. I will say the growth in new business applications is being driven by an increase in submissions or quotes. Because we've been raising rates, and, we believe, raising rates ahead of our competition, our conversion ratios are actually flat to down, depending on the segment you look at. So it's actually growth in new business quote that's driving the increased new business volume.
With that, I'll turn it back over to Brian.
Brian C. Domeck
Next question. This goes to John Sauerland. Advertising spend is up, but yet we say direct quotes are down. Can you provide some commentary on these things?
John P. Sauerland
Brian, as a numbers guy, we give him a formula in response to that question. Cost per response equals the cost per impression times impression per response. First component there, cost per impression. What we're showing you here is a cost per impression from our largest external media buyer, Rubin [ph]. This is television costs. As you can see, it's been on its way up. Maybe a little flat during the late '08, '09 period when things slowed down. But it's on an upward trend and is projected to continue to rise. So this is -- this case we see pretty much across all media types, cost per impression is up.
Impressions per response, so how many eyeballs does it take to get someone to call or click Progressive? A little harder to measure because an eyeball and television not necessarily the same one online, on the billboard, et cetera, but a couple key things to consider there: our advertising spend continues to grow. That said, our share of voice, and this is just offline, so television, radio, et cetera, measured media by Nielsen, continues to drop a bit. So the industry or the category continues to advertise more and more at a pace outpacing our own, and so our share voice is challenged. That's the challenge for Progressive. Breaking through becomes even harder. That was why Great Creative becomes even more and more important.
Speaking of Great Creative, we've also been working in a complementary campaign with Messenger. We want to ensure that Flo continues to be fresh and invigorating into -- in our advertising, the Superstore, we continue to introduce new characters there. That said, we are employing the Messenger as a complement to that, that lessens our load on Flo and Superstore. Generally speaking, when you put a new campaign out, it takes a while to burn in and get attribution up for Progressive, so we think that is affecting impressions per response as well.
So net impression per response is also up. Cost per response, per the question, is definitely up. The great news is, is that we managed the cost per sale. We've walked through our economics around this with you previously in these meetings, but the short story there is we built in loads into our direct prices to recoup our advertising costs. And as long as the amount we're spending per sale is within a reasonable range of the amount we are recouping over time, we are fine spending on advertising. And if you recall, from the graph from the previous slide, we are spending more, actually high single-digits more this year than last year, on Personal Auto advertising, which means that we are very comfortable with our cost per sale. Not surprisingly, that means response per sale or the inverse of conversion, if you will, is headed down. Conversion is this graph for our Direct Auto programs, conversion has been going up. So the net of those 2 has gotten to us to a cost per sale -- continuing to have a cost per sale that is well within the range that we deem to be acceptable to continue to spend in our direct business.
With that, Brian?
Brian C. Domeck
Next question. I think we'd be remiss if we didn't have at least one investment question. Investments are big part of our business.
So Bill, Europe is in the headlines these days. What is our exposure to Europe, and any implications to our portfolio?
William M. Cody
Well, no slides, this is pretty straightforward. We have no European sovereign debt. However, we do have some corporate exposure. Our exposure to the periphery of Europe is less than 0.5% of the assets of our portfolio, and since it's only 2 positions, I might as well just tell you what they are. We have $42 million of Telefonica debt and $25 million of BBA preferred, on which we also have $25 million of credit protection.
In the core of Europe, we have $288 million of corporate debt. In the U.K., it's $163 million of corporate debt and $93 million of financial preferreds. Now these are all large multinational companies with significant assets and earnings outside of the core of Europe and should be able to withstand pretty much any negative scenario for Europe that you can imagine.
When I think about the impact for our portfolio, it's clearly something that we do think about. We've run scenarios on it. We're prepared by running different scenarios for that, all of which have some increase in the price of safe assets, which is becoming a smaller and smaller pool, and a decrease in the value of any other asset. But the best thing we can do to prepare for that is really just to make sure that we own fundamentally sound securities in a reasonable size such that any price decline is manageable and to have a cushion. And right now, our cushion is approximately $4.5 billion of cash in U.S. treasuries. That should provide us some balance and flexibility in most of any outcome.
Brian C. Domeck
Okay, final question, and then Glenn will join me.
What is our assessment of our capital position?
In Glenn's opening remarks, he used the word solid. I'd use that word as well. Also strong. We ended the month of May with $8.3 billion of total capital, that's debt and equity capital. I would call that comfortably above that which we are required or need for regulatory purposes, also for our own internal assessment of contingent needs and also at a level to help fund the growth as accelerating, which we're quite pleased with. At the same time, we remain committed to returning underleveraged capital to our shareholders. Last year, at the end of last year, we actually accelerated share repurchases towards the end of last year. We did over $600 million in the last 6 months of this last year. So far this year, share repurchases have been pretty moderate. But we also had a variable dividend that was paid in late January, early February, that was approximately $250 million. And we had $350 million of debt-debt matured in January. Right now, we believe our levels are comfortable for where we're at. And our philosophy of returning underleveraged has not changed at all.
And with that, I would say lightning round has come to a close, and we will now open it up to the more formal Q&A, and Glenn will join me here on stage.
Glenn M. Renwick
Actually, let me start -- I know we're running just a little bit long, but we'll certainly be around for plenty of questions. Let me start with a couple of announcements that I think relate to some things that we've talked about today. The first is a, what I call a strengthening of our relationship with one of our key PHA partners. Earlier this week, we signed a letter of intent to take a non-control equity position with ASI, and I would tell you here, think in terms of this as a market positioning statement. We want to be able to come to our agents and provide them with a very strong assurance that this isn't just 2 companies coming together conveniently to offer products in the agency channel. This is something considerably more than this. This is 2 companies who are very committed to trying to provide a great bundle of 2 superior products to our agents. And by taking a little bit more of a equity interest in the company, we hope we send a very strong message to our agents that this is something that we and ASI are very committed to. This is a natural outgrowth of 2-plus years of working together, and frankly, the synergies between the companies are incredible at every level. ASI will also invite me to join their Board of Directors. And for us, this is just taking that PHA concept, just turning the dial one more notch and when we first outlined it to you on Progressive on a Page [ph] or at least updated you with it. We said reasonable integration. Well, it's time to go beyond reasonable integration, it's trying to get something a little more. So this is just one step, they're not a public company. So that's all I intend to say on that today. The investment is not the material announcement here. The material announcement is what this can mean and what we intend to make it mean for our agents. So we're very excited about that.
Second thing, John touched on some of the relatively positive issues we've had with our intellectual property as it relates to Snapshot. Clearly, there is an awful lot of interest in Snapshot. We're well aware of that. Both John mentioned in his comments and I've been very consistent in saying this is all about execution. This isn't about sort of, necessarily, patent protection and so on and so forth. However, we have put many, many years of research and development into it. We have taken it to market in different forms. Some of those forms didn't necessarily produce the kind of profitability we look for today. We expect to have the benefit of that patent protection to, in fact, allow us to benefit from our own discovery. Pretty much the way the constitution drew it up.
However, we're very well aware that there's a great deal of interest, and we want to be very respectful of not only our industry, our regulators, clearly, our shareholders. So we've thought of this, we'll try to think of this from all sides. And I'm going to announce today that starting in the second quarter of 2015, we will make available to all who are interested a license to our intellectual property. We will not be licensing our algorithms, just the right to the intellectual property. And we will announce later this year what the basis, what the reasonable fee or licensing fee will be, and we'll get that out so there's no ambiguity in the marketplace for that. And clearly, and to be very clear, I hope we will structure this so that others will see that, yes, we deserve, we think we deserve some reasonable time to bring this to the marketplace. We'll provide it to the industry. We'll support our regulators who have supported us, and we'll try to make the fees such that we believe, at least, the cost of the fee is a whole lot better than fighting.
But in the meantime, we intend to enforce our intellectual property right as we see them. So just a couple of announcements.
All of the people who have presented today are mic-ed, and I think with some technical wizardry, we can all get them involved in questions if we need to, but why don't Brian and I try to take most of them?
Can you -- 2 questions. In terms of the Snapshot patent protection, how convinced are you that there won't be a decent workaround over the next couple of years through some other type of technology, mobile phones, something that you could use in the car already? And secondly, on the mobile apps and the time you're spending in mobile, what are the pros and cons of that? Are there -- does it directly benefit direct insurers more than all insurers? And is there any risk to, let's say, increasing frequency of changing through policies? And can you go over the pros and cons of your mobile strategy?
Glenn M. Renwick
Sure. First, with regard to mobile, are there any risks in different technologies come up? Yes. There's always risk. I alluded to it. We clearly have a cell phone app that does a lot of the sort of things that we know we can do. So we're very well aware, and we have different alternatives. We choose to use the one that we have and what John outlined. Let me say it again, it's all about execution. And no matter who comes up with a new technology, a lot of that will appear to be meaningless scatterplot of almost sego [ph] like situations, those were our insurers. It's massive amounts of data. What really has been the issue here hasn't been so much the technology curve. We knew that would come to us. We're not in the technology business, we're in the data business. So yes, I expect there'll be all sorts of things that will develop over the next several years. But I'm not -- I can't concern myself with what I don't know. I'll concern myself with what we do know. And could you just repeat your second part of the question?
Just in terms of does the cell phone and mobile apps, does that disproportionally benefit direct underwriters like you and GEICO? And are there risks to it, everybody doing this and people -- retention going way down for the industry?
Glenn M. Renwick
Yes, certainly, you've heard today, clearly, we -- and I've stated it several times, I think we're in a very advantaged position, having both an agent distribution and a direct distribution. Lots of times, people sort of feel that perhaps we control the distribution. We don't. We have to appeal to the way consumers want to shop. And I think what you heard from Jeff and Matt, there's -- the consumers who are using the mobile device, they would more likely be direct, of course. But that doesn't mean some of the things that Karen presented to you that we won't be able to take in some way shape or form and help our agents as well and keep them relevant in the mobile world. So I think that's one that you'll see, just as you saw with -- if you're around, our internet strategy we developed over a good number of years, we'll just keep that one rolling. The really key message today is we're not going to hedge our bet on this one. We're not going to wait and see. We intend to lead on this one.
I had a question just on Snapshot. So just quickly, could you remind us what you measure today with Snapshot for the driver, the data? How that evolved, clearly, you looked at a lot of things, and there was pushback from privacy issues. Where is it going and how do you change the metrics that you measure? And do you have to wait for renewal to change the metric, saying, "Okay now, we're going to measure this."? Because you're going to get pushback from the consumer. And then can you help us at least frame what the payback is for this new -- I mean, you give it a discount. What's the payback and how does it help the EPS, over what time frame?
Glenn M. Renwick
Why don't you sort of build on John's comments for the EPS question, if you can?
We are always revising our algorithms based on the data set that we got. You mentioned a privacy issue, and I think we've been very forthright with that before. We actually took out GPS, for example, from the chip. So do we believe or know that a GPS could help us? Yes, probably. But right now, we have so much marginal gain in the data that we have that we're very comfortable that we're not even exploiting the full depth of that. You've heard John refer to that. So we're not dramatically changing the variables that we're measuring. Last year, Tom talked about horizontal axes. So lane-changing behavior, something that's certainly is in the scope of the types of things that we'll consider. But we've been pretty public about a few of the things we measure, time of day, actual usage and braking and acceleration behavior. Beyond that, we really don't need give any away to anyone else as it's somewhat proprietary. And to the extent that when someone's got a renewal, they have, if you're one of our customers, you will have now pretty much gotten your discount and returned the device. So we're not -- we would never -- this whole thing will come down on its ears if we don't have the integrity to live what we tell our customers we're doing. So ultimately, once you've got that discount, unless we came up with something fundamentally new that you think that you can benefit from, and we do reserve the right to resample your behavior, and we, of course, we'll manage that. No one could expect that whatever we knew today is all we'll know for the future. But realistically, it's not that dynamic from the consumer perspective.
So just on the EPS question. Just can you help us frame when we're going to see some payback?
Brian C. Domeck
Can I give you an exact number? Here is what to put into earnings per share, et cetera, no? But here are the numbers. Tom referred to -- it's over a $1 billion of our Auto premium right now. So now it's going to be 7%, 8% of Personal Auto premium and growing. Growing nicely in Direct, and we think there is a big opportunity to grow it in the Agency channel. It's still relatively new, and the adoption rates are lower there, but we think the opportunity is there. And as more and more consumers become acceptance of it, try it out, et cetera, we believe there's growth in the agency channel from this as well. Tom also referenced this meeting, particularly the newest model, and the data time reference was the model we introduced sort of about 1 year ago, it is meeting our cohort targets. So -- and we measured things relative to new and renewal targets, sort of comfortable there. Previous model iteration, I would say didn't exactly meet it. So that still has to earn out a little bit, but the vast, vast majority of the business we're earning right now is on a new model and meeting targets. And I believe the big win, a -- one of the big wins is the retention gains. The retention gains, the PLEs of this business will continue, we believe, potentially, even to be higher than our own estimation. So the lifetime additional earned premium, the lifetime additional underwriting profit will play out over time. And the final thing is with the test drive. If we can get more consumers to just check it out, we think they're going to be pretty positively influenced, and if that generates more new business quoting activity with us, we think that's a good thing. Can I give you exact numbers? No.
Just last question, so then, you would assume, then, you might be attracting the best drivers. Does that correlate to getting the best homeowners?
Brian C. Domeck
In terms of the best homeowners risk profile? Well, we -- Tom has sort of mentioned there is a little self-selection process in Snapshot where good drivers tend to think of themselves as good candidates for the product, which is great. And more of those tend to be homeowners as well in some of the preferred segments. So whether they will end up being better loss ratio business for homeowners insurance, I don't know that per se, but the combination of being very attractive to the people that have multiple products helps increase that lifetime premium. Well, the big thing for us is that lifetime premium of a policy, not just the current period. But if we can generate much more lifetime premium, that leads to future years' growth rates. And we're quite pleased so far this year at 7% premium growth. But as Glenn mentioned, we wouldn't mind double-digit premium growth, and given our capital position, we can fund all that. So that's how I'd answer that.
Glenn M. Renwick
Well, the point on the homeowners -- well, last year was a pretty tough year all around. Certainly, the carriers that we did business with, I think, fared very well. Now I realize that's a function of pricing, but relative to the price, they're not getting a loss ratio return from the business that we're putting on that's sort of unexpected.
I have a question on the BI severity issue. First, wanted to know if this is Progressive-specific or this is industry-wide, increasing the severity that you're seeing? Secondly, how confident are you that a rate changes you're taking, the rating position you're taking are sufficient put you back on top of that issue? And then finally, I wanted to know if you expect any margin improvements based on the ratio you're going to take from now to the 4? And then I have a second question, which is -- one more, you're gain share has been increasing even though the profitability has been declining. Wanted to get a sense for whether the waiting between the profitability and the growth is changed for the gain share calculation?
Glenn M. Renwick
You want to take the industry general severity question?
Brian C. Domeck
I'll take as many as I can, and then you can fill in. First, in terms of industry severities, I would say yes, we have started to see more and more competitors talk about higher severity changes. Some of it is not just isolated to bodily injury, I should mention, though. Some of it is also in the property damage coverages, where more and more companies are seeing severities are going up. I would say our 5% or 6% severity increase in bodily injury, for us, that's a year-over-year in number. And last year, our BI, bodily injury severity, at least what we estimated then, was smaller or much more moderate. So some of it is the year-over-year versus the denominator. But I would say in general, industry is saying severities are going up, and I think that's pretty consistent. In terms of rate level changes, I'm pretty confident -- we're structured where there's product managers responsible for individual states. They were very concerned and look at their profitability in growth relative to internal measures as well as competition. So I think there's a lot of diligence around rate level. And in fact, I think some of the rising signs of severity increases -- and I should mention our bodily injury frequency is actually up so far this year, which is a little bit changed as well. Product managers are becoming aware of it, and the level of rate activity in the last 3 or 4 months has been much more accelerated than what it was, say, in 2011. So we're cognizant of it. We remain vigilant, and we try to act fast. That factory that John referenced enables us to get rate revisions out in the marketplace. That does take some time to earn in. The one thing we are, for Personal Auto, 6-month policies, so it, fortunately, is able to earn in faster. Specialized products and Commercial Auto more of those are annual terms. So it takes a little bit longer to earn in. And then am I confident we will over time meet our targets? Yes. Yes, because that's what we do, and I think we do it consistently and exceedingly well. And lots of people are looking at it, including both loss reserving, marketing claims, et cetera. Final question, on gain share. And I've heard it from a few folks. Gain share has gone up even though the combined ratio has gone up as well. How does that all fit together? I would acknowledge it's not all intuitive but remember, the aggregate gain -- we report one gain share factor, but it's actually a -- there are 5 different components to that gain share factor. There's one for Agency auto, there's another one for our specialized products, one for Commercial auto, and 2 for our Direct auto business. So that aggregate one month we reported is the aggregate of the 5 component pieces. So in any given month, some of it may go up, some of it may go down. It's an aggregate. What drove May's going up relative to the higher combined ratio is actually our Direct auto performance was actually better in May than year-to-date, both in terms of our profitability versus new business targets, which is 1/3 of the direct score, and the second component score is our estimate of our retention in Direct, that's continuing to go up. So we believe and we know we're keeping more customers this year than we did a year ago. So that drove up the Direct Auto component score for the month. I should note, Agency combined ratio was higher than targets. Their score went down. Commercial went down due to the higher combined ratio. The aggregate score -- there's 5 different components, so some could go up, some could go down.
If I'm an agent, why would I sell Snapshot? Aren't I going to get paid less if my customers are getting a discount?
Glenn M. Renwick
I've had that conversation with a lot of agents, and quite frankly, it's almost a good conversation. I go, "Well, recognize when we sell perhaps at a discount, that affects commission." They get that. They get -- more than anything else, they want a viable carrier in their business. So while we'd all like to have more business and higher commissions for agents, they certainly get that. So the fact is, you're right, they'll get lower commissions, but they will have a product that literally only can be had from Progressive. And puts them at an advantage. We will also be doing television advertising that promotes Snapshot. And our agents are telling us more and more that the brand is helpful to them. So, yes, there's always trade-offs, but that is not one, and I'm speaking from very personal experience. I'll see another large group on Tuesday. That is not the issue that is paramount for agents, a reduction in commission. They're more excited about the fact that this is now what I call commercial-grade and that they can put it to their customers without the fear of coming back 6 months, a year later and saying, "Hey, it didn't quite work out the way the company expected. I've got to rewrite you," and so on and so forth. That's a great concern for agents and one that I happen to share.
Brian C. Domeck
The other thing I'd just add to that, I think agents also think about the lifetime commission. And so the policy life expectancy of that customer is longer and stays longer, the lifetime commission from that customer and being in their agency might not be lower.
And then when we think about the customers getting the discount, what percent that apply are getting it in both the Agency and the direct channel, and then talk a little bit about the retention of those who are not getting the discount that apply for it in both of those channels.
Glenn M. Renwick
I don't think we have any more specific data to put to there. Actually, Tom referred to it as money slides. So we actually saw the retention of people who, by our determination, are not entitled to a discount, which is also euphemistically saying they probably could have or should have had a small surcharge or a great surcharge, they don't tend to stick around. But let me be very clear here. Long term, we want a rate for everybody. Let's not be confusing here. Right now, we are introducing a very powerful notion on top of already an existing product. Over time, and this is going to be just stylistic, if today, just for example, simplification sometimes helps, sometimes they hurt, our base rate was 100 and we can discount 30%, it might be over time, Tom's dots going up very slowly over time, because this cannot happen quickly. If we can take base rates up to 200 but have a 65% discount, so I can still get that individual that would've gotten 30% discount, they get it, but I've got a great deal more scope now for the people that fell in that 0 category. So just an important point as you ask that question, we're not seeing sort of any great concern from people who are staying at rates that we would otherwise like to have surcharged, but it's a risk. And from an agent perspective in terms of who gets a discount and who doesn't, we don't see any dramatic difference. You heard earlier today that agents tend to put some of their better customers with us, so the mix is even better.
I'm going to be greedy here and ask 3 questions. The first, in recent months, the Commercial auto business had -- the results have been fairly erratic. Is that a function of the underlying loss trends changing or a couple of very large settlements that happened to fall in like May and it would have been March? So that's one question. Second is you talked about going after these bundled customers that have been long term with, say, an Allstate or a State Farm or Travelers, I'm just wondering if you could give us a rough percentage of those kind of customers who are actually in your PHA program? I'm assuming it's a pretty low percentage. And last, just the cost of one of these new concierge centers.
Glenn M. Renwick
You want to take the first one?
Brian C. Domeck
The first one on Commercial auto, just given the limits profile of that business, which has, on average, higher bodily injury limits, there is going to be a little bit more variation in terms of those results, and we expect that. But we've priced for over a longer period of time, and some of the variations, say, from April to May and March, can be from higher-severity bodily-injury-driven events. The other thing in the month of May that drove it, we actually had some reserve increases in Commercial Auto from the actuarial perspective in the month of May which contributed to that May calendar year combined ratio.
Glenn M. Renwick
One of the things, and I think you know this, John, but producing results monthly is great in one dimension, but it also gives us all something more to worry about. And you're never going to hear me dismiss or suggest not to worry about things. I worry about them. But it does give you some insight into our business. It doesn't run on dead straight parallel rails all the time, and we're going to have some ups and downs. And certainly, as we get closer to a 96, there'll be sometimes we go above that. If you think about hitting a 96, there's always going to be some degree of error function around that. And we talk in these meetings in terms of aggregate. What really matters is the slide John Barbagallo showed you, it's 300-plus rate revisions. There may be more this year with some of the changes that we'll need to do. So it's all very small steps, every one of those has a small error functions. Those error functions for, again, the stat geeks, that should be central limit there. They should be reasonably normally distributed, so we should be able to target our 96. We know what we're doing, but it doesn't mean every month will come in exactly on target.
My memory must be going. Second question is, prompt me?
The final in the bundled environment.
Glenn M. Renwick
I Yes. I think we're probably -- that's a fair question. I don't know the numbers off the top of my head. I'm sure we have them. Let's be honest. We are a relatively small player in the bundled environment. What I think we're trying to give you today is a sense of how Progressive has gone from being a monoline writer, very successful, to something that now can be a destination writer. Easy to say, but I wanted to give you more proof points today of why that could be the case. So certainly, we attract people that have been with other carriers. We keep the proof of prior that we know we're attracting. But frankly, that's still a small story to really talk about at this point. Certainly, one that we'll be willing to talk about later. As to the cost of our claim centers, they cost a lot.
Brian C. Domeck
I'll provide a little bit more commentary on that one, which is okay. One, certainly land cost in different geographies are very different. So we'll take land aside. The construction cost of the facilities themselves, call it, $4 million to $6 million apiece. But we've changed a little bit in terms of the model from where we first started. All new centers will likely have a claims office co-located with it. So in the past, we have had claims offices, which we primarily leased, but future claims service centers will have offices co-located with it, which helps on, actually, the real estate cost, and we get leverage from the real estate cost by doing that. It also creates a little bit more flexibility in terms of the workforce handling the work at the service center, so we see that as a benefit. And over time, we have learned as to how big to build these. And the reality of it, certain of our early service centers, we just built too big in terms of the number of bays, et cetera. We've gotten smarter, more diligent about that. So the future service centers and any ones that we're retrofitting now are smaller in terms of the number of bays, have co-location claims offices, so the real estate component actually is not as large as that $4 million to $6 million I was referring to.
My question is in regards to reserves. So in the 10-K, it says that the implicit assumption on severity blended between, let's say, Personal and Commercial is, let's say, percent in terms of severity increasing. And I think, Brian, you said in the lightning round that -- I'm assuming it's a couple of percentage points higher, so let's say 3% year-over-year increase. In the 10Q, you've talked about severity being plus 5% to 7%. So if what I said is correct, are you assuming that severity gets better? And if so, why?
Brian C. Domeck
I think there's 2 different pieces to that. The 5% to 6% to 7% severity that I referred to in Q&A are based upon this year's losses relative to the same period a year ago. So it's a year-over-year comparison. In the K, what we talk about in terms of the accident year severity, in with that, we said bodily injury then at that point in time was relatively flat for 2011 versus 2010. Now what we're seeing is that, that severity estimate should have been more like 2% or 3% relative -- our estimate at last year end. And that's what's driving some of the unfavorable development. What we know now is we probably should have had a higher accident year severity for 2011 for bodily injury coverage. That's the bottom line. And this year, severity trends seem to be a little bit higher. I should mention that, that 5% severity that I mentioned for BI includes some of the reserve changes that we've made. Our paid severities is a little bit less than -- our paid severity so far this year is about 3.6%. But severity trends, we think, are increasing and are increasing relative to what we saw in the last 2 or 3 years. There's no doubt about that, which means for us, we need to be pretty committed to getting that rate revision factory going when we need to. And it's not every state. It varies by state. We have that capability.
On the severity question, I guess, you must have some theories as to why severity is going up. It just seems to be very risky to be growing at a time when you're supposedly -- I guess you're being surprised by severity trends. My second question was actually on Snapshot and the 38 trial that you're talking about. Have you given any thought to try finding a way to actually putting the Snapshot piece into people's hands? Because it still seems like it's a voluntary process that you're asking for. And I actually did have a third quick question is, did I see the mystery man selling blue jeans? Is there something wrong between him and Flo?
Glenn M. Renwick
Let's start with the severity issues. The answer is, we always have theories as to what we're seeing. We try to track as many different things, some are more tangible than others. Clearly, in the physical damage side of things, we know what's happening in the used car marketplace and so on and so forth, so we can relate to those. Bodily injury sometimes can be considerably trickier. I don't want to be in any way unresponsive to the question, but I would tell you 99% of what we do is try to make sure we're measuring it accurately, then try to see if, in fact, it is a real trend, because obviously sometimes you get data points and they don't quite trend in the right direction, and then act as quickly as possible. So it's not always the most fulfilling answer to say sometimes we don't always understand what's driving trend, since many of these services are provided by medical institutions and so on and so forth. But we must price for it. So we're much more attuned to measuring it and pricing for it. We're much better at that than we are in explaining it. I remember the blue jeans or Gap or whatever, I assume that's the character act to doing another ad for another company?
Yes, yes, yes.
Glenn M. Renwick
Yes, it is. And your second question again?
Brian C. Domeck
But there is no conflict between the Messenger and Flo. That's not what it was.
Glenn M. Renwick
That's not what it was?
Glenn M. Renwick
Yes. We've got -- what John outlined today was really a pretty big announcement. I mean, that's a really big announcement. We're trying to take this to a mass audience. What he showed you was sort of a mass media introductory television campaign. Jeff used the words in several different settings more related to mobile but more of an integrated campaign. So expect to see something like last year when we introduced Snapshot, a more of an integrated marketing campaign to try and get the Snapshot devices in people's hands. And while we're not ready to sort of talk about anything specific, we definitely have some discussions with nontraditional suppliers. I think you heard from Tricia earlier that, certainly, even our own claims people may have the opportunity to be able to -- small number, but may have the opportunity to be able to say to someone, gee, you just had a great claims experience. I don't want to lose that instantaneous goodwill. You want to take a Snapshot, take a test drive right away. So I think we have sort of an array of ways that we can get this into people's hands and a marketing challenge, and frankly, I would tell you, my opinion, we've been a lot better at taking those marketing challenges on in the last few years. So expect us to see us be creative on this one too.
Staying on the issue of loss cost, can you discuss the potential for frequency to start the increase you're seeing an uptick in miles driven? If that would be a driver. And then my other question is probably more broad. In light of growing as fast as you can at a 96% combined ratio, taking into account that auto insurance is short-tail lines, given the low interest rate environment, is 96 enough?
Brian C. Domeck
I'll talk about the frequency. Most of what we have seen in terms of loss cost trends so far has been the increase on severity. But I did mention that so far this year, bodily injury, frequency has gone up. And Glenn referenced in his first quarter shareholder letter the combination of rising severities, and if frequency goes up, it can be pretty dynamic at times. Our other coverages, like property damage, frequency is relatively flat. So BI being up with probably damage being flat begs some more questions, but those are the stats. Collision frequency is actually down so far this year, but I would caution there, I think we all benefited from very mild weather in the first quarter in lots of the Northeast and Midwest states. So not only did we see much lower collision frequency, I'm certain the rest of the industry did as well. Comp frequency so far this year actually is a little bit less than last year, and it really relates to GAAP losses were actually higher last year than so far this year. But the troubling one or the one that we need to remain pretty vigilant on is that bodily injury frequency if that goes up. Because again, particularly with relationship to loss reserving and loss reserving accuracy, bodily injury is a major coverage, and then the secondary coverages would be UM and personal injury protection. Property coverage is like property damage and collision and constants, they sell pretty quickly. You don't see a lot of development over time there. So it's really -- but we're watching that bodily injury frequency.
Glenn M. Renwick
With regard to your second question around the 96. It's not hard to engage in the academic discussion about could you take a greater margin in one piece of the business when the margin on the other side is not so great? We're not going to do that. We get it. We understand the question, very reasonable. 96, grow as fast as you can at a 96, is much more than just a goal. It is a cultural imperative to us. It's how we operate. Well, none of us know when. I'll assume that at some point these [ph] rates will come back to a little bit more of a favorable environment. In the meantime, Progressive -- and this is not a forecast, this is not spin -- but we tend to do well when the market is a little more disrupted. So I'm happy to get growth at a 96. If sometimes we don't get growth, I might be a little happier to find the [ph] ratio is a little bigger. You've seen that, just take a look at sort of the history, but from a pure goal perspective, grow as fast as you can at a 96 is our mantra. It is deeply seated in everything we do in Progressive. It's not sort of just a this week we'll do that. And don't look for new change on that. So let's hope we get the growth.
So question on Agency-written premium growth. Can you just kind of talk about the trends that we've seen there. Obviously, there's been a lift there recently. Is that the bundled homeowners product or something else that's been contributing to the lift in written premiums? And then would it be possible to talk about what was growth, either Agency, Direct or both, excluding Snapshot? Just to understand kind of what -- obviously, that number is starting to get more material. And I just have one really quick follow-up after that.
Glenn M. Renwick
I'll just take your first question on Agency to start with. I would tell you it's mostly price. I've written about this, trying to get your pricing accurate all the time, the smaller bites of the apple, if you like, and sometimes the competitive environment, perhaps, reacts a little bit more. So we think we're actually positioned very nicely on price. You heard John say, "You may need a little more but not necessarily big chunks." So the biggest driver of our growth in the agency channel has been price and having a great product. And think you'd find genuinely that if you ask agents, they always like our claims, they like our technology and they like our service, and when price comes into play, that foursome really works very well. So we're in a nice sweet spot right now. I wouldn't tell you -- and certainly, a contributor might be some PHA, some Snapshot, but they would not be the biggest drivers of what you're observing today.
John P. Sauerland
The other gains on Agency-written premium, some is price competitiveness and new business, but also we're retaining more of our customers. Our PLEs are up and our retention, our decay curve is further up than it would have been 1 year ago or 2 years ago, so we're retaining more, too.
The Snapshot question?
John P. Sauerland
In terms of what would growth rate be without Snapshot, I couldn't tell you that. How much of Snapshot is incremental customers versus current customers, I mean, that's the base that we have internally, but to say what will growth rate be without the growth in Snapshot, I think that'd be hard to say.
And then just one question. You mentioned growing at a 96 combined as much as you can as a goal. Where in the pecking order of your objectives does growing earnings fall?
Glenn M. Renwick
I make a pretty clear statement to shareholders in every possible format that I do. I mean, it's not like we don't understand every other business measure. But having clarity around your primary business objective and expressing it is what we do. Grow as fast as we possibly can at a 96. You can derive from that based on our capital position, return on equity. Clearly, we say we're not going to chase yield. We'd love to see that complement our earnings and underwriting a great deal more than it's doing, but we're not managing earnings. I know that you might like to hear something differently, but we're going to grow the company and we're going to continue to invest. And a lot of what you've heard today is those extended policy life expectancies won't show up in every income statement, but it's starting to build the asset value of the company, and we're really, really pleased to know about that.
I'm going to ask Jay Gelb's 96 a different way. Glen, just trying to get your temperature gauge on the 97 on how much of a fever you think that represents? I'm not sure the presentation would have been much different if you'd reported a 94, 95. So I didn't sense that you're at a point where you think steps like cutting advertising back, dialing down are necessary. Is that correct read?
Glenn M. Renwick
I think that's a fair read. That doesn't mean we won't do everything. Let me be very clear. I'm not that happy with a 97. But the one thing I can be very assure of, we know how to handle these situations. Now whether we get the timing right of any severity or even frequency turn, that's a lot easier to see in retrospect than it is at the time it's happening. But we're on it, this is one thing we can do. So no, you're right, it wouldn't have changed this presentation at all, because that's instantaneous. We know what it is. We'll react to it. Frankly, we see some positives from that as well in terms of average premiums going up. But most of what you're going to get today is sort of our sense of how we're really trying to change our proposition to consumers if their [ph] either distribution channel and grow the business. So yes, we'll get the margins back in order. That's always priority #1. That's very tactical. And I think for those who observe Progressive over a long period of time, hopefully you've got a reasonably high comfort level that you know that we know how to do that.
So I'm right that advertising isn't a tool that you need to dial down?
Glenn M. Renwick
Well, you saw from John's presentation the sort of cost per sale. As long as cost per sale stays in a range that we're comfortable that we leave a price for [ph]. Now if we have to increase the prices to accommodate the cost per sale, John may or may not choose to do that. So don't assume it's not a variable we wouldn't play with, but also don't assume that, necessarily, 1 month over 96 or even 2 or 3 months would necessarily relate to cutting advertising for us. One thing I would say to you, if I can, if I can extend your question, is I said I feel very lucky that we've actually got 2 distribution channels, and that sometimes just seems like words. But we do have a situation where if the cost per sale gets forced up by the cumulative number of players in that environment, it's really nice to have an alternative distribution channel that's on a variable cost basis. So that combination is just very nice. Clearly, we want both to be growing, but it's nice to have that situation where we don't -- we're not captive to one channel and have to continue to generate demand regardless of what's happening to cost per sale.
I'm afraid that was the last question...
Brian C. Domeck
Just on advertising a little bit. We've talked about this before, some of our advertising is seasonal in nature. We actually advertise generally more in the first quarter than we do in subsequent quarters. So that's one influencing factor. And two, it is based upon the cost per sale relative to what our acquisition targets are, and we will watch to see how conversion plays out if we raise rates, and then we might have to adjust accordingly. But we will always remain diligent about how we spend our advertising dollars. It's a lot of money. The industry spends a lot of money. We've got to make certain it's smart. And if we think we should spend less on it, we will. But I do think it will be a dramatic decrease if we ever spend less.
Thanks for all the questions. I'm afraid that was our last.
Glenn M. Renwick
Thank you very much for coming. Really always appreciate your interest in the company. I think you can get the idea that we're pretty passionate about it. Just a quick announcement, my expectation, my expectation, I'll let you know, is that we're going to host this next year in combination with our annual general meeting. So that will be moved back to May. We'll probably move this one forward to May, and we will host that one in Cleveland, where we're developing a center where we feel it's much more conducive to quiet quality webcasts, because clearly, we've got to move in that direction. That's where a lot of people are actually now starting to do this. Obviously, you're all invited, but next year it's highly likely to be in Cleveland.