Good morning. It's always good to have a chance to tell the Progressive story, so thank you for coming. Each year, this time, we try to give you some sense of what's on our mind, what we think is important. We'll also try to, from time to time, bring other members of the Progressive management team to introduce, so we've done that again today. Ultimately, if the opportunity is appropriate, introduce some new ideas into the marketplace, and I think we did that a couple of years ago with Name Your Price, and we may have a couple of things today. Over the years, and I recognize there are some new faces in the audience, I think we've had the opportunity to cover a fair amount of territory, and hopefully what we say, we will continue to come back and reinforce of how we've done on those things.
Sometimes it's exactly as we intended, sometimes not quite as we intended, but ultimately, hopefully, you get the sense of real intensity around follow-up on the issues. As I think back over 10 years or so, we've covered Concierge Claims Service. We've covered our approach to reserving, our targeting of combined ratios. We've certainly done a fair amount of work on retention science and our activities around that. How we've started to target our customers a little differently. Some work on Net Promoter Score and how we use those as a diagnostic inside of our business. Certainly, of recent times, a little bit more work on our branding activities. While not complete, that's a fairly representative area of things that we've covered over a decade or so.
Today, we'll do a lot of that, but we're going to try a slightly different format, and hopefully, that'll give you a sense of just maybe how all these pieces come together. In your packages, you all have a blue document. Recognizing that readership on the screen is always a challenge, you'll have this to take away for words. This is something we call Progressive on a Page. Right now, it just looks like words. I hope, after a couple of hours, you have a sense that those words are really words that define Progressive uniquely and govern our actions. It was actually a fairly objective comment that was passed along to me that suggested maybe this would be a good format for a meeting like this. Regroup a little with regard to the activities we've been working on.
Please don't feel slighted by this comment, but this wasn't put together for this meeting at all. In fact, this was put together so we could present it, and the senior leadership of Progressive could present it to every employee in the company. The senior leaders that are with me today, and I'll introduce in just a moment, and many others, have actually used this to talk to every single person in the company about what we're doing and why, so they have a sense of what's expected of them and what their role does, and how it contributes to the overall company. Hopefully, at the end of the day, that this will mean just a little bit more than it does right now. Consistent with our prior practices, we don't intend to reformulate any numbers that we've already shared with you.
Hopefully, you get those in a timely fashion, we're not going to reorder them and do anything with them. We'll talk more about our activities in the company, what we're doing, how we're thinking. When I think about strategy, I guess that's a word that we use, all of us, probably pretty commonly. Question is whether it's the same thing. I think this way. Being abundantly clear about what we're doing and why. Almost by virtue of being very clear about what we're doing, we're equally as clear about what we're not doing. Then for us, it's very much about making sure that everybody that's involved in executing understands it and understands their role, and that's what we've used this for.
It's actually been a very insightful exercise to go through the company and very rewarding, I think, for those that it's taken us about a year to get that done. Clearly, it's not quite in the same format as you'll get today. Hopefully, by the end of the day, what appears to be words right now and words that could arguably be used for other companies in the same industry will have a little bit more meaning. Again, as I said earlier, uniquely define what Progressive's all about. We should enjoy that. The way we'll do that is, to some extent, I'll try to narrate some of the story, at least around the key activities. My colleagues will explode the bullets, if you like, in some Not all of them.
We won't get time to cover all of them, but a good number that we think are relevant and important to what we're up to today. Let me just quickly introduce from your right to left. Brian Domeck, our Chief Financial Officer, is with us. By the way, in the back of your book, there are bios on each of us. Jim Haas, to his right our Director of Research and Development for private passenger auto. John Sauerland , our President of Personal Lines. John Barbagallo, our President of Commercial Lines and also responsible for our agent distribution and sales efforts. Those will all become relevant as they have different speaking points later on today. That'll be our approach.
Let me first start, if I might, by just a couple of comments that set a few of the things we've done over the last year in perspective, where I think we've tried to say we have aspirations and intensity around our activities, just how they've come about. Let me start with that for about five or 10 minutes, then we'll get into exploding Progressive on a Page. Just quickly, you all realize before this slide, there was the safe harbor statement. We all know what that means, we'll respect that during the course of the day. The purpose of this slide is not to be very convenient and array data that just puts us in a good position. Just a quick snapshot of the first quarter.
The real point I want to make here is while Progressive Direct and GEICO have been fairly long-term residents of the more desired upper right quartile it's great to see our Progressive agency business coming into that zone and starting to find some growth. We have said consistently we wanted to be very nimble. We wanted to respond to the market conditions. We didn't approach this position by going above our targeted combined ratios. We think of that as an asymptote. We've approached from below, but we've absolutely positioned ourselves very well. We have rate adequacy. We have a service of the business.
Right now, I recognize you have to take my word for this, just a lot of things feel right at Progressive in terms of how things are coming together, both on a pricing and a servicing basis, and that's really allowed our agency business to also prosper. Demand for our product is up. Great. Somewhat evident from the first graph, but it's very true. The exercise in intensity around branding has worked quite well for us in direct. We can never actually measure the direct strength of that. Name Your Price has been something that we've shared with you for some prior periods. That seems to be contributing to growth. You know the numbers on direct, very pleasing. So far so good. We're never, ever comfortable. Profitable growth is what we're all about, and we're always looking for ways to accelerate that.
In the agency environment, we'll talk more about this today. We've accepted that the environment has changed. We've talked to you in, I think it was 2006, about comparative rating and how the environment inside of the independent agency channel has really changed. We've accepted that, and we've adapted to that, and we've retailed our product into that in different ways, and ways that we think are actually putting us at a slight advantage. We'll continue to find ways to work within what we now see as a channel acting differently and play to our strengths. All of the things that we've talked about, and I know there's a good number of people in the audience that have seen us talk about retention. We've given you actions that we've done. We've given you the way we think about it, diagnostics.
We've given you a little bit about the science of policy life expectancy. We've talked about nature and nurture, all those sorts of things. I'm not going to talk too much about those today, but I actually hoped it would get to the point where I would have a really boring story about retention and that the graph would just be year-over-year over year better. The economic impact of extending policy life is incredibly strong. The good news is, so far, I am becoming really boring and happy to be so because these are graphs year-over-year of our policy life expectancy as a book. Now we break that down to a lot finer detail in terms of different customer segments, so on and so forth.
For right now, a macro view and a fair takeaway is the Policy Life Expectancy of Progressive customers, on average, are longer and continuing to get longer as we create reasons for our customers to stay. We continually work hard to eliminate reasons that might cause them to leave. It's as simple as that. Eliminate the reasons that they may leave and create reasons for them to stay, whether it's loyalty programs or the like. I would suggest to you that the percentage of our income now being derived from renewal business and the future life expectancy of customers in the book is starting to make Progressive a very different company. In my opinion, the earning stream is considerably more valuable and likely to be sustained over a longer period of time. Something we're very pleased about.
A simple line graph, I think there's a lot of power behind that. Demand for our product is up. Retention is extending. Good things. Clearly not endgame, but on the right track. Our capital is strong. Certainly, we've had a session in this room a couple of years ago where that was a little bit of a different picture. Actually just pulled out the same picture. Hopefully, a couple of you will relate to it. Recognize the zero line here is the capital above our statutory capital. At a 3 to 1 level, think of that as close to a $5 billion and then on up. The orange zone, I use the same labels as I used at that time, would suggest that we're in our Contingency, our self-imposed Contingency Layer for capital, and that would give us some concern to be in our Contingency.
At the time that perhaps we were talking about this a little bit more intently about exactly where we were, I used the term several hundred million above our concern level. That always was true. You can see the low point there back in 2008. You can see now that clearly we're in a very strong position, having had a very nice run and a consistent run of underwriting profitability and being able to put that back into the capital mode and clearly some recovery in the investment marketplace. You all presumably know that just last week we launched a tender for $350 million of our hybrid debt, and that's conditioned on getting a removal of a replacement capital covenant. We don't know any more about that at this point in time as to the success. That'll be next week. We're avoiding questions on it.
We just don't know any more. That's out in the marketplace right now. About the 23rd, we'll know a little bit more. Bottom line, a nicer picture for us on a capital basis. We'll continue to manage our capital in the way that we've consistently communicated, that when we have more capital than we can effectively use in the business, we'll find ways to return it to shareholders. We'll put it to a very practical use, I think the opportunity to retire some of our hybrid debt at this point in the market cycle is a very effective use of capital. We're on track for that, we'll know more next week. Also, our dividend, certainly, I don't want to start projecting or any guidance, all sorts of whatever caveats would be needed.
If the year continued on the same course and speed as we see ourselves at this point, we will be looking at the largest regular dividend ever paid from Progressive. The dividend that most of you know how it's calculated. You see the Gainshare score and the monthly results. It's actually on a course for being a significant number. The year clearly hasn't ended, so we're a long way from that and things could happen. It's interesting to note that our return of capital now, we actually have multiple methods. We clearly have our share repurchase. We have a meaningful dividend. We've used a special dividend in the past, and here we've got an opportunity to repurchase SCAP. Those would be three big takeaways.
Just a couple of points that are a little bit more macro strategy over many years, and I would say things that matter in this business beyond price. Price is clearly important, cost structure, everything that goes into price. A few things that we really, really wanted to do, and I'm just going to give you a sort of health check on those. First is positioning. In our direct channel, we said starting in 2000, I know I said many times we wanted to become the leader in the internet space. When I say the leader, clearly, other people want that position as well. I think we have been very credible in positioning ourselves as a company that really is a significant player in the internet space for auto insurance.
We recognized, I think last year I had the opportunity to say that Forrester Research had recognized our website as the best website in all financial services, not just insurance. We, just a couple of weeks ago, won another recognition from Keynote as the insurance space winner for website, and that's our 15 out of 16 consecutive wins in that category. In and of itself, that's not important, but it does give you a feel for it's not just luck. Ultimately, what matters is if the dog hunts, and we're starting to see their Progressive production on quote initiation be in a position with GEICO to really say a significant position in the internet space, and we're very, very happy with that. So far, so good. Clearly, this is going to sort of explode into the mobile space. We're going to be looking at very different opportunities.
This is a place that the talent in Progressive and the resources available are something that take a special place, and we want to make sure that we're really set up to succeed. The agent retailing, actually, John Barbagallo is going to talk us through some of that, but really this is the issue of the environment changed somewhat. We changed with it. We think we became a better retailer. Frankly, our brand is not hurting us in any way with our agents, it's helping with our agents. Now agents or independent agents actually have a nationally recognized brand to sell in their agency. Certainly, while that's a hard one to put an absolute value or contribution to, it is definitely meaningful. Our reach is expanding.
I'm hoping that at least some of you remember, I think it was 2 years ago, that we used the format of this meeting to talk about sort of profiles of customers. They're not quite the profiles that we use internally in Progressive, but they're certainly good enough for this kind of a descriptor. We have a little bit more complexity to the customer tiers that we use, but we defined 4 tiers. Sam maybe is a little bit more of the classical non-standard. We're trying to move away from a lot of those terms. If I remember correctly, someone in the audience felt that Sam was a very strong profile to at least their brother-in-law or some other family member. Hopefully, that recognition is still there. The Dianes of the world, more upwardly mobile, a little bit more consistency in their insurance behaviors.
These profiles, by the way, are also in your appendix. The Wright family, certainly moving into a home, not necessarily a bundled family, the place that we're very strong. The real interesting thing here is the Robinson family. That's a much more complex needs family. Not a target that we, for the most part in the career or the history of Progressive, have really said our products are targeted at that group. We've done a lot of work. You've heard about our entry into home. Home Advantage, which is depicted a little bit by sales on the other side of the graph there, has given us an opportunity to really speak to that audience, and we're starting to grow the audience. Clearly, the numbers of 200% are a little silly because that has to mean we had very little to start with. That's not the point here.
The point is we now have a product that we can start to bring into that space. Even if that doesn't attract a lot of switches for people at that stage in their life, it gives us an opportunity to make sure that we create our own as they start moving through and their needs develop over time. That may be the more powerful point. Our reach is expanding. We've changed Progressive in many ways from a company that was extremely successful at a target group of customers. We've expanded that because it was consistent with our skill set and not our strategy. I want to make the point very, very clearly. The Sams, even though the percentages, in this case, this is policy enforced growth for the last 3 years.
While the percentage of our business might shift a little away from non-standard, we don't want to give up an ounce on non-standard. That's a very good business for us. We know it. We absolutely do not intend to be moving away from anything. It's more we're moving to something else in addition. Our reach is expanding, and it's starting to work. We feel very good about the opportunities to do that, the potential just seems so great when we realize We're not as good. We're not as good at the Robinsons, but we'll get better. Our brand. Brand health, and we'll talk a little bit about brand today. Brand health measures are always awkward for me because they're survey-based. Sometimes they're glaciers, and they don't really move, and then maybe you infer more than is really there.
What I look for when I see any of these measures is sort of meaningful deltas over any reasonable period of time and seeing if you could relate them to market actions. Here, and I'm really just going to focus on Progressive because any other measures we give you are just as a relative measure, but Progressive's all we want to talk about. I think we've really sustained a fairly significant delta in the last year. Our brand health measures have started to actually be certainly more than just noise. I think we've got some real signal there. Consumers are relating to Progressive and the Progressive brand in ways that we certainly didn't see only a few years ago. Our brand, I think I can actually conclude that the consumer perception is becoming more positive to Progressive and more well understood.
I happen to like the consideration one. Clearly, I ultimately want to see that show up in preference, but consideration now is actually a fairly strong measure and a nice delta year-over-year. Let me just quickly finish on these comments. I'm going to get a chance to talk about this as part of our strategy a little later, but it really is about the culture at Progressive. We've talked a lot about our actions, and I think we're very good at the math. We're very good at sort of the continuous improvement, the Deming type model. The culture of Progressive has changed, and here is just a slide to sort of at least give you a touch of that. Really, it's the people.
Yes, there's been a set of actions, and I think here, in terms of retention, we've talked about rate stability, Net Promoter Score. You can read the list. It's that combination of culture and actions that really has allowed us to become what I've called a true customer care culture. That's showing up a little bit. Again, measures always have their degrees of accuracy. This is a combination of value for money, where we've always done very well, and cares about the customer. If you put a combination of those two, we're now actually getting real recognition in the combination of those two, caring and value, that is meaningful. What's exciting to me is to observe the culture of the company really sort of making this a reinforcing prospect.
Just a minor example of this, literally, this is just one that sort of came to me, is that a few weeks ago, we had significant flooding in Tennessee. Apparently, we extended billing leniency to people that might have otherwise had their vehicle flooded and perhaps have other things to do to get on with their life. We actually ended up buying radio advertisements to tell people, "Look, we understand you've had a significant upset in your life. If your bill is due and you've got other issues, just let us know. We can handle it. Just let us know. You're our customer. We can work with you." The only way I even knew we did that was getting an email from a customer in Tennessee that was thanking me, which frankly, I didn't do anything in this case. "Hey, thanks.
I didn't need it, but I really appreciate knowing the kind of company I'm with." The point is, this just happened from the people who know what's expected of them and what the customers want, and that's the kind of culture that's starting to build on itself at Progressive. I expect this combination of value and caring is something that we care a great deal about, and hopefully, we can keep those in balance for a good long time to come. Let me move on now and see if we can make the Progressive on a Page sort of come to life in a sense of, yes, it's words, but what do they mean and what's the context for us?
I'm not going to go in order from top to left. Actually, I'm going to come back and summarize some of the more foundation statements a little bit later. Perhaps one of the things in the middle panel or close to the middle panel, you'll see that we want to be distinctive and competitive. You can read the words around that. Think as we go through today, is this a company that truly is trying to make itself distinctive and competitive? Ultimately, we don't want to be just another player in the space. We want to be a recognized player, and when you think Progressive, you think, I've got a brand perception of that company, and it's not the same as XYZ. I'm not talking about better or worse.
I'm talking about a clear understanding of who we are and what the character of the company is. Let's start with our financial goals. You know most of those, I won't cover the ones that I think are pretty well understood, but take a quick read of those and think about this as a top-line strategy. Again, clear communication, hopefully, to our owners. Manage to a 96 combined ratio. If you actually read our annual report, you'll see that the 96 is really a balanced blend of states, products, different maturities, new, renewal. It's actually not sort of just one number. It's a roll-up of a lot of different pieces. In fact, in your appendix, because we like to be very open about these things, I've given you a sense of how states are performing. That changes sort of monthly.
Not all states are making their goals. When they're not making their goals, we act. When they're making their goal, great, we try to grow. If they're not, we don't grow. We get it right. The 96 that we report is really very much a balanced blend of all things, all channels. In the annual report letter this year, I wrote that under certain scenarios, we'd be happy in our direct channel to report a calendar year 97. I somewhat predicted that in the first conference call, we might get a question on that. We did. We said we'd talk more about that today. I'm going to ask Brian Domeck to come up and sort of suggest why that isn't inconsistent with the longstanding 96 goal that we have.
Thank you, Glenn. Good morning. Managing to a 96 combined ratio. It's been part of our financial objectives for many, many years. What do we really mean by it? Glenn in the shareholders' letter this year provided a little bit color, hopefully, this morning I'll just add a little bit more color to those statements. Included here are excerpts from the shareholders' letter. I'm not going to read the complete statements, but I would like to provide a few of the key messages. First, our goal of an aggregate 96 calendar year is unchanged. We've talked about a 96 combined ratio for many, many years. Certainly for me, it was ingrained in my early years in product management and controller roles. I've heard it for a number of years.
It's simply, we seek to achieve at least a four point underwriting profit in any calendar year. That goal is unchanged. It has served us well, and it is unchanged. Second, for our variable cost businesses, commission-based businesses, think of those as personal auto in the agency channel, our specialized products, and our commercial auto products. For each of them, a calendar year combined ratio is an accurate and appropriate measurement scheme. For each of those businesses, we have a calendar year combined ratio objective that is at or below a 96. I'm pleased to report that so far this year, each of those businesses are meeting or beating their calendar year combined ratio objective functions.
Finally, for our direct business, particular our direct auto business, under certain high new business growth scenarios, we'd actually be happy to report a combined ratio over 96, as long as we are meeting or beating predetermined new and renewal combined ratios that will ensure that we achieve a lifetime combined ratio of 96 or less. As Glenn mentioned, we've had a few questions and comments regarding that statement in particular. Today I'm going to focus most of my time on trying to explain that statement a little bit more. Keep in mind, we manage all to an aggregate 96 combined ratio. Why do we think of the direct business a little bit differently? I'll try to explain this a little bit more with what I will call a very simplified example.
It's really dependent upon the cash flow differences between the direct base business and our variable cost businesses. Consider the following. This is two policies, personal auto policies, one in the agency channel and the other in the direct channel. For both of these policies, consider that they are in force for four six-month terms, two years in total. For both of these policies, the lifetime premium is the same. $1,000 each term, $4,000 lifetime premium. Also for this example, consider that the loss and loss adjustment expenses are the same, as well as other expenses associated with these policies are exactly the same. What differs is in terms of the acquisition cost, and in particular, the timing of the acquisition cost. For the agency channel policy, we pay, in this example, a 10% commission each and every policy term.
$100 for each of the policy terms, in aggregate $400. On the other hand, for the direct channel policy, the acquisition cost in this example is also $400, but that is all paid upfront to acquire the policy. Keep in mind, we've expressed this before, we expense our advertising costs as they're incurred. We don't defer our advertising costs. Actually, for our own cost accounting purposes, we allocate it all to new business. One could argue that you could allocate some to renewal business versus new business, but for years, we have consistently applied it all to new business. I'll talk a little bit about the pricing implications of that a little bit later. The real difference is in terms of the timing and cash flow of the acquisition cost. You'll see in both of these examples, they both achieve a lifetime combined ratio of 96.
What is different is that timing of the cash flows and how that translates is in terms of the differences between new and renewal combined ratios. In the agency channel, and this is again, a simplified example, the new and renewal combined ratios are much closer. There is a variance, but it's much more narrow. That's why we believe a calendar year combined ratio is an appropriate measure for that business. On the other hand, in the direct business, you can see that the new and renewal combined ratios are very, very different. Renewals with no additional acquisition costs, the margins are much higher. That's why we have talked for several years about the value of increasing retention.
The difference between new and renewal combined ratios is the reason why, and the fact that we spend the advertising upfront is why under certain high new business growth scenarios, we'd actually be pleased to report a calendar combined ratio over 96, as long as we are confident that we'll ensure a lifetime combined ratio of 96. You might ask, given the cash flow differences, how do you really figure out how much you're going to advertise for our direct businesses? We've talked about this before in the past that we do not have an advertising budget per se. We certainly have plans as to how much we'll spend in any given period, but we will adjust that according to what I'll call our economic yield. Let me explain that a little bit more. First, we have what we'll call target acquisition costs.
You can think of that as what is the amount of acquisition cost that we have incorporated into the pricing of the policies. That cost on a per term basis is really a function of policy life expectancy. For those policies that would stay with us and we think will stay with us for a much longer period of time, that per term cost is going to be much smaller. For policies that stay with us a shorter period of time, Glenn may refer to them as Sams or non-standard, the per term acquisition cost is going to be a much higher per term acquisition cost. We compare that targeted acquisition cost to our actual cost per sale.
Cost per sale is really how much have we spent in terms of acquiring policies. That would be media costs as well as quoting costs. Then compare it to the yield of policy sold. We calculate the cost per sale in an aggregate measure. We look at total cost per sale, average cost per sale, but we actually try to do it at much finer and finer levels of detail. We try to do it at media types, for example. A little bit later on in the program, John Sauerland is going to talk about paid search advertising. That is a media type where attribution of sales, attribution of a cost can be very fine. So we can more readily determine actual cost per sale for paid search advertising.
Other media types are not as kind and not as easy. We try to do the best we can in terms of trying to measure cost per sale at finer levels. I should mention in terms of cost per sale, we also do look at the incremental cost per sale. What is the efficiency of our last dollar spent? We want to ensure that our last dollars spent of advertising are as efficient as they possibly can be. A key question that we always have to answer, is our cost per sale less than our target acquisition cost? Is our actual experience better than what we had anticipated in pricing? The simple answer is, if the answer to that question is yes, we feel confident that we can continue to spend more.
If the answer to that question is no, we would likely reduce our spending activities. All of this is subject to the two constraints that Glenn articulated in the shareholders' letter. First, that the direct auto lifetime combined ratio has to be at or below a 96. We will measure the aggregate company-wide combined ratio to be at or below a 96. That goal and objective function has not changed. Let me try to provide a couple different illustrations to provide further evidence of how we think about these things. Consider this as a direct auto policy again. Again, it's simplified. Here I have indicated what we'll call target combined ratios. I've used the same target combined ratios for new and renewal business that were in the previous example.
They're illustrative and not exactly the ones that we actually use in terms of our pricing. In this situation, our target combined ratio for new business is 129. You'll notice that the renewal combined ratio target is 85. If 25% of the premium is new business in that four-term policy, 25% was in new business, that would equate to a calendar and lifetime combined ratio of 96. You'll also notice that we actually have targets at much finer levels of detail than just new and renewal business. We have targets at loss ratios levels, we have targets at loss adjustment expense levels, we have targets at expense ratio levels, et cetera. I should point out that we also have targets at state levels and tier levels. Think of it. We have hundreds of targets that we measure ourselves against each and every month.
In this situation, we're shooting for 129 new business combined ratio and 85 renewal combined ratio for an aggregate 96. Let's consider case 1. This is a situation where we are actually missing new business combined ratios. New business has 135, and we've missed our loss ratio target by six points. Not a good thing, and something that we would seek to remedy. Let's consider that this is a scenario where new business growth is at a little bit lesser levels than expected. Let's say, for this example, only 22% of earned premium is in new business. This actually equates to a calendar year 96. We're not meeting our targets, but equates to a calendar year 96. Case 2. This is a situation where we're meeting our targets exactly, both for new business and renewal.
We're actually meeting our targets along each of the subcomponents. A great outcome. Let's consider this as a higher new business growth environment, and 30% of our total earned premium is in new business. On a calendar year basis, this would equate to a 98.2 combined ratio over the 96 combined ratio you've heard for many years. Which scenario would we prefer? It's actually case 2. In this case, if we normalize and earned premium earns out over a period of time to be the expected 25%, case 2 will equate to a 96 combined ratio for the lifetime. Similarly, case 1 is actually not good. We missed our targets. Even though that current period calendar said 96, we believe over the lifetime, it would equate to a more equivalent combined ratio of 97.2.
That is why the case we continue to say we measure against targets, and under certain high growth scenarios, we'd be very comfortable with reporting a calendar year combined ratio or calendar reporting period combined ratio over 96. Just so you don't think this is just hypothetical, I'm going to share with you first quarter results. Now, the actual results are not in your book. You're going to have to pencil them in. I didn't want you to go to the punchline too quick. I want to share with you sort of the first quarter results. Remember, these are not exactly relative to the targets of 129-85, but to our true pricing targets. What happened in the first quarter? In the first quarter, we beat both new business and renewal aggregate targets.
In particular, we beat loss ratio and loss adjustment expense targets in both new and renewal business, and for that, we were very pleased. We actually, on renewal business, were slightly over our targets on other expenses. In aggregate, we still beat the renewal targets. In fact, in the new business, we actually were a little bit higher than our targets in terms of acquisition cost. Again, that was a conscious decision. You've heard me say before, we adjust our spend accordingly, and in fact, we spend more in first and third quarter in terms of advertising dollars than in other quarters. That's because there is a little bit of seasonality to shopping season. We actually consciously chose to spend more in the first quarter. We believe over time, as this earns out, it will meet that acquisition target.
For the first quarter, it was slightly over. In aggregate, this generated for our direct auto business a 97.2 combined ratio for the first quarter. I'm certain this is news to you because it's not separately disclosed in our monthly releases, it was a 97.2 in the first quarter. We were actually very pleased with this. Why? It was a high new business growth scenario. We reported in our first quarter the growth in new business, and it was very strong. Because we are meeting and beating both the new and renewal targets, we feel and are comfortable that we expect it to return a lifetime combined ratio of less than 96, and that is how we measure the direct auto business. I mentioned the aggregate 96 combined ratio objective function hasn't changed. What has changed?
Well, certainly for us, our advertising spend has increased, and you can see from this chart it has increased for the last several years. A little bit later in the program, John Sauerland is going to give a little bit more detail on that. Not only has it changed for us, but it has also changed for the industry as a whole. Advertising costs are up. We have actually coupled our increase in terms of advertising cost with what we believe an improvement in creative. I'm certain you have seen and are very knowledgeable of our Superstore campaign, with Flo as the main character, and we feel very good about how that expresses our products and services to consumers. Both the increased advertising spend and improved communications has led to demand generation and demand increase.
By that I mean our quotes for auto insurance in the direct channel are up, they're up on a year-over-year basis, and they have been for the last couple of years. At the same time, our conversion is increasing. Not only are we generating more quotes, but we're converting more of those to sales. Some of that is a function of some segmentation and product improvements from the marketing side, and Jim Haas is going to talk a little bit about that a little bit later. Certainly, we have continued to improve our retailing on the website of our prices. In a few cases, we've actually lowered rates in a few select states, all of those combinations have generated a higher conversion rate of quotes to sales.
Glenn mentioned our customer mix is changing a little bit to what we would call higher retaining, higher policy life expectancy policies. We think that is a very good thing. All of this, the increased advertising, the increased demand, increased conversion, and increased policy life expectancy, is leading us to the growth that you see in our direct auto business. I'm pleased to say that while we're achieving that, we're also achieving efficiency gains. Included in your book is one of the key measures that we report on regularly internally, and we've expressed before, Policies in Force per Employee. So far this year is up 11%, and that is on top of a 9% growth that we reported last year. I'm very pleased by that. We need to continue to try to continue to improve our efficiency, we are making gains.
That helps keep the circle going. Increased ad spend, increased demand, conversion, life expectancy, scale helps us to enable us to continue to spend more, which we're very comfortable doing, subject to the two constraints that I want to leave you with and hopefully are ingrained. First, the direct auto lifetime combined ratio has to be at or below 96, and finally, that the aggregate company-wide combined ratio has to be 96 or below, and that goal is unchanged. Thank you. I will turn it back to Glenn.
Great. Thanks. For me, that's sort of not all 96s, or 97s for that matter, are created equal. It is an important part of the strategy. It's not just a set of words there. To be consistently at or below 96 for any long period of time is actually quite difficult and requires a great deal of controls inside the company. Certainly I'm not going to contrast to anyone else, but it's easy to sort of be below for a while and above for a while. Our intent isn't to be on average at 96 over a 10-year period. It is to consistently be at 96 or below. That balanced blend requires a lot of controls to make it happen. Thank you, Brian. I think that reinforces a very significant part of our strategy.
It's almost strange to sort of not have claims be front and center for a session like this. We had Tricia Griffith with us last year to basically take the stage and give us a good overall overview of exactly what she's working on to extend the claims strategy. What I'm going to do is just comment on a couple of quick things because we're going to explode the bullets that we think are more important for this year. Claims is actually a real highlight story for us. If you remember, she talked a great deal about sort of how to preserve the local presence, the fourth bullet point there, but a lot around matching employee skills with the claims severity and making sure that we don't over-skill or under-skill. All of that work has gone on. I'm not actually going to report on that today.
If there are questions, I'd be happy to take them later. I want to give you some sort of sense of the output and has this worked. If you recall, and if you didn't, I'll repeat them, she focused around four guiding principles. They've been longstanding guiding principles for us in claims. The accuracy of the settlement, accuracy and fairness of the settlement, the efficiency by which we get the settlement, the customer satisfaction, and the work environment that we create in doing it. Very simple. Everybody would write those down. It is really hard to get a great balanced blend of all of those things together. I would suggest to you that while we've had some great results through the decade, we have mostly had one or two or at times three out of four working really well.
I think we can actually conclude, and we're pretty excited about, that we've really actually got the best balance now of four out of four than we've ever had. Let me just quickly take you through the small graphs. I know they're different time frames, I want to make the point, which I think the point has integrity. The claims quality, that's a graph that you've seen many times. The only thing that's relevant there is it's sort of upward sloping to the right. The slope of that orange line, I don't know how to measure the economic value of that, I assure you that's a significant value.
The fact that we've been improving the quality of our claims settlement against the standard that no one else in the industry would measure themselves, this is just a self-constructed standard, we know what we expect in a claim file when we see it. Just because a file doesn't have 100% quality may not mean that it was paid incorrectly. It just wasn't the process by which we think is reproducible and sustainable over a long period of time. Our claims quality has done just exceptionally well, I thought I would probably be running out of the opportunity to say it's higher this year than it was last year, so far, that's continued. At some point, that will flatten out, hopefully, it'll flatten out at a very acceptable place.
What's perhaps the most pleasing is the loss adjustment expense, you see that coming down fairly dramatically. A point of loss adjustment expense. We don't publish that in the monthlies, it's a point of loss adjustment expense in aggregate year-over-year. That's huge in pricing advantage. We're now carrying no degradation, in fact, improvement in claim quality at a lower cost. Can you really have those two things working? We think so. The other two are every bit as important, maybe should really be on the top. The overall claims work environment. You can see there a somewhat dramatic, you might say, "Well, gee, there's no continuity to the data there." If you remember, Tricia last year was talking right after we had made a very significant restructuring of our claims organization.
The management group was a fairly significant reduction in force to better meet the needs of what we see as the process going forward. Unfortunately, it required we have a new process that's taken almost a decade to get there, we had to adjust for it. Clearly, the internal environment took a hit right after that's the dip that's down. I would suggest to you, I am more than prepared to provide additional data points as time series develops, the fact that we came right back on this claims work environment, again, an internal measure, came right back to the all-time high is really a great reflection of the leadership in claims the comfort with which everything has been communicated what's going on in the claims organization today.
I believe that the work environment now is maybe as good as it's ever been, and on its way up. That means that these other things are very sustainable. The customer environment and customer satisfaction, I think the orange line there might be a little liberal, but hopefully, that'll be one that we're more on track with. As we measure Net Promoter Score, we're very aware, and I'm actually very aware from other industries, that in the last couple of years, it's been a little interesting to calibrate Net Promoter Score. It seems to be the economic cloud over our economy is not necessarily making people sort of that comfortable. That seems to be reflected in some of those types of measures. That's not an excuse, it's just a level of suggesting that in our claims environment, we're still going up.
It may even be more favorable for the future as we continue to do the sorts of things that we think have already shown some nice improvements. Perhaps as the economy recovers, we may even see a calibration change there as well. The point here is four out of four feels very good. I've included something that for some of you will seem familiar. We first used this and put it in the annual report for 2003. I think I may have introduced it in 2000. For me, it's important. It's just the way we think at Progressive. This graph, think of the darker line there. The vertical axis is total cost, what we pay in a claim, plus what it costs us to adjudicate that claim as a function of loss adjustment expense or what it costs to determine how much to pay. Very simple calibration.
If we don't put any effort into the claim, we're likely to overpay the claim. If we put too much effort into the claim and do an accident reconstruction for a letterbox claim, it's probably going to get awfully expensive. What's the optimal place to evaluate a claim, and ultimately pay it? Where's the derivative zero? How do we drive down that curve? Everything we do at Progressive, we're trying to drive to that minimum point. We're also trying to reshape the curve. We're trying to move the curve, as you see on the dotted line, and Tricia used that last year. I would suggest to you that I think while we can't calibrate this perfectly, we've shifted the curve, and we're going to keep shifting the curve.
Those marginal changes in that curve become real competitive advantages because the marginal changes get harder and harder, they ultimately also give us advantages we believe in the marketplace that may be harder to replicate. As you saw with our financial part of the strategy, we have great intensity around something that really is dear to us. This is equally as important to us in terms of execution, what makes us just that little bit more focused, a little bit more intense, and that's how we get the competitive advantage, along with the distinction that we're trying to achieve as the foundation statement suggested. Our Concierge Claims Service has definitely given us a great deal of these types of gains, we've been able to lift those gains and apply them into areas where we can't actually offer the Concierge Service because of scale.
All things are working well, I hope, in fact, I'm sure we'll come back and probably share with you these four guiding principles and the measures that match up to that. If we can keep getting four out of four and keep pushing that curve slightly to the lower and left, I think we're creating real competitive advantage through our claims organization. Claims to Progressive is not just the back end of the process. It is very much a part of what we stand for. Let's move on now to sort of industry-leading product and offering. I said I'd try to give some sense for why these words maybe are a little bit more meaningful to us and perhaps just aren't words that would otherwise fall out for someone else. Let's just take the first one for a second, open and price for all drivers.
As I think about it, all of us have been product managers at Progressive at one point in time, and in the mid to late '80s, I would say my product was open and priced for all drivers. You might have to be a little crazy to buy it if you were otherwise a very clean driver and had no issues because that really wasn't our focus in the late '80s. We had a price for you, but it wasn't the target. It wasn't that reach that I was talking about before, where we've now started to target our audience and act with purpose. We are extending our reach. We really do have a price. All of us are insured with Progressive and not because we work for the company. I mean, probably yes, probably no, but it works.
I wasn't when I was the product manager of my own product because it wasn't really open and priced for me. That's a change. What's even a bigger change now is designed for those who intend to maintain coverage. Again, you've heard me go on about this, and you've seen the comments about rate stability, loyalty programs, NPS, creating reasons to stay, eliminating reasons to go. It becomes sort of really self-fulfilling. John Sauerland coined the term inside the company of we want to be a destination insurer. I love the term because I think at times when in the past, which clearly we wouldn't be here today without the past, sometimes we were the training wheels for insurance, and that's not where we are today. We've changed greatly as a company. We're a destination insurer, and we're building on that concept.
Consumer segment focus, features, and coverages. That was what we dedicated a session a couple of times ago, and that was the first introduction of the Sam and Diane and Robinsons, et cetera. Let's take the second and third points. Industry-leading price segmentation. I think for those who know Progressive well and perhaps for those who don't know us quite as well, I think it's fair to say that we are generally associated with being a fairly good segmenter, statistically a strong company. We get the math. I also hear from time to time commentary that maybe segmentation is not quite as powerful as it used to be. I think maybe you might think just a little differently after you hear from Jim.
Jim's our director of R&D. He spends a lot of time thinking about segmentation and even marginal gains in segmentation, again, to give us the kind of competitive advantage we seek. Jim?
Thank you, Glenn. Good morning. Thank you all for joining us today. As Glenn mentioned, one of the most longstanding elements of our strategy has been to have industry-leading price segmentation. I'm here to talk to you a little bit about what we mean by that, why we think it's so important, and what we've been up to on that front over the last couple of years. First, what do we mean? We want to have industry-leading price segmentation that covers the cost of each of the risks that we insure. We want to price the expected lifetime costs of insuring a risk, including their loss costs, their estimated loss adjustment expenses, the operation costs of servicing those policies, and the acquisition costs that Brian talked a bit about before, and I'll talk a bit about later.
We also want to tailor that to the channel in which that person bought that insurance, whether it's agency or direct. As we see, certainly acquisition differences, excuse me, and also some loss cost differences, and we'll talk a little bit about that. At this point, some of you are probably thinking, wait a minute, in many, many industries, we've sort of heard pricing the cost is kind of a bad idea. I should price to market. I'm going to give you a little bit of an example here why I think that's a little bit different in our industry and some other financial services industries, that pricing the cost really is probably your best choice.
I apologize to some of you who for whom this is a pretty common and a known topic that you're familiar with, but it's pretty important in our industry, and I think there's a few new faces. I'll spend a couple of minutes, but I'll go quickly through this to explain a concept called adverse selection and why we think it's so important in insurance. Imagine a world, if you will, where there are only two insurance carriers, creatively named Carrier A and Carrier B in this example, and they're insuring some pool of risks represented here by the blue vehicles. Let's say both carriers have figured out that to price those insureds correctly and to cover their costs or earn their necessary profit, they need to charge about $1,000 each to cover those.
In a world where price is the only thing that matters, and of course it isn't. We know about brands and claims and all the other things. In this world, let's assume just price is the only thing that matters. What happens if they each charge $1,000? Well, they each get some fair share of those risks, and it works out just fine for them. They need to charge $1,000 each. They do charge $1,000 each. They earn the required return. They split the market. It's a pretty happy world for both. Now let's say Carrier B figures out that these risks aren't all the same, that some of them are somewhat different. That let's divide it here in terms of the orange and the blue cars.
They figure out the orange cars, they need to actually price at $1,200 a year to make the required profit to cover their costs. The blue cars, they only need to price at $800. There's about the same number of each. The average is still $1,000. That was right. Still is right. The price that you need to charge for each different group is different. Let's say Carrier B prices that way. Orange cars, $1,200, blue cars, $800. Carrier A still hasn't figured this out. They're still going to charge $1,000 each since that's how it always worked for them. Again, if price is the only thing that matters, what happens? Well, the orange cars go shopping, and they say, "Well, Carrier B wants to charge me $1,200 a year, and Carrier A only wants to charge me $1,000.
I'll go to Carrier A." Meanwhile, the blue cars look, and Carrier B is only charging them $800. Carrier A is charging them $1,000. They're going to go to Carrier B. Pretty simple. For Carrier B, this works out just fine. They need to charge them $800. They are charging them $800. They're going to earn the profit they need to earn on those policies. For Carrier A, however, this is a very bad outcome. They suddenly are insuring policies and vehicles that they need to charge $1,200 to, but they're only charging $1,000 to. They're going to lose $200 each per year. The trick here, too, is that Carrier A may not know this. All they know is their loss costs have gone up by 20%.
If they haven't figured out that the orange are actually different than the blue, all they know is loss costs have come up, and the only recourse they may have is to raise their prices. They'd have to raise them to 20% in this example. Now Carrier B is going to be competitive on the orange cars at $1,200 each, and on the blue, they've got a great price for them. Carrier B will gain more and more market share in that example. Obviously, life is not this simple, right? There aren't only two carriers in the world. There are scores of carriers. There aren't only two segments. That's just orange and blue cars, and they're not 20% apart. It's a pretty big difference. There are lots and lots of segments.
In fact, we have millions and millions of different segments that we're pricing to, different prices that we can generate. Our competitors also have millions and millions of different combinations of risk factors. By the way, they don't all line up. We don't all use the same thing. I can't just say, "Well, here are the segments. Let me line up the prices for us and all of our competitors and see which one works the best." They've got to use different information, different formulas. It's very complicated, and it's shifting all the time. Everyone is changing their prices and their pricing structures constantly. The key thing to take away from that is segmentation is always happening, either by you or to you. The trick is to stay ahead of that and to continually make your segmentation better.
How do you stay on top of that? Well, the first thing, of course, you do is you look at what your competitors are doing. You try to find out when they launch a new product or a new segmentation variable, you take a look at that, and you say, "Will that work in my own product?" In some cases, the answer is no. It's tailored for theirs, and it really won't work in yours, or you're picking it up some other way. In other cases, the answer is yes, and you try to incorporate that. More importantly, perhaps, and the thing you can control a little bit better is to continually try to make your own segmentation better. Let me talk a little bit how you can do that. Fundamentally, there are three ways you can do it at a very high, simple level.
You could add new variables or information, you could interact the information you have a little bit better, or you can essentially apply better math, figure out what the underlying patterns are a little bit better using different techniques. Let me give you quick examples of each of those that you're probably familiar with. For instance, the biggest change to come along in the last 15 or 20 years has been the introduction of information from consumer credit reports. It was introduced in the mid-'90s, and it was probably the single biggest segmentation innovation in the last 15 or 20 years. Companies that got to that early, that was worth hundreds of millions of dollars. This was new information. We just never had it before. It provided some incremental value over the way we used to rate policies.
When you think about interactions, let me give you an example of that. This is combining data you already use. We already rate on the age of people. We know that a 17-year-old, all else equal, is riskier than a 50-year-old. We also rate on the type of vehicle they drive. We know sports cars are riskier than family sedan. It may be the case, however, that a 17-year-old with a sports car is not just the product of those two things, but even worse. This is what we call an interaction. The fact that it's a 17-year-old with a sports car is different than just the average sports car. That would be an example of an interaction. Then finally, there's just better science. This is just running the math a little bit differently and potentially better.
That could be saying, instead of having all these things multiplied together, I'll add them together. Instead of trying to predict my expected loss costs, I'll predict how likely someone will have an accident and how bad that accident is likely to be where they have it. It's an empirical question. You run the math, you see which version of that, which formula works better. That's all the math, but the reality of this stuff when you try to put it in is not just about the math. You're subject to a series of constraints that you're trying to work this within, and sometimes it removes some of the segmentation you had hoped to gain. The easiest one to think of are the legal and regulatory constraints. While we were insuring the U.S., that's not one market. It is 51 different markets that we operate in.
The laws and regulations are different in each of them. Things that one state finds perfectly acceptable is not in another. We will always comply with all the laws and the regulations that are in the industry, but we have to tailor what the math answer might be to fit that in some cases. Second, you have to think about consumer acceptance. It's another constraint. I could have all the data in the world that said blue-eyed people are safer drivers than brown-eyed people. I don't know how many consumers would really accept that. I'm not sure how many regulators would really accept that, regardless of what the math happened to say. Last example I'll give you is of what I'll call a logical rate. Around customer experience. Most carriers give you a discount if you insure multiple vehicles with them.
If that data were happen to say that that discount should be very large if you had two vehicles rather than one, that might make sense in the math part of the world. Think about the customer who has two cars and says, "I want to sell one. Now I only need to insure one with you." If that was a pretty low price vehicle relative to the other one, when they take it off and they lose that multi-car discount, the price could actually go up. I don't know about you, but most people, I think, believe that when they insure less with you, their price should go down. Regardless that the math might say that discount should be that big, we can't really do that. We have to actually consider how the consumer is going to interact to that.
That should give you a flavor of some of the constraints we have to work within when we introduce new segmentation. How have we been doing? I'm going to borrow a concept here from social sciences, I'll relate it back to insurance, and spend a little time setting it up because I'm going to use it throughout for quite a bit. In social science, a lot of times they try to show you income inequality, they basically lay out the axes this way. On the vertical axis, they look at the cumulative share of wealth in dollars. It's a percentage of cumulative wealth dollars. On the horizontal axis, they sort the population from highest wealth there on the far left to lowest wealth out there on the far right. Dollars, people. The easiest way to think about that.
If the wealth was completely evenly distributed amongst the people, then you get a line like this, a diagonal line, where about 50% of the population would have 50% of the wealth, 80% of the population would have 80% of the wealth. Very straightforward. That would be perfectly evenly distributed. The other end of the spectrum, you could have a distribution look something like that. A very uneven distribution. In this case, about 15% of the people, here on the horizontal axis, control all the wealth, and the other 85% have none of it. This is your feudal monarchy view of the world, where a very small part of the population controls all the money. What about our world? How do we translate this to insurance? Let's change the axes around.
Still people and dollars, but on the vertical axis now is the cumulative percent of actual loss dollars. On the horizontal axis, it's the population, in this case, sorted by predicted risk, from the highest risk drivers on the left to the lowest risk drivers on the right. If everybody had the same risk of loss, or your model was really terrible, what you would see is that half the people would account for half the losses. They all have the same chance of loss. I'm going to use loss and pricing here interchangeably because remember, we price to our expected loss cost, so in this example, you can use them interchangeably. Perfectly even 50% of people have 50% of the losses, 80% have 80%. All works out the same. What we observe in a single year doesn't look anything like that, though.
What we observe in a single year looks something like that. 10%-20% of the people have an accident and account for all of your losses. The other people don't have any. This makes sense, right? Most people don't have an accident in a given year. Only about 10%-20% do. Does this mean we should charge the other people nothing? Of course not. There is some risk there, and we're pricing to the expected risk, not what we actually observe in any given period. You can use this to start to think about what does it really look like? What's the perfect curve? If this diagonal is, I don't know anything, I'm going to charge everybody the same price. This is the blue cars, everybody's $1,000 example. What would the real curve look like? The perfect curve. No one really knows.
From that data, we can begin to estimate it. Our best guess makes it look something like that. There is differentiation in the risk of drivers. I think we'd all accept that. A smaller share of the population will account for a disproportionate share of the losses. You get a curve that looks like that. Now you can start to put on your actual pricing models and compare it from diagonal of completely awful model, I don't know anything, to out to that perfect curve and see where you lay in that range. How have we done? About 1994, we had a product which we creatively called 2.0, which was actually the second generation of the product, and that contained most of your traditional rating variables. Your driving record, how old you were, what kind of car you drove, where you lived.
The basics, the classics, right? It did pretty well. It gained all that space. They're the traditional rating variables for a reason. They do predict loss pretty well. By 2000, we had a model which we called 4.0, and the big change between those two lines was the introduction of credit. A couple slides ago, I said, wow, credit was a really big deal. It's a big innovation in the industry, and it's hundreds of millions of dollars of the people who got there early. If you look at this chart, you say, wow, the line really didn't move that far when we did that. Those are not incongruous. Small improvements here are worth a lot of money from a business impact point of view.
There's still a lot of room to go, that you didn't add that much segmentation in the grand scheme of things, but it's worth a lot of money in the industry if you get there early. Where are we today? This line represents the segmentation power of our most recent model, which we've called 8.0, which we launched last year in 2009. You can see, again, some improvements since where we were in 2000. Made some good progress there, but still a long way to go to get to that perfect model. I want to talk to you a little bit about that 8.0 model, and what it was designed to do and how we got to that point. If you remember back in 2000, we had a reorganization where we created separate direct and agency business units.
The goal of that was to really bring a lot of focus to each of those channels and the differences between them, so we could optimize the direct business for its dynamics and really learn about that business as it was still relatively young and really optimize for the agency business. We did that. We created, in that process, completely different direct products. By product in this case, I mean rating algorithm. We created very different direct products and agency products. We had different quoting experiences, so the way an agent would get a rate was different from the way a consumer would get a rate. Because of that generated different servicing and follow-up procedures in our call centers and internally.
What I mean is servicing or follow-up procedure when a consumer called in and wanted to make a change to their policy, there are rules dictated by the product about how that has to happen. Because the products were different, all these procedures were different. These were also different systems and different everything else. Fast-forward to 2007, we did another reorganization. This time, we brought those things all back together because what we had found was we had optimized those things really well, but that had led to increasing differences, and we started to pay something too high of a duplication tax on our minds. We brought those back together to gain some efficiencies, improve our speed to market, simplify things so that we didn't have different sets of procedures, we could get closer to one, and that's what we did with this 8.0 product.
We took the best of the agency and the direct products, we created an integrated product here in the middle. We still have observed, though, differences in the loss behavior between those two channels, differences in the experiences, differences in the acquisition economics. We have some direct specifics still and some agency specifics, but we're only different now where we think that's material and important. Because we have a similar product, though, we now can have much more integrated follow-up and servicing procedures. This helps lower our costs, simplify things in the call centers, and because these are all in the same system now, it makes it easier for us to roll them out. We have maintained, however, different quoting platforms. We find that the experience needs of an agent and a consumer are still very different, and we're treating them differently.
We think that is a material difference that's worth keeping. This is what we've done to create the ADO product. We've brought these things back together. That's led to some of that segmentation gain you've seen. I'm going to use the direct product here as an example. The blue bars here represent policies. We basically said, what was the rate in the old product? What's the rate in the new product? Let's take the difference and see how different they are, and we sort them. People on the far right here, policies on the far right will be seeing rate increases, policies on the far left will be seeing rate decreases. When you do that, one simple thing to do is say, let me check out the loss ratio of the people in each of those bars.
I would hope the people I'm raising rates on have higher loss ratios. Right? It's a pretty simple check you ought to do. The orange line is that loss ratio. It's also the indicated rate change. People on the right, fortunately, have higher loss ratios. They need rate increases. People on the left had lower loss ratios. They deserved a rate decrease. We'll throw up what we did-- how much did we actually change the rates? That's the blue line. Lines up pretty well. That makes you feel, as a research guy, when you see that, you feel pretty good that I'm moving rates in the right directions. I'm actually improving my segmentation over what I had before. Some of you are probably trying to do that mental math of, you needed to raise rates 10, you increased to 12.
What's your new loss ratio look like? If you think back, we want to price to our costs. We want to price everybody, so they are priced about the same loss ratio. We'd like the new loss ratio line to be very flat. It is pretty flat. It's a lot flatter than an orange line. It's not perfectly flat because I'm not out of that perfect model yet, but moving in that direction. We've made significant progress in improving the segmentation here. To get done with that, however, it's not that agency rates and direct rates are not identical. That is still not true. I will tell you now that the loss assessments have gotten a lot closer. This is what it looked like before we had ADO.
We took a group of policies, we rated them in direct, we rated them in agency, and this is the difference. People on the far right here, direct had much higher prices. People on the far left, agency had higher prices, people in the middle were pretty close. This is what our loss cost estimates looked like before ADO. After ADO, they tighten up quite a bit. They look more like this, these orange bars. Much tighter distribution. There aren't as many people where the rates are quite as different. Our loss costs have tightened up considerably now that we have added ADO, which makes sense because we brought the products back together. That is just the same chart. I just took out the other bars to make it a little clearer. These are not final prices, though. This is just the loss cost.
It does not include the acquisition expenses. Brian has already hinted at that we treat those somewhat differently between the two channels. We layer on the indemnity, these loss assessments, we layer on the acquisition on top of it, we get final rates that look like this. The difference in price between agency and direct, these policies on the final price looks more like this. Tighter than it was before, but certainly not identical and certainly broader than just the loss cost. Implies that the acquisition differences and the way we price for that contribute to a lot of the differences in final rates between agency customers and direct customers. Let me explain to you a little bit about how we do price agency and direct, and why it leads to some of these big differences.
As Brian mentioned, we have chosen to expense all of our direct acquisition costs in the first term. It is a choice. Does not really affect how we price it, but it is what we do. In agency, we have a variable commission. It is a function of premium. Let me take two example policies. One on the left will be a low premium policy with a relatively short life policy expectancy. The one on the right is a high premium, long life expectancy policy. On the direct side, which is represented by the blue bar here, the cost of acquisition is identical. We do not really care what the premium is. It costs us the same to get them in the door. In agency, however, the expenses are very different. It is a function of premium and for how long you are going to be paying it.
For the short life expectancy policy at low premium, we are going to pay a fixed percentage across a small premium for only two terms in this case. Whereas on the far right, it is a high premium policy, the commission dollar expense is going to be larger. In this case, it is for eight policy terms, it is for a long time. When we go to price that, what we do with direct is we take that fixed expense and we divide it over the expected life of the policy. In the far left case, we just divide it in half because there are only two expected terms. On the right-hand side, we divide it by eight because there are eight expected terms.
The output of that is for these low premium, short life expectancy policies, agency is going to be much more competitive because they're paying a relatively low commission relative to a high direct acquisition cost. Meanwhile, on high premium, long life expectancy policies, direct is going to tend to be more competitive because, again, that variable commission is going to go up and direct expense can be prorated out over a longer period of time. This will generally be true. Direct will be more competitive at higher premium and long life expectancy policies. Directionally, that's consistent with the renewal economics that Brian shared. Those renewal terms in direct are very profitable. This makes us more competitive on longer PLE policies in direct, which only fits that strategy pretty well. How's this new product doing?
We have it out in 10 states today, and we'll have it out in states representing 60%-70% of our net written premium by year-end. When you take two products that had seven years to get different, to drift apart, the rates are going to be pretty different. When you put them back together, there's going to be a lot of rate change for individual customers. We've invested in the last several years in rate stability initiatives. I know we've talked about that at prior meetings. That will help to mitigate that rate change for those customers and ease them onto the new product. That's good for retention. We think that will help us retain more customers.
It will, however, slow the speed of the adverse selection towards our competitors that our new segmentation is generating because our customers, the customers we think we need to increase rates on, won't be seeing those as quickly, so won't be as likely to go to a competitor. All in all, we think that's a good trade-off. We think we'll get the retention benefit. We need that. We will have the segmentation at new business where elasticity is the highest because we think it's most important there. We think that's the right balance on those. The early results of this new product are encouraging. The loss ratios look okay so far. The data is very thin so far, though.
So far, it's hitting the targets that Brian described, and we're seeing a slightly more preferred mix of business in terms of dimensions like multi-car policies, homeowner policies, full coverage policies than we had in the past. We're encouraged by that as well. Where are we going from here? This is a chart we had up earlier, where we had seen the progression from older products to the most recent product. We don't stop working on these things. We don't just wait for a reorg to try to come up with a new product. We're always working on those new products. We already have things identified sort of back in the lab that we think will add even more segmentation. We're beginning to start to implement those. Sort of proven out the math, now we're starting to actually do the implementation.
When we look at the segmentation that will provide, we get a curve that looks something like that, which I've creatively called nine-zero, which shows a pretty nice segmentation gain. I'm encouraged by that. I will tell you now, I don't think we'll get all of that because remember all those constraints I showed you, we're just starting to vet all those. The segmentation is against that, we'll have to make some modifications to really make it a usable, marketable product. We're still going to see some segmentation gains out of it. I'm confident in that, this shows good improvement over where we've been, with still a long way to go before we hit that perfect model.
This, however, does not include what I think of as the most exciting segmentation we may have in the next few years, which is usage-based insurance, which I know we've talked to you about for some time now, John Sauerland is going to talk more about it. We already have a model we're using in the market that we think provides great additional segmentation. I'm here to tell you, I think there's a lot more yet to go on that we're really just at the beginning stages. I say that for two reasons. I think usage-based insurance is so much better than the traditional rating variables for two reasons. One, it's less of a proxy for driving behavior and more of just a measure of it. Rather than just being correlated, it actually measures the driving behavior that we're interested in.
Second, it is a much richer source of data than any of the variables we've had in the past. I'll tell you a little bit more about that. Our traditional rating variables, if you think about those, they tend to be a single piece of information. How old are you? You're 45. Great. There's one piece of information. What kind of car do you drive? I drive a 2005 Honda Accord. Great. There's one more piece of information. Just individual pieces, and when you put them all together, maybe you come up with 50 different data points. One of the reasons I think credit was a big step forward was it added so much more data. On a typical credit report, there are about 2,000 individual transactions that make up that report. It's very complete and very dense. Think about your driving record.
How many of you have ever driven unsafely or exceeded the speed limit and not gotten caught? Nobody here I know. Maybe on the way to work, you saw somebody else doing that. How many times do you think people have not paid their credit card bill, and it didn't show up on a credit report? Almost never. Those guys, it's there. You really get to distinguish between safer risks and less safe risks. Credit has up to about 2,000 data points, lots of information there that we were able to parse through and really add a lot of segmentation to. Now think about usage-based insurance. In a single trip that someone takes, we get about 1,000 data points. We're at 2,000 today on our current model. One trip, we have 1,000 new data points.
Over a single term of driving, that equates to about three-quarters of a million data points. It's a vast, rich source of data that instead of being closer to proxies for how you drive, like how old are you, it's actually measuring the driving behavior we're interested in. That's a great new source of data. I'd submit to you, we like data. We've done good things with that in the past. I'm confident we're going to find new segmentation in that. When I look at the whole, our progress in hitting this goal of having industry-leading segmentation, I feel like we've made good progress over time. We've got some things in the lab that will help us maintain that position. With the usage-based insurance, I feel like there's a lot of potential out on the horizon.
It's always a challenge to take on a topic like segmentation because our factory doesn't sort of lend itself to tools. It's a statistical factory. Hopefully that gives you a sense. For me, the takeaway was Jim's comment, segmentation is either happening to you or by you, there's no standing still, you better know which are the blue cows and the red cows, whatever analogy that works for you. That's a huge part of our strategy. Hopefully that gives you some insight there. We're going to quickly roll into John Sauerland now. Almost all of you at some point in time have heard me perhaps say that I always felt the substance of Progressive was better than our representation, meaning our brand and our communication to consumers. I hope I don't have to say that too much longer. I think we're catching up.
I think our presentation of ourselves is getting just a little bit better. The Leadership Brand element of our strategy is clearly a very important one, perhaps one that has taken on even more life. We've always been pretty good at the math, we've been pretty good at the claims, been pretty good at technology, this is sort of that next leg in our very stable stool now. We're going to ask John Sauerland to take us through sort of the work that's been going on in Leadership Brand, it's not a totally distinct skill set in some cases from our segmentation skills.
Thanks, Glenn. Good morning. You can clearly see our aspiration is to be widely recognized as an industry leader with broad presence and powerful awareness. As Brian pointed out, Glenn pointed out, we now have a fantastic vehicle with our Superstore campaign starring Flo to deliver messages about Progressive to consumers. To date, I think we've done a really good job delivering messages broadly, especially around acquisitions, service focused messages. In case you're not completely familiar with our campaign, frankly, because we're kind of proud of it, we like to show it off, I'm going to share with you now a reel of clips that, again, I think show that we do a great job with broad-based acquisition and service messages.
I hear Progressive has lots of discounts on car insurance. Can I get in on that?
Are you a safe driver?
Yes.
Discount. A homeowner's discount. The e-sign discount, and the paperless discount. The multi-freckle discount. Saving money with our discounts and customized policies. Oh, thanks for dropping in. Just making sure you're listening. Not to mention the multi-policy- Discount. Isn't getting discounts great? Yes. There's no discount for agreeing with me. I'm looking for a deal on car insurance. I think we might have a coupon in here. There's an easier way. We've got the Name Your Price option. You do?
I'm just trying to save money on my car insurance.
You know, with Progressive, you get the option to Name Your Price. You tell us how much you want to pay. We'll build you a policy that fits your budget.
Wow. The price gun.
Feel so empowered.
Power to the people. Yes, we're looking to save on car insurance, even if that means we have to shop all day. We can help you save because we instantly compare your Progressive Direct rate with rates from other top companies. Watch this. Nice savings. How can I know I'm getting a good deal? Folks of all shapes and sizes. Saving time by comparing our Progressive Direct rates to those of other top companies. Look at the deal we just got him. That's a new pair of shoes. Yeah, or a big tricked-out name tag. I can just drop off my car, and you'll take care of everything? Yep, even the rental. Concierge Claims Service, local response claim service, and twenty-four/seven live support, all at no extra charge. We'll assess the damage, coordinate the repairs, and call you when it's fixed. It all comes with your policy.
Cool.
Hold on, cowboy. I'm not done. For less than $1 a month, you also get 24/7 roadside assistance.
Wow.
Wow. I know. I'd say it louder.
Hope you like those as much as we do. As Brian pointed out, we are able, given the great response we've gotten from our campaign, in concert with the increase in conversion, the increase in policy life expectancy, we're able to spend a lot more. We don't spend to a budget, so to speak. We are sharing with you here a plan number for 2010, we are forward-looking here a bit. Know that if circumstances change, response changes, conversion, et cetera, our spend will change as a result as well. You can see from that graph, we're at about a $0.5 billion spend this year. That's up about 30% over last year, call it more than 50% over two years ago. That's great. That said, it's a challenging category.
You can see our share of voice here on the right, and we calculate that using Nielsen data. What Nielsen says we spend relative to what the industry spends, and you can see, even though we've been spending a lot, the industry growth in spend has been outpacing our spend until recently. In 2009, year to date through 2010, we have grown our share of voice, and we are now in a strong number 2 share of voice position ahead of State Farm and Allstate. Again, our plan on spend is to spend to what is allowable, understanding what's going on with our business across those 3 attributes and trends that are good. My expectation is that we're going to continue to be able to grow our share of voice.
The spend and the creative is working not only from driving prospects to Progressive, it is also working in terms of building brand equity, if you will. You can see the graph there around unaided awareness. This is the % of people who, when asked, "Name an insurance company," say Progressive. You can see this is up, and it's up to historically high levels. We've also done well positioning Progressive in consumers' minds. Glenn shared with you the graph on the left here already. That's value relative to caring. The graph on the right, on the bottom right, I like a lot because if you think about our foundation summary, again, competitive prices with distinctive products and services, I think this says that's working. This is personable and easy as a summary, and we're distancing ourselves from the competition nicely in this space.
If you think about what drives that, you think about immediate response claims, you think about 24/7 service, Concierge Claims, comparison prices, Name Your Price, are all great examples delivered by Flo as well, helps us get a lot of distance from our competition around personable and easy to use. I think we're doing pretty well in the mass marketing of or mass delivery of acquisition and service messages. I'd offer that we are starting to get better at targeted messages. I'm going to share with you a reel here of some of our more recent examples of targeting, and then I'll come back and talk a little bit about each of those segments we're talking to.
Tom, check this out.
Sweet gravy, Bill.
Our insurance company doesn't have anything like it.
Magnificent, isn't it? With Progressive, it's easy to cover all of your favorite rides.
This is why we do this.
Oh, it looks like you guys have everything.
We're number one in motorcycle insurance, a leader in boat and RV. Ding-a-ding, ding. Fun fact, Progressive is the number one truck insurer.
I dig that.
Most bikers do. That's why Progressive is number one. Woo.
Woo!
Do you guys insure Airstream? Yep. Everything from travel trailers to mega motor homes. When your RV is covered, so is your pet. Do I have any bugs in my teeth?
Nope, you're good.
Attention, Progressive shoppers. Introducing new Pet Injury Coverage. Now, where Progressive covers your car, we cover your pet at no extra charge. Pickles, one policy, please. Our service is top-notch. We'll take care of you, your family, even this little guy. Oh. That's a cat. Hi, welcome back to progressive.com. How's that car insurance?
Great. Just bought a house.
Big news. We have another way to help you save.
Oh, really? How?
By bundling. If you get your homeowners and auto insurance together, we give you even more savings.
Ooh.
Home and auto together, it's like peanut butter and jelly.
Anything else I should know?
Yes. Make sure you stay away shag carpeting.
If I switch to Progressive, I could save hundreds.
Yep, that's just the start. We have a great Loyalty Program.
I'm listening.
Stick with us, and you keep earning bigger discounts and benefits, like accident forgiveness. Plus, it's free.
Wow. Great.
Now let's get you initiated into the program.
What is that?
Watch. Here you go. Automatically enrolled. Painless, right?
Totally painless.
All right, hopefully you caught those segments there. First, we were talking directly to what we call multi-product households. As Flo told you, we are number one in motorcycle. We are a leader in boat, we're a leader in RV, obviously a leader in auto as well. We do okay penetrating multi-product households, meaning selling that household more than one product. We think the potential there, we know the potential there is big, we're talking directly to those consumers to try and increase our penetration there. Second segment we were talking to were pet owners. I think we've done a great job with our Pet Injury Coverage, really talking to a segment that cares about that. If you look at our brand metrics, you'll see we significantly over-index with pet owner households relative to non-pet owner households. We're doing great with that segment.
The third you saw there were bundlers, as Glenn shared with you, the Robinsons, in this case, representative of the bundling crowd. We're still small, we're growing significantly, that segment is about 40% of U.S. households and a larger portion of the U.S. auto market in terms of premium available. We are going to continue to go after that segment because we believe the potential for growth there is huge. The final segment, if you will, really wasn't a segment, it was a spot around our Loyalty Rewards Program. We actually don't have that spot in market yet, we are just finishing that ad.
We'll start playing that in about a month or so. I think we are now starting to get onto new ground where we are able to talk to our consumers more directly or in a broad sense about reasons to stay with Progressive. We're excited to see how that plays. The great news there as well is that in tests at least, those spots do very well with non-Progressive consumers in terms of increasing propensity to consider Progressive. I think we're doing a great job in getting better at the targeted marketing. There's one more spot I want to show you. Unfortunately, we haven't shot it yet. The topic of this spot or spots we will shoot in about a month or two is our usage-based rating. We're going to begin to call this our Snapshot discount.
Last year when I was on this stage, I talked to you about our MyRate program, which is currently what we refer to our usage-based rating system as. I told you we had made nice advances in our ability to deploy the wireless device broadly. That means it works in virtually all newer model-year cars now, and that delivers data to us real time. I told you that we were a lot better in improving the consumer experience, so the onboarding of a consumer into our usage-based rating program. I told you that we were really focused on a target for MyRate. It was affluent, urban dwellers, and we thought that was really the right target to shoot for with MyRate.
If you read our annual, you saw that we have more than 100,000 units in place today, which is okay, but my big news for you here today is that we now think we can go after the broad market with usage-based rating, and we don't have to target market. The main reason behind that is because we redesigned our product. We studied product attributes that really drew people to the product. Studied different facets of the product that detracted from the propensity to consider it. We've now created what we call the Snapshot discount. The Snapshot discount works as follows. Within 30 days of plugging the device into your car, you get a significant discount, and a discount up to 30%.
After six months of having the device in your vehicle, you return it to us, the device, not the vehicle, the device, and you lock in your discount. That's it. Up to 30% discount within 30 days of plugging the device in. After six months, you return it, and you lock in the discount. We have tested this with a lot of consumers, and we think it has broad appeal. We have that product design in place today in two states. We don't have the marketing out yet, but we have that design out in two states today. We will introduce that same design in six more states in July. We expect to have about half the country rolled out by year-end, and we will continue to roll it out, obviously, into 2011. We, as I said, will be making ads here in a month or two.
We will test those ads in one or two cities between now and year-end. If things go well across the product and the marketing, we expect to be broadly advertising usage-based rating in the form of Snapshot discount with Progressive in 2011. Jim told you the math works really well here. I'd say that I think we're getting close to having the marketing work as well as the math, if you will. To what Jim said, I have to add that. He said the math is really good, and the math can get even better. We have over 1 billion miles of driving data now in our systems, obviously with the associated loss cost. We have some patent protection in this space. We have worked with the regulators where we have MyRate today.
In those states, it's around 20 states, we worked with the regulators to provide them the information they need to be comfortable with our usage-based rating program. We have filed the algorithms actually under trade secret protection. The algorithms aren't public. The math is not public. I think you get where I'm going with my positioning statements here. Again, it'll be exciting to see if we can get the marketing as good as the math, and we'll know that pretty soon here. Look forward to that. All right, I'm going to shift gears now and talk a little bit about our media spend. We share with you the actual media spend over the years. We've done that for at least several years now. While we don't give you the numbers within the internet segment here, internet is growing. I think you know that.
Paid search is a significant portion of that. Display advertising is significant as well. I talked to you last year about display advertising and the amazingly rich data environment we're operating with there. We're doing more and more of that. It's going really well. Paid search is going well as well, but it also presents a nice little case study around the power of brand. Just to ground everyone in at least the terminology we use around search, this is a Google search for car insurance. Car insurance is the most popular, what we call generic term that people search for when they're looking for auto or car insurance. We call that a generic term. When you click on that, for Progressive, for everyone, on average, think of it as about $20 per click.
We also bid on what we call brand terms, Progressive insurance, Progressive agent, et cetera. Those terms cost us around $0.50, a little less than that. You might wonder why we have to bid on our own brand term. There are players who bid on brand terms. We don't bid on other competitors' brand terms, but nevertheless, we have to ensure we're number one on the listing there. We bid a small amount. The final item there you see is organic, or some people call that natural search, that is below the paid listings, but prominently on the first page of listings. With that as the base, I'm sure you've already figured out that, well, you're much better off with a brand click, right, than a generic click. $0.50 versus $20, that's sort of 40x.
People who shop for a brand are more likely to go buy that brand than people who shop for car insurance. I don't have all the pieces of the funnel here for you, but suffice it to say 60x relative to 40x, it's big. Surprisingly, I'm also going to tell you that our brand mix relative to generic is up nicely. You can see the graph over time. We think that's, to a large degree, a function of our spend and our creative, which is great. We're also beginning to understand, we think there might be a fundamental shift in the way consumers shop for auto insurance. The data on the right in those charts is Google total category data. For the auto insurance category, all Google data.
Brand terms would include State Farm, Allstate, et cetera. You can see people are increasingly shopping more for brands relative to for car insurance. My point here is clearly, if you're an established brand in this space, this plays really well for you, and if you're not, it makes the going a lot tougher in getting in, right? Okay. Just to round out the search terms I just talked about, organic search, we're doing well there as well. The graph on the left shares with you our average search ranking. When you're at the top, you're number one, which is at the bottom of the graph. You can see on Google, we've continued to improve, which is great. We've also done really well on Bing. Yahoo, you can see we've been somewhat up and down.
The good news there is that we think, at least by year-end, Bing and Yahoo are entering into, I think they refer to it as a co-marketing agreement, where they're joining forces for search marketing, and we think Bing will be the engine that they migrate to, that will work well for us. Net here, organic quote starts are up about 30% per year, which is fantastic. Before I round out the search discussion, I also want to make sure you're aware of the power of online marketing and our ability to manage it proactively when we are having profitability issues as well as we want to grow up. We always want to be priced adequately in all areas.
In this example, we recognize that PIP trends, personal injury protection loss trends, were outpacing our prices, so we were not price adequate in Downstate and in Buffalo. You can see that overnight, literally, or even intraday, we can change that and decrease our marketing spend in those areas and maintain it where we're making money. It's powerful stuff, and I think we're playing this game pretty darn well. It's a highly dynamic game. It changes every day, even by the hour. I think our team here is doing a great job. Okay. Glenn has told you our strategy involves online leadership for sure. What we pointed to as evidence that we're succeeding with that over the years has been quote initiations online. He already showed you the graph over here on the left.
Neck and neck, if you will, over time with GEICO in terms of quotes initiated online. Yes, we have one data point above them at the end of the last quarter, which we're certainly happy to see, but not the point of the slide. The point of the slide is on the right, which is policies initiated. Think of purchases. What I'd offer to you there is that we're much more close to the leadership role there in purchases than we've been historically. Over the past year, we have significantly closed the gap on policies initiated. Why? I just talked to you about a shift in mix of quotes to brand, and that certainly helped us. If you think about the other things we've introduced recently, you should also recall Name Your Price.
With Name Your Price, I told you last year, we increased our conversion around 5%, and we're also able to, deploying Name Your Price messaging and creative increased response to our ads. We also saw a small decrease in average premium, call it less than 1%, when we rolled out Name Your Price. We continue to evolve Name Your Price. We're going to continue to enhance it and make it part of other portions of our experience. We think Name Your Price will be a very valuable part of our offerings for quite some time. We have also done some deep dives around the shopping experience. What you're seeing here is the result of a study of around 3,000 to 4,000 consumers who shopped at both progressive.com and geico.com in fairly proximate neighborhood of each other. First, just let me remind you, this is a study.
I don't contend that it's perfectly representative of the country, recognize that. At a minimum, I think you'll find it instructive. The graph on the left shows our presented rate relative to GEICO's. You go on, you enter your information, you see a price. You go to the other website, you see a price. This is the ratio of those plotted against the percentage of people who go on to initiate purchasing a policy.
There's a lot of things you can infer from these graphs, the key point I want to actually make here is on the right-hand side, and that is to say while the rates presented at Progressive and GEICO were significantly different in this sample, again, 3,000 or 4,000, when you dissect the difference, the difference of $185, only about $21 of it was due to apples to apples, same coverages offered, difference in pricing. The other were simply a function of the package presented, the coverages presented. Our strategy or our foundation summary, as you have there, is competitive prices with distinctive products and services. We will not chase the market down to the lowest quoted premium. That is not part of our strategy or our foundation.
That said, understanding this is important for us. We need to understand where our competitors sit in concert with what we think we should recommend to consumers. That's a subjective thing. We have taken action around these understandings. I'm going to share with you some actions we've taken here in a second. Before I jump into this slide, though, let me make sure you aren't swayed too far in thinking around changes in average premium. If you've watched our results for a while, our average premiums have been going down, especially more recently in our direct business. Two key points. I'm going to talk about changes in coverages here. Those matter to average premium for sure. Our state mix matters a lot as well.
As we've had pricing adequacy issues in some of our PIP states, larger personal injury protection states, those states have higher average premiums. As we've grown more in other states, that shifts the mix. When you're looking at a countrywide average premium number, that drives that down. Recognize what we're going to talk about here is not the sole driver at all of average premium. The other caveat I'd offer going into this is to say we've been raising prices. I think we give you numbers around that, but last year, we raised prices. This year to date, our rates are up as well. For an apples to apples risk, in aggregate, on average, for Progressive, our prices are up. Okay. Given all those caveats, you probably already read this slide.
Bodily injury liability limits, so a significant portion of premium, and you can see the distribution of what we recommend in a quote is significantly different than our competitors. We have taken some actions as a result of this. If you look at the new business mix slide on the bottom left or graph on the bottom left here, you'll see, as an example, the 100/300 line. That's a, relatively speaking, higher priced coverage. If you look across those bodily injury limits, think of it sort of around 10% difference per bar, if you will. 100/300 has gone down, so we are recommending it less frequently. If you look at the percent of times that we're selling it, call it 20-ish%. That is still almost two X what our competitor recommends.
The minimum allowable limits, so less than 25/50, we're at sort of 18-ish%, if you will, and our competitor is recommending that more than 30% of the time. Big point. We will change again relative to what we're seeing competitively and what we think is the right thing to recommend. This has helped conversion. As an example, on the right is our vehicle coverage. This is coverage for the vehicle, and this is a set of older vehicles, so if you have a loan on your car, you have to have full coverage or physical damage coverage. If you don't, and you have an older vehicle, it's up to you. A lot of people decide not to. Understanding our competitive offering here, we used to offer or recommend that coverage to 69% of shoppers. We now recommend it to 50% of shoppers.
As you can see, conversion has gone up, average premium has gone down, but the net total new premium is up. A sort of more holistic example, if you will, from our special lines business is as you're seeing on the screen. Historically, we used the presentation on the left, and you can see, or you may not be able to read that, but we're offering 1,500 limits there in addition to other coverages. The premium, call it $100, or it is $100, I should say, specifically, the pay-in-full premium. And today, we offer that same exact package as our recommendation, but then we offer a plus and a basic, a lower and a higher option. Net, we've seen conversion go up, as you can see on the slide, 17%. Average premium's gone down 8%, but net total new premium is up 7%.
The point here again is simply understanding where we sit competitively and making sure we're doing all we can online and in our call centers to take a quote into a sale and do it with the recommendations that we think are right for our customers. I hope you've gained an appreciation that I think our presentation is as good as our content or closer to our content recently. The Superstore campaign is absolutely helping to drive prospects to Progressive. I think we're starting to get a lot better at target marketing. You're going to see a lot more of that going forward from us. We look forward very much to seeing how we can market Snapshot discount or our usage-based rating broadly, and you should see ads in targeted markets before the end of the year. Hopefully, again, rolling that out big time next year.
Hope you've also gained an increasing appreciation of how we deploy our skill sets of analytics and segmentation skills online and really maximizing our spend. Hope you're gaining an appreciation that we are getting better at retailing, so doing smart things online, again, in our call centers to get to the sale. In aggregate, you can see on the graph here, prospects or quotes are up. Conversion is up a lot. Yes, average premium is down, and this is for our direct business, but I think you can see the trajectory on our new premium. New total premium is looking pretty good. Turn it back now to Glenn.
Great. Thank you. For me, I hope the long-term carry away is that the concept that John said, our marketing is starting to catch up with our math. If we can get those sort of in sync, I think we've really got something that's a very different dimension of Progressive. I want to be conscious of leaving enough time for Q&A, so we'll make sure that we get to that. Let me just say a couple of words on service that is critically important for us and that strengthens the relationships with our customers. Couple of points here. The first one is really something that might seem a little different than strengthening relationships, but online transactions, it's very clear to us that's what consumers want to do when the transaction is something they feel comfortable doing. We've got to provide a great website for them to do that.
Anything less than great gets customers annoyed. We're now actually at the point, we've chosen not to put % on here, but I assure you that the lines represented there in terms of the number of transactions that consumers are doing for themselves online is a very high % of our transactions. We see that going only really in one direction. Ultimately, it gives us some opportunity to create leverage on our expenses. Clearly, the mobile platform's subtly emerging, but they're going to be a big piece. This is not like we think. It will happen. What I find intriguing, and this is obviously very small numbers, but on the mobile application up with the iPhone application, is just the number of payments that people made on the iPhone in the first three months that we had the application available. We never told anybody you could.
It's just expected that you would have that sort of thing. This is not even something today that you really have to market at all. When we make these things available, people will use them. That gives us a great deal of opportunity to be able to affect our policies in force for a full-time equivalent. Just earlier this week, we had a business review on a lot of these issues, we start to see not only the kinds of things that Jim talked about in terms of reducing that duplication tax as we moved our products together and have easier servicing parameters for our folks. That's an effective reduction in terms of the number of people we need. We're starting to talk about and already have implemented a fairly large standard, but we think we could even larger.
The technology that is allowing our people to work from home. There's a lot of things that we can do. It's not the subject of this discussion, but there's a lot of things that we can do to get those non-acquisition costs headed in the right direction. The graph that we've actually shown you for a couple of years now on the right there, cumulatively over three years, that's about a 27%-28% improvement in policies in force per FTE. That, for us, is the fundamental measure of efficiency. I include here something that I think has to be done in balance, and that is sort of the service and cost slide. This is a derivative of some publicly available information I think we've shared with you at different times in the past. J.D. Power.
Power surveys about service and where Progressive was, a very strong intention for us to change that picture. We have consistently moved up and moved up more than any other single carrier, admittedly it's easy to move up from where we were, in the last five years than anyone. Now if we sort of plot service with cost, we come up, I'm looking obviously at only the blue segment there for Progressive. We've actually been able to reduce cost slightly, but not with any trade-off at all in service. That combination, that vector, that's what we want to see go even higher. We have no reason to believe that the things that we've got in store won't continue that trajectory. Lots of things happening there that are important. Now, if you read the wording on the last bullet point, it says reasonable integration.
That might seem a very weak word to use in your strategy. Let me give you some insight into exactly what that means. With multi-product households, John already talked about it wasn't a big push. Now it's becoming a bigger push as we've got more of the Wrights and the Robinsons in our book of business. We're starting to see something close to 10% of our policyholders actually have multiple products in the household, whether we're the original equipment manufacturer or not. In the case of homeowners, we're not. We didn't design Progressive around the household. We designed Progressive around the policy. Nothing wrong with that. Now we've got to start to take a household view. A household can have many policies. That's not the way our systems work. At first, we said we'll have reasonable integration with products within a household.
We'll make it work to test the use case, to test the business case. Will people buy multiple products from us? Will the bundling of homeowners work? Really good news. We think that use case is now being tested. The next iteration of this strategy will remove reasonable and put something closer to seamless integration, so when customers buy multiple products from us, we will know all of their household products and be able to service them as a true household. We do a proxy for that today. We will have a version of household view in place later this year. I think the important point is we didn't necessarily test the systems before we tested whether consumers would accept Progressive as a multi-product offering. We believe they've accepted us as a multi-product offering. Now we'll build the systems to respond to that.
On sales and service, that really strengthens those relationships. There's lots of things that give us the opportunity to continue on our cost curve, just as we've showed you in claims. We can reduce those costs, we can reduce them in the non-acquisition expense, but at the same time, having those service vectors. From my perspective, that's the only combination that's acceptable to us is to make sure service is stronger. The branding kind of metrics that we've shown you clearly are important to us. We don't ever want to see the spiral on that get reversed. It has to be compounding. This is a big part of our strategy, not necessarily something that we'll talk a great deal about, but it's a very important part and continues to give us competitive advantage. We think there's more gas in the tank there.
Let's quickly go to the one more bullet point that we want to explode and push through that. Broad distribution. Some things here haven't changed for a long time with Progressive. The last bullet point, direct-to-consumer, I think we've covered that a fair amount today, so we're not going to talk more about that. I've asked John Barbagallo, who talked to the group in 2006, specifically around broad distribution and our agent distribution specifically, to give us an update on that, and more importantly, some of the things he's thinking about with preferred agency distributors.
Thank you, Glenn. When you talk about broad distribution as an element of strategy, certainly broad distribution through independent agents has been a key element of Progressive strategy for a long time and continues to be. We choose to work with more than 38,000 independent agents around the country. We recognize many of our competitors choose to work with fewer agents, and in some cases, substantially fewer agents. We like the broad distribution reach this gives us. In fact, it gives us more local distribution outlets for our insurance products than State Farm and Allstate have combined. Through that broad distribution reach, we have been able to garner significant share within the channel across a number of lines of business, from monoline personal auto to motorcycle and our specialized products and our commercial auto products as well.
These lines of business are important to our agents and their customers, but they're not necessarily core lines of business in many of these agencies, nor is Progressive necessarily a lead market in many of those 38,000 agencies. Nonetheless, through broad distribution, we've been able to garner a not insignificant share of packaged personal auto, which is a key line of business for many of these agencies. When I say packaged auto, I'm talking about auto insurance customers that own homes and typically require other personal lines coverages. Think of the Wrights and the Robinsons from Glenn's earlier slide. We're going to talk a little bit more about packaged auto in a minute. I had the opportunity to address this audience or a similar audience a few years ago.
At that time, I shared some of the key elements I believe are necessary to effectively support broad distribution. Things like technology that works for agents, great customer service, and a respected company brand. At that time, I shared Progressive's relative strength on these attributes versus some of our key competitors in the channel. Today, I can tell you that in all cases, we've either maintained that strength or in a few cases, built upon it, which is the case with respected consumer brand. A couple of things have grown more important in the channel over the last few years. One of those things is brand. Strong consumer brands are now more valuable than ever to independent agents. In my role, I get to talk to agents all the time. In fact, I'm going to be meeting with a group of them after this meeting this morning.
One of the recurrent themes in those conversations, particularly for personal lines business, is that the old methods of generating new business prospects simply aren't working nearly as well as they once did. Increasingly, agents are looking to their companies to help them market their agencies and provide them with strong brands that attract and retain customers. I think Progressive has a distinct position in the channel as a company with proven consumer marketing skills and the most recognizable consumer brand in the channel. Another thing that's changed in the channel, and we've touched on this a few times already, is the influence of technology in the placement of business. What we're talking about here are real-time comparative raters.
When I spoke to this group a few years ago, I talked about the accelerating adoption of this technology in the channel and what the implications for that would be. That adoption has largely happened. Today, comparative raters are the primary source of personal auto quotes in the channel. That's certainly true for Progressive. I also talked at that time about the importance in this kind of a rating environment of having accurate rates and really good price segmentation so you don't get clobbered by the kind of adverse selection Jim Haas described earlier in his example. Hopefully, Jim this morning gave you some sense of the level of intensity we bring to that activity every day, because it is very important. Beyond all that, comparative raters have created opportunity for Progressive. Opportunity in the form of many more quotes per agency.
That is particularly true of preferred auto quotes, which is depicted by the orange line in the graph in the lower left. Today, when I talk about preferred auto, you can apply the following general definition. Preferred auto customer is a customer who's owned a home, has maintained continuous insurance, and has a generally clean driving record, so no major violations, and no individual driver on the policy with multiple minor violations. Comparative rating is creating lots of opportunity for Progressive. To fully capitalize on that opportunity, we have had to adapt. We've had to adapt to this new agent technology. The way we've adapted is by applying the online retailing skills we have learned and developed in our direct business to this agent application. We don't own this application. We don't necessarily control this application.
What this has taken is a fairly intense level of relationship management, working with the providers of comparative rating to ensure that their quote flows capture all of our relevant rating information and identify and apply all applicable discounts so we present our most competitive rate every time. By working with the providers in this way, we have been able to consistently drive up our relative conversion on comparative raters over time. More quoting opportunities, higher conversion means more business. Here you see our growth in preferred auto applications over time, and the dark line, you see steady, sustained growth in preferred auto policies in force. That last point is particularly significant to Progressive when you consider the much longer policy retention and the lower associated cost of this preferred business. Sustained growth in preferred auto can materially change Progressive's competitive position in the channel.
In fact, preferred auto represents Progressive's greatest opportunity to grow with independent agents. All right. This is kind of a busy chart, what I want to do here is share with you the auto premium distribution by channel, that's on the left, and Progressive's estimate of how that premium breaks out between packaged customers and monoline customers. I'll call your attention to the middle box, which is the independent agent channel. What I'm doing here is I'm using packaged auto, those customers that own homes and typically require other coverages, as a reasonable proxy for preferred auto. You'd have to apply the additional filters of driving record and continuous insurance, the numbers would change a little bit, the point remains the same.
The independent agent channel is relatively rich in preferred auto, and our share of that is relatively modest at a little bit over 5%. You compare that to a greater than 20% share on the monoline auto. Our opportunity, our headroom here is significant. We can grow. The last point on this chart is on the right, and that's the fact that this auto premium in the channel is not necessarily evenly distributed across all agents. In fact, a lot of this auto premium today is controlled by agents where Progressive has not historically been well penetrated. That's our challenge. How do we grow preferred auto with the right agents? What's that going to take? One, we're going to need to fill out our product suite so that we can meet the needs of preferred customers.
I think we're doing that. Secondly, we're going to need to deliver more value to those agents that can and will partner with Progressive on preferred auto. Let's talk about filling out the product suite first. Glenn and John have already touched on our activities around Progressive Home Advantage. I'm happy to tell you today that Progressive Home Advantage is gaining real traction with independent agents. Here I share with you growth in weekly homeowner sales with our agents. Admittedly, this growth is on a relatively small base of business, but we are converting more agents every week, agents that are presenting Progressive Home Advantage and Progressive Auto as a lead market to their preferred customers. To use Glenn's words, we've proven the use case. We're on our way to proving the use case within the channel.
It's time for us to make the investment in improving the agent experience of quoting, selling, and servicing homeowners with Progressive and achieving that reasonable product integration. We're comfortable making that investment because Progressive Home Advantage in the agent channel is starting to deliver the desired business outcomes. First, we are creating true bundled customers. We've only been at this for a very short while. Already, about 7% of our preferred auto customers in the agent channel have an associated homeowners policy, and that number will grow. Those agents, those early adopter agents that have really engaged with Progressive Home Advantage, actually have a higher percentage of true bundled policies. We're getting the bundled customer. Second objective of Progressive Home Advantage was to stimulate auto growth overall and preferred auto growth in the channel. There, too, we are making progress.
Agents that have engaged with Progressive Home Advantage, and by that, I mean agents that sell two or more homeowners policies a month, have substantially higher preferred auto growth. Even though, as I've already told you, preferred auto growth in the channel is up across the broad distribution. Specifically, those agents that have engaged are producing substantially more preferred auto policies per agency. The number of engaged agents at this point is relatively low. A little less than 2% of our agents in space with Progressive Home Advantage fit that definition. The leverage with those agents is substantial. The key for us is how do we engage more agents. One important way we're going to engage them is through our Signature Agent program. Signature Agent is a program designed for those agents that can and will partner with us on preferred auto.
Signature Agent is a program we piloted in 2009 and are rolling out countrywide right now. The basis of Signature Agent is to provide those agents with more value and more benefits. Signature Agents will have access to the full product portfolio, including home and umbrella. Signature Agents will receive marketing support from Progressive, financial support, and support in terms of access to Progressive experts in the areas of advertising, media buying, online marketing, and public relations management. Signature Agents will be paid higher commission on preferred auto business. Paying these agents more commission on preferred auto, will that materially change Progressive's agent commission expense line? In the short term, the answer is no. We do not anticipate a material change in commission expense.
As more agents choose to partner with us and make Progressive a lead market for preferred customers and this program grows, yes, we will see upward pressure on commission. I would also anticipate we will see some offsetting cost reductions, including loss costs, as we assume the position of a lead market with more and more agencies. We will monitor our results, and we will adjust our pricing accordingly to ensure we make our targets. To ensure that our Signature Agents are successful, we are committed to focused account development planning. This means taking our high-touch, high-involvement sales resources, working side by side with our Signature Agents to build marketing plans and budgets. Marketing plans that utilize proven Progressive brand assets in print, radio, and particularly online.
Marketing plans that will be coordinated with Progressive's media plans, national media plans, and local media plans to ensure we maximize the impact of these agents' investment. Progressive remains committed to broad distribution through independent agents. We will continue to deliver value with products that are competitive and easy to use, with great agent technology and excellent agent customer service. Beyond that, we are now committed to developing a core group of agency partners, preferred auto partners. Partners that will be supported by more product, better compensation, and much stronger brand ties. Good. We're just a quick breath away from turning it over to you for Q&A. I'm just going to make two very quick points here. I sometimes get put on restriction that I'm not supposed to talk too much about technology because I kind of like it.
There are two points that I want to make. We're actually able to get some scale leverage. I think we perhaps maybe even got into a place that we needed to re-correct, but our intent here, just as it has been in claims and in servicing, is to get scale leverage on our technology costs. For the most part, that's less about cutting back. It's just working on the right things. That's the third bullet point, making sure we have a governance system that we're working on the right things all the time. There's lots of things we can do in technology. I'm sure you know that. The question is, pick them wisely.
The second, really to architecture, one that doesn't sort of lend itself to a discussion too much here, but we want to have an architecture that allows us in the future not to be necessarily dependent on developing every application that we need. I think in the history, we've generally developed what we needed ourselves. We'll continue to do that, but we're going to have to make sure that we have an architecture that allows us to take into account anything, perhaps developed elsewhere, cloud computing, ASI, whatever might make sense to us, because architecture is all about speed to market. We've got to be able to make sure our technology speed to market matches our ideas and innovation within the company. Things are good on technology. I think there's opportunities in pricing, the architecture will be a long-term and consistent commitment.
There's never one architecture that we can say is set in time for any period of time, but it's all about speed to market, and we're redirecting our IT resources there. I would just tell you, so far so good. No real issues. Let me sort of bring this to a close, but frankly, on the one topic that I think is the most important, we even colored it a little differently on your handout. It seems simple, really, doesn't it? We charge premiums and pay some claims. How difficult can this business be? It's all in the details. We have money, and we have people. Well, money's pretty fungible. The people aren't. The culture that we try very hard to create at Progressive is really what makes everything else work and possible.
I didn't add up, but I think there's 100 years of experience here, and we're all fairly young people, so we tend to want to stay at the company. I think that's also true of a lot of other people. These statements here, while sort of nice, they really are important. I wish I had a better way of describing what the culture really is and can be at Progressive. The foundation of our core values, which clearly is also not only on your handout, but will be on almost everything, is incredibly important to us. The one thing that's very interesting, this is tough for sort of a math engineer to start talking about brand implications, is the first bullet point. The core values combined with people who are becoming brand ambassadors.
I talked a little bit in the annual report about what I believe to be the congruence of our external brand and our internal brand. I think for us, the breakthrough in our branding efforts was that we started to actually just display ourselves. Flo is just an idealized Progressive employee. She is a Progressive employee. Don't worry about the hairdo. Don't worry about the lipstick. It's what she represents. Progressive people really get that. Really get that. I have the opportunity to spend a lot of time with different groups, and they can spin off the brand characteristics that you might expect an advertising agency to come up with. They're natural because we've got a great role model for who we are and what we represent to the public.
I can actually prove that point because I've done enough surveying to know that I don't know the specific answer, but I bet you I'm close, that we have about 20,000 people that are customer facing that will talk to customers today. I will estimate 10,000 to 15,000 of them will be asked at least once today, "Do you know Flo?" Which in some sense is a really powerful notion because the consumer is recognizing that Flo is just one of us. We, of course, have to turn that around. We have to be Flo without the hairdo and lipstick. That's a hugely important notion and has added to the construct of our culture built on values to really make us brand ambassadors.
Just a little story, just because I'm proud of it. We obviously named the stadium in Cleveland Progressive Field. As a result, we get to do something special for opening day, and employees can hold the flag and do some things at the stadium. Well, to do that, obviously, we've got a few more employees than could reasonably hold a flag in a stadium. They actually apply for that privilege by writing a small essay as to why they feel that they're brand ambassadors for Progressive. No obligation to do that. I think this year it was something like 7,000 people just submitted a small essay just to do something like that. For me, it's really about can you have this culture that is sort of building momentum all the time to be doing the right things.
I would tell you that maybe from all the things we've talked about today, even though that's the least able to be described, it may be the most powerful thing that's happening at Progressive. Here's where I think we are. I hope that that blue handout, which I didn't cover the first couple of panels, even though they're most important, I think you know those very well. Our aspiration to be a company that really does provide a distinctive service to consumers at competitive prices. Our brand is supporting that. All of the key activities on that page, that's what we do. If it's not on that page, we're probably not working on it. Those words have some real depth to them.
Maybe more importantly than all of that, and I think I can be very basic at times, and I read recently, I forget exactly where I read it. I think what that does for us, and I hope maybe is communicated to you today, the statement I read was, "The main thing is that the main thing is the main thing." What you've got there is the main thing for Progressive. We've told a lot of stories over a decade. I think we've got a lot of intensity around execution. The strategy doesn't change year-over-year. It just molds and modifies to our talents, where we are, and the intensity of execution on different pieces becomes a real powerful notion.
Obviously, I'm very proud of Progressive, I think it's showing up in the results, ultimately, that's where we have to make the dog hunt. It seems good so far. Hopefully we'll see you again next year. We'll take an hour of Q&A now, give us just a second while we move some chairs. In your package, you have a white card at the end with, or somewhere towards the end, with a Q on it. We don't see you as well as you see us with the lights. If you would hold them up, someone will come with a microphone. It is important that we get the mic there because there are people listening on the webcast. Let's try to wait till the microphone comes. We'll be able to hear you a lot better as well.
All right.
Thank you. Jay Gelb from Barclays Capital. I have two questions. First, can you talk about what potential issues there could be from a customer perspective on issues of privacy and trust when it comes to usage-based pricing? How do you get them more comfortable with doing that? Then also looking at some of the data in the appendix, it looks like personal injury protection, loss cost trends are deteriorating. Maybe you could talk a bit about that and perhaps overall, with a nationalized healthcare system now, does that mean that healthcare costs are going to get pushed into the non-managed aspects of healthcare like auto insurance?
Could you just say the last sentence?
With nationalized healthcare, is there a potential for overall healthcare costs to get pushed more into the non-nationalized healthcare aspects of-
Got it.
Like auto insurance or workers' compensation? I know you're not involved there.
Right. Let me take a shot at the usage base. John, I was thinking to add whatever I leave out. Clearly, he premised that by saying, "This is our thinking, so we've given you something that's a little bit forward-looking." We're excited about that. The real notion there was we know usage base is good. We've got the math. That's not a surprise. We've been on that for a long time. The marketing was a sense of how do you make meaningful trade-offs, meaningful to the consumer, and more than acceptable from a data integrity perspective? The Snapshot that John outlined was that the data collection period would be finite. That goes directly to one of the detractors of the service, and we don't have to go far to imagine that people don't necessarily like being continuously monitored.
You can always come up with some scheme of why someone else would be doing that, right, wrong, or indifferent. Of course, we have no interest in anything other than auto insurance. The fact that we can, I'm going to be a little careful with my words here. We can extract a great deal, and I mean great deal of the knowledge that we need to create pure premium determinations within the time period that was selected. We are also, just assume, quite knowledgeable that many of you right now are thinking, "Gee, I could behave for that long." We're also quite good at determining true behavior from behavior that may not be reproduced. You can't do these things unless you really have a lot of data and some good theories on what might be going on.
Perhaps the first 10 days of driving may not be very indicative. Nonetheless, those are sort of trade-offs that we make. That trade-off directly speaks to those, and it was a fairly large number of people who said, "I reject the notion because I don't want to be monitored." Couple that with the fact that there's cost to this device. Now, instead of having a permanent device in the vehicle, we actually get multiple turns on that piece of inventory. That changes the fixed cost add and actually reduces it because we'll literally get to use that same device. The question is, are there true curves here of marginal return for information against marginal cost?
The design that John's outlined, and I suspect, quite frankly, we'll probably have tweaks for that for some time to come, is really designed to address consumer needs that we know people want. They want to feel like, hey, we know you and your driving better than maybe your current insurer, and we'll reward you for it. That's not a problem for the consumer. They're willing to take that. Here, if we structure it as a discount, now normally a discount can be given at the time of purchase. This is a little bit strange. Now it will be a observed discount into the term. It also is giving you the opportunity to do no worse than you would have done in the initial quote. We've actually very directly taken. There's more to it than that. I'm just hitting the high ones.
Here are the things that were attractive. Here are the things that were detractors. Here are the acceptable data integrity trade-offs that we can make. Here are the economics on the turn on the technology, we try to put that in a way consumers understand discounts.
To that, Glenn clearly said we've done the studying around the product design and tried to encapsulate preferences and detractions into that design, for sure. We also watch and do our own surveying around consumer attitudes towards usage-based or pay-as-you-drive. What we see is that more and more consumers are aware of that as an option, and more and more consumers are willing to consider that as an option. Those trends have been pretty marked lately, so we feel great about that. The other key facet that I want to make sure everyone understands is that location is not part of our rating algorithm here. If you ask consumers about usage-based rating, you talk about how they drive and how much, they're all good with that. The one point where you lose a lot of them is if you start talking about location.
We have explicitly not included location in the design.
To your second question, there was two questions, hit the second. I think it was last year that, maybe it was I, it could have been any one of us, commented that we were starting to see and clearly were concerned that there would be a shift of medical costs from medical insurers to auto insurers. A simple example for those that didn't hear that, you go to the emergency room, perhaps it's a $4,000 bill, I'm just making up the numbers. Are you insured? Yes, I'm insured, and you give the name of your health insurer, Aetna, UnitedHealth, whatever it might be. They subrogate us because they ultimately determine it's an auto claim, perfectly valid to do so. We pay them $4,000. Cost shifting occurs when you go to that same emergency room, and they say, "Was this an auto accident?" Yes.
Who's your auto insurer? Now the bill is something other than $4,000 because we simply don't have those negotiated rates at that level. There is a very clear sort of first law of thermodynamics doesn't work in this case. There is actually a premium created. We will pay those costs. That's very important for us to make sure that we're very, very nimble in responding to those cost shifts. I'm not saying that's the driver of all PIP. In fact, there are many other drivers of PIP. A year later, I can tell you we're absolutely seeing that, and that we will have to be very much on top of that trend. I gave you the appendix there on frequency and severity. That clearly is a severity trend. It could relate to almost all coverages, BI, UMBI, and PIP.
We'll have to be right on top of that because I think that is a very important severity trend. I think there are other trends that will likely affect this, as I just extend your question a little bit. I think there could be very significant frequency trends as well. We're seeing changes in the licensing behavior of youthfulness. In fact, youthfulness now at ages 16, 17, 18, 19, about 20% less being licensed at that same age than they were only a few years ago. Whether that's a reflection of a texting environment or whatever it is, it's actually a lower licensing rate of youthfulness that could have an effect on frequency. I can speak to my household. I don't go to shops very often. My shop is a big brown truck that comes down the drive with anything I need, and that changes your frequency.
I talked about work from home. There'll be some very clear frequency drivers and severity drivers that on most occasions, I tell you our strength is really making sure that we're not trying to necessarily predict the trend, but be able to capture it in our analysis as quickly as possible. I do think we're going to see cost drivers, and PIP will probably be the most obvious coverage for that to come through quickly.
Hi, my name is [Keith] Credit Suisse. Two questions. One is a follow-up on the usage-based auto. If you could clarify whether the purpose is to collect data or to increase sales. Also associated with that, if you actually realize that the driver is a worse driver than you thought, you've already locked in the discount, and the discount remains even six months after the policy. How do you sort of work around that to just maintain profitability? The second question is More of a trade-off between short-term and the long term. What incentives do the managers at Progressive have to keep the combined ratio below 96? You've had lower than 96 combined ratio for a long time, and there's a trade-off, I guess, between long-term growth and extracting maximum profitability. How do you balance those two?
On the usage base, I can make it very clear. We like data a lot. We like premium more. It's absolutely about sales. We've done a lot of data collection over the years, and as we've said, we think we have the math really well outlined and in a form, again, that we can use in a broad manner. Your second part of that question was around what do we do with the people who don't drive well? Those folks will pay our normal rate. When we say up to a 30% discount, your discount is derived from your driving behavior, first within the 30 days, and then for the duration of the six months' term. As we just discussed in the previous question, we don't believe you can behave well for six months.
We think six months for sure is highly representative of your driving behavior. We wouldn't rule out potentially the prospects of resampling, if you will, sometime down the road if we think situations have changed materially. We're pretty confident we're working out the price right after six months.
As it relates-
The price will always, I'm sorry, but the price will always be lower than the price he had before. Let's say that the person had a better-
No, it could be the same.
Okay.
It's an up to 30% discount. You will get a discount based on your driving behavior, or you will get no discount. If you're a poor driver, you will not get a discount with usage-based.
Oh. Sorry. Yeah.
As it relates to the incentive question? Well, certainly it's in objectives, personal objectives, that most folks in the product management roles, control roles, et cetera, be part and parcel of that. But for all of Progressive employees, it is in the construct of our variable compensation program, the Gainshare program, that is a measure of growth and profitability. Certainly, if we were to go over our objective of 96 combined ratio, and it's even done at each individual business unit, there are penalty functions constructed in the Gainshare programs to discourage that.
Right. My view is more towards, you could do a 91-
versus a 96, what incentive do you have to maybe do a 91 and make a lot of money for the company versus, say, push for a 94 and gain maybe a 1% more growth, yet you'd be making much more money in 91. What incentive, short term versus long term?
Feel free to chime in.
Yeah.
I'll go back to the Gainshare program is also designed as profitable growth. Remember last year, we had pretty good profit margin, but our Gainshare score wasn't high, super high. It was 0.71, that's because we weren't achieving the growth that we expect of ourselves. The Gainshare program tries to put a balanced blend between growth and profitability. Obviously, this year, we're much more pleased with the growth we've achieved this year. Combined ratio is actually a little bit higher than last year, but we're pleased with that growth profit trade-off. I don't know if you want to add to that.
No. We understand the question, we understand the question very well. What's really important is to know what your objective function is. If it's sort of all over the place, it's very hard to run a large corporation like that. We have essentially a series of indifference curves in our Gainshare. I would take 91, but it depends what growth rate. 91 at a low growth rate versus a 94 at a higher growth rate, then that's where we express, no, we're not going to give it to you if that's the follow-up question. That's where we express our management preferences in the ordered pair of put and growth. It doesn't have to be too magical. We're not going to share that, but it comes directly from what we put in the annual report. First, 96 or below, then grow as fast as possible.
It's important to get that order right. Yes, as long as we're below 96, we're going to favor growth. Just to add, to make sure you're aware, if you read our releases or the filings, we also have a long-term incentive plan. There's the annual Gainshare, we call it, which we also derive the dividend from, there's also, for senior managers at the company, long-term incentives for growing faster than the market.
Thank you.
Hi. Ian Gutterman, Edge Capital. I want to go back to the segmentation, the A, B charts. It seems to me like it's a big prisoner's dilemma game, isn't it? If you only have A and B, and B can underprice or can lower the price and still make the same profit or a greater profit, obviously prisoner's dilemma says they should do so, you've brought down the profitability of the system, right? Before the average revenue was $1,000 per customer. Now it's $900. You have half the market underpriced at $1,000, half fairly priced at $800. You've lowered profitability of the system, although for the individual company, it makes sense to do it. My question is, things aren't that simple, right?
Historically, so far, we've seen it play out that way, where there's been this ability to cherry-pick, it's been good to go that route. As more and more companies do segmentation, not everyone's going to do it as well. Some are going to make mistakes, and maybe they take that $800 customer and think they're pricing it right at $700. Right? Then the company B basically has no customers. We have an entire market that's underpriced, half at $1,000 that should be $1,200, half at $700 that should be $800, and all we've done is given away industry profits. Why is segmentation, in the long term, not going to be a winner's curse?
You want to answer that? You can go, Jim, here.
Make sure we first we all agree that the loss costs of the system are constant.
Right.
Right. Just because somebody went to $800 doesn't mean the system in aggregate has lower costs. We have costs we, as a set of companies, have to cover. That's the marketplace, right? The point we're trying to make is if you can get the competitor to go to $700, yes, we won't write that risk, but they're going to choke on that in a pretty short period because they're going to be losing money. If they don't know they're choking on that, they then have to increase their overall rate level because they don't understand the segmentation that needs to happen in order to be priced adequately by segment. Yes, there could be short-term dislocation, if you will.
I don't know if it's the right word, in the long term, as long as we think our prices are more accurately aligned with loss costs, we're okay with that happening because that puts the competitor in a worse position.
Doesn't that make it hard to grow in that environment? If we have an environment where industry profit is coming down over time because of what we described, doesn't that imply that the smarter people are actually going to lose market share? There's always going to be someone who's, again, winner's curse, there's going to be someone who comes along with a lower price that isn't a good price, where the last five or eight years we've had the lower price has actually been the right price.
If you again assume the loss cost for the system stay the same and you assume that the industry profit target remains the same, then there will be winners and losers for sure. In aggregate, I don't think what you're describing happens. There are timing differences there, though, and that can be challenging. We have seen companies, knowingly or unknowingly, pricing segments below cost, and at a given time, that shows through, and then those prices end up coming up.
The other thing I'd add is a lot of the segmentation was talking about loss costs. Even if we all had the same loss costs and price is exactly the same, one of the reasons we believe a low-cost strategy works is so that the end price to the consumer is lower. That's why we focus on the efficiency of our own internal costs. We all have the same loss costs and do it the same, but we have a lower cost structure. We will get those risks accurately priced and meet our profit objectives. That's why I also want to share, there's lots of focus on the expense structure as well that gets incorporated into the pricing.
Okay, fair enough.
We like our position that way.
That makes sense. Just one quick follow for you, Brian, is the slide where you showed the 97% in direct for Q1, how does that compare to what you report in the financials? Because the monthly release is sort of like a 94%, 95%.
Right. In our monthly release, we report agency and direct personal lines, which would include both personal auto as well as our specialized products.
Okay
Like motorcycle, RVs, et cetera. That would also be included in those monthly numbers.
Okay, what you showed us, the 97, was just auto.
That was just auto.
Got it. Thank you.
Just one build on that. Let's just assume for right now, I think you get this, since we made a big deal out of it, I'd like to make sure there's no ambiguity. Brian and I are otherwise indistinguishable. Drive the same, we have the same characteristics, we live in the same neighborhood, same age, whatever it might be. A usage-based product might very well determine that there is a distinction between our driving behavior, one that a company is willing to recognize. Jim made the point, they may not even know that there's a difference in red versus blue. The costs of the company that insures Brian will make him the better driver at this point in time. That'll be just fine. In fact, they now are able to offer him a product that actually is to his best advantage, they're making a good profit margin.
I, on the other hand, just lost a customer that was providing me a subsidy. I have to work my prices up, I don't really know why I'm doing it. Why did it happen to me? Yes, you're right, the profit never works out perfectly at all points in time. From a pure game theory point of view, if we have a profit margin that we're targeting and we can attract more customers and meet that's what our commitment is to shareholders. We have a clarifying question from the web. If I get a new vehicle, does my Snapshot data roll over to my new vehicle? That's a good question. We are deriving data based on a vehicle, and if you change out a vehicle, in the near term, we will allow that similar discount to apply.
Thanks. Brian Meredith, UBS. Two questions here for you. First, could you talk a little bit about the economics of the homeowners business, the bundling package discounts? Is it the third-party homeowners provider that's eating that? Are you eating it in the auto rate? Also, are you getting any commissions or fees right now in the homeowners business? Going forward, as you advertise this homeowners business, are you anticipating you may actually try to get a commission or fee from the third-party provider?
You want to take that?
Let's reinforce the reasons that we're in that business to start with.
Sure
for retention and testing. You asked a lot of questions in there, but in short, I tell you, our ultimate intention in providing the home in whatever manner we provide it is to continue to grow and retain auto. To date, growth has happened a lot in that segment. I'd point out, important to note that the majority of the people to whom we sell Progressive Home Advantage are current customers. We are starting to try to attract the bundled household, but today it has been a tool predominantly that has allowed us to not lose them. Glenn talked about being the training wheels, if you will, for a consumer, and historically, that's been our positioning. We were there at the start, and once people got married and maybe bought a home, they have then historically sought other insurance needs, and the auto ultimately followed.
Today, instead of that happening, we can have that conversation to say, we can meet that need as well. Beyond price, we survey people who leave Progressive consumers, and beyond price, the second biggest reason is because I want to bundle. We are trying to eliminate that as a reason. In terms of structurally how we accomplish that, you mentioned getting a commission around Home Advantage. If we sell the Home Advantage program to a direct customer, there is a commission involved, so there is a revenue stream there. If our agents sell the Home Advantage program, they are the receivers of that commission. Those are becoming somewhat significant numbers, but again, I want to stress that is not why we're in that business, and that's not why we deployed this approach.
I'll also say we're going to invest significantly in what Glenn told you was sort of the household view. Our customer service reps and to some degree our agents as well, today don't have the household perspective when they're talking to a consumer. We can have sold them multiple products, but our systems today aren't set up well enough so that the rep can have a meaningful discussion about your household needs as opposed to just the policy. We've done some pilots where we have systems in place now where we can provide that view, and when you give the customer service rep that view, cross-sell rate goes up markedly. When you can provide that rep the ability to sell immediately the other products, cross-sell rates go up even more.
We think we've got a lot of opportunity to, again, sell a lot more auto, but also retain a lot more.
My second question is, can you talk about the relationship between the agency and the direct business from a profitability standpoint? For instance, you said you're willing to take your direct distribution combined ratio is above a 96 obviously, if there's some great growth opportunities. That obviously means that your agency and other products have to be below a 96 combined ratio. How do you think about that? If you have that great growth opportunities in the direct distribution segment, are you going to start raising prices on the agency and try to improve profitability or get better combined ratios there?
As it relates to the agency channel, we're continuing to price that at a 96 combined ratio, so four points or so in terms of profit margin. We are not intending to sort of have a greater margin there to subsidize growth in the direct channel. In fact, our other products, commercial auto and specialized, actually do have slightly lower calendar year combined ratios than that 96. That factors into the equation. The intent is not to have agencies subsidize direct by any means. We want to grow agency as much as possible relative to a 96 calendar year combined ratio in the agency channel.
I think the big point we were trying to make there is there's the lifetime combined ratio and calendar. We saw even though we were running direct auto at a 97 calendar, we're telling you we're below a 96 lifetime.
Would you have to cut back on potentially growth in the direct distribution if all of a sudden the commercial auto starts seeing loss cost inflation starts to pick up there or something and profitability starts coming less so?
There's always a combination of math or the algebra that you could say you have to. I mean, Brian said very clearly there's a couple of constraints, the one that's not moving is our commitment to shareholders is a calendar year 96% for the entire company. We also wanted to be very clear about what that could create for We talked about this in 1999. It was just a little early. It wasn't ready for prime time then, that in direct, it's just a slightly different animal, and we'll try to make sure that we tell you exactly what that means. The commercial and even specialized products, we absolutely run those at targets that we think are appropriate that give us enough room for the subsidy effect that you're trying to sort of mentally get through. The math actually works out without big swings.
Just to add to that, especially special lines is quite variable through the course of the year in terms of its combined ratio. You need to watch that from a calendar perspective. From those that join us on the conference call, it's absolutely fair game to ask, could you tell us more about that number? Because, as I said earlier, not all 96%s or 97%s are created equal. We're more than happy to do that. We wanted to try and provide the first introduction in the annual report letter. This is a bit more of a briefing. It just gets awkward to answer those questions unless we're all coming from a similar point of view.
Pardon me. If I can quickly follow up on Brian's question, has there ever been a case when the overall 96% calendar year combined ratio has constrained growth in direct?
Where it's constrained growth in direct?
Right.
Not recently. Not recently, no.
I guess my second question, Jim Haas did a good job of describing the increased convexity of the segmentation curve, if you will. Has the gap between Progressive's convexity and whatever the external market is, has that been increasing or shrinking in recent years?
Been increasing by a lot. I don't know. Unfortunately, slight levity. We don't know. That's a very hard thing on. I think what Jim Haas was really giving you is a construct that would allow you to sort of understand what we're really trying to do, and that's very reflective of our internal process. To suggest that we know somebody else's curve and could measure the deltas, probably the best we could do is it would take a lot of work to re-engineer sort of their formulas and then do some sort of average square error of delta between the true premium and what we actually charge. That's probably not something that any of us are actually going to spend a lot of time doing. We have to be relative to our costs, not relative to somebody else's costs.
Right.
Hi, Matthew Heimermann, JP Morgan. Two questions. First, on the usage rating, is there a particular customer segment that you think this will appeal to? Secondly, when you look at the improvements in retention over the last several years, can you maybe speak to what factors are driving that? How much of that might be mix, how much of that might be relative price stability, how much of that might be— I had a bunch of things written down, but I can't find them on the page now. Oh, product changes, discounts, et cetera.
Why don't you take this?
Sure. We think our Snapshot discount product design is actually appealing to a broad set of consumers. In the MyRate product design that we've been employing for quite some time, know that we took a lot of 2009 to test a lot of those different product attributes that we've been adjusting across all consumers to understand what were the detractors and the attributes that increase propensity. We think the Snapshot can be broadly appealing to virtually all customers. I did tell you that for MyRate, when you do attitudinal studies around who is interested in usage-based rating, the segment that we were focused on was urban affluent folks. For those consumers, they're a little more tech savvy. They're willing to understand what you're talking about, and the financial gains for taking part in such a program because they pay higher insurance rates is great.
On the retention basis, we understand the question. We have a very clear view of mix versus other actions. Not going to share that, but we do understand the sort of nature versus nurture kind of issues. With regard to the nurture type actions, unfortunately, while each of them we tend to test as sort of a single variant model and get some degree of response, once you put them together, and of course, we clearly want to put them together, we've really got a multivariate environment where it's very hard to know exactly what contributes to what. Things like rate stabilization, yes, we think it works, or we wouldn't do it. Some of the things that we used in an example of Net Promoter Score as a diagnostic where we actually recognized some electronic funds transfer issues in terms of our payment schemes and how that affected retention.
We're going to put those in. Ultimately, it's the sort of things we want to do all of them, but the absolute contribution of any one of them gets a little harder to determine. Perhaps even more pleasing but complicating factor is that I would tell you that brand has a level of stickiness to it. There are effects in conversion and retention that probably don't relate directly to some of our actions but are strengthening simply because the brand is strengthening.
Thanks. Just a question on MyRate. Is it an agency or direct product primarily? If you're a customer, do you have to change all of your cars at the same time if you have several cars? Just to pick up on a previous point, if you're a really bad driver during this test period, and you don't raise the rate, right? Because you said it would be a discount, but you're not going to give somebody an additional premium. Would you un-know that information at the renewal period, or would you reflect that next time you price?
Thanks.
Good question. We are offering the MyRate program today in direct and agency channels. It isn't in both channels in necessarily every state in which it's offered, but it is in both channels. You can opt in to MyRate or in the future Snapshot discount at the vehicle level. You needn't have all your vehicles on the policy participating. Yes, if we know you're a significantly worse driver, you will pay our normal rate even at renewal.
On the discount, if you're a homeowner and you take a discount as part of MyRate, does that interact with the bundled discount on the advantage side?
We have a lot of different rating variables. Home ownership would be one, and if you bundle, you would get additional discounts with Progressive. In terms of interactions with our MyRate or Snapshot discount rating methodology, today, there are interactions, Jim Haas, correct me if I'm wrong, by customer segment, if you will. Let's generalize and say youthfuls versus more mature drivers. If you are better or worse than average as a youthful, your driving behavior looks significantly different than better or worse than average if you're a mature adult. We're interacting with some of those variables. I don't believe specifically we're interacting with homeowner, but you clearly get a homeowner's discount if you're a homeowner. If you bundle with Progressive, you get an incremental discount. I don't believe we're interacting there with usage.
Thank you.
Just a minor follow-up to your question. We are very aware of the idea if you reduce the rates for some, you have to increase the rate for others. Most of the people who come to shop for us obviously currently have somebody else's insurance, which is good from our perspective. Ultimately, the number of people that we give discounts to being offset by those that are at current rates is probably not going to start swaying this balance over any quick, immediate time period. There will be a renormalization over time, just as it was with credit. To the prior question, or maybe the first question, the idea of saying to a consumer, "Hey, we really have a neat tool. Let us take a look at your driving, and we might give you a discount, or we might increase your rate." That doesn't fly.
We can sit here and say academically it's pure. What we have to do, and that was the point I was trying to make before, is find acceptable trade-offs so that the math and the marketing are in sync. The only thing I know for sure is we won't have it perfectly. We don't have anything perfect. This will be very exciting, and John talked about the reasonable intellectual property protections that we have as we can sort of get this into the marketplace and move from a target to a more broad base. Frankly, obviously, I'm biased, I think it's a very exciting proposition, I think we will be talking about this for some years to come as we continue to find the model.
How's it going? Thanks. Keith Walsh, Citi . Glenn, I will save the LeBron questions for after the conference. First off, just want to focus on retention. When I think about the slide you put up on page 14 about claims quality, is there a strong correlation between these drivers and how you retain customers, or should we just be focusing more on purely price driven? I know you talked to this a little bit earlier. Then, do your customers shop every year? Are they going out there year after year looking for lower cost, or is it more of a service factor that's going to keep them there? Then finally, within the direct versus independent agent space, are there differences or magnitude of retention factors within those spaces we should be thinking about? Thanks.
Maybe we'll split them up. I'll go with the claims question. The observation in claims is actually very powerful. I think we've shown, but maybe a couple of years ago, that our retention rates for people who have had a claim with us are some of our highest retention rates. Unfortunately, I really don't want to extrapolate that to the entire population of 12.5 million people, that could get expensive. It is a testament to the claims service that we offer. It's also true, although I think that we still have a lot of things that we could do better than we're currently doing to actually attract the claimants that we serve in the claims environment. Obviously, we serve our own customers, but we serve customers that our customers have created a problem for, and those customers rate us very highly as well.
Claims very definitely is another contribution to retention.
Data point I love sharing. If a non-customer experiences our Concierge Claims Service, we hit somebody and they go to concierge, they are 50% more likely to show up within six months as a Progressive customer than the average consumer in that same area. Claims service can absolutely drive sales. We call that brand, right? It also absolutely drives retention.
Second, please.
What about the differences independent agency between direct and independent? The differences in retention factors, independent agency versus direct.
Sure. In aggregate, our retention, our policy life expectancy across our agency business and our direct business is different. What you really need to get down to is it different by segment? For the same set of consumers, is it materially different in agency and direct? When you get down to that level, they're, I'd say, more similar than they are dissimilar.
Thank you.
Joshua Shanker, Deutsche Bank. I was wondering if you could give us some industry background a little bit. Where's internet purchasing right now as its share of the industry, the way people buy their auto insurance, and what's happening generally in the industry in terms of retention compared with what you're seeing at your company right now? A second question is how you prioritize hybrid debt, retiring debt, giving a special dividend, and share repurchases with what you're going to do with excess capital. A third question, the future of the Spanish language audience in the U.S. and targeting that market.
Online shopping, we have our internal trends. Obviously what we do in terms of marketing online and offline does influence how consumers shop with Progressive. I presume a lot of you read industry data. Comscore is an example, and I think you see there that the trend is, generally speaking, there is a little ups and downs, but over time, it continues to grow in terms of online quoting for sure and in online purchasing. That said, I want to go back to the fact that if you look at the business available to Progressive today in terms of where consumers are across independent agents and direct, independent agents have about 50% more business than direct companies in aggregate have today. While a lot of trends show online shopping
Growing, it is for sure. Our potential to grow with our independent agents is still huge and is predominantly in the preferred segment as John Barbagallo was telling you. Mike Hudson?
As it relates to the prioritization of capital management activities. Certainly, in the past, you'd say share repurchases has probably been the biggest component of that. Then 2007, we did the extraordinary dividend. Shortly after that, we came up with the variable dividend component, which Glenn has articulated could be the largest regular dividend at the end of this year. Share repurchases will be still part of the plan. It's just in this instance of trying to retire some of the hybrid security, we put that as part of basically there are some little things that we can do in terms of capital management. It's not to say that we won't do share repurchases. We've had some share repurchases throughout the year so far. Maybe increase the activity a little bit, and that will continue to be part of it.
I think the variable dividend is our dividend vehicle. We haven't thought about an extraordinary dividend. We did it four years ago, and extraordinaries, you don't do it every year. I think the variable dividend is our dividend component piece for now, and you've seen what we've tried to do in terms of retiring a portion of the hybrid security.
The last question was around?
Spanish language advertising.
I'm sorry, say it again.
Spanish language advertising.
Oh, Spanish language advertising. We've done that over time, and frankly, I think that's one of the things that we haven't necessarily done well in the sense of we haven't ended with the objective of that. As many of you know, we're actively looking for a CMO now. I've got a great opportunity to talk to some very talented people. That's high on the list to see how we might go about making even greater penetration in the Hispanic population. Something we're very aware of. We've done multiple times in the last 10, 12 years. I couldn't say that was one where we could put a nice slide up and say, we're really making the breakthrough. We've got a lot of segments. We talked about pets. We've got several other segments. We're making great breakthroughs. We certainly attract the Hispanic population, mostly through our agents.
We'd like to give them a more valid option through our direct channel. That's high on our list of things to do.
Paul Lushin from Sandler O'Neill + Partners. If you could go into a little bit more detail off of Josh's question about the capital management prioritizations that you currently have. I don't know if the tax laws change anything with that as well, but you're at your desired or above your desired capital level. This change in your hybrid security is a pretty small number. You're pretty profitable. The question I think we want to know is, are you going to actually do something and when? In lieu of that, give us a little bit more detail as how you think about it.
Well, as I said before, it's a good position to be in. That we have the capital that we do have, we're very pleased that we've actually created lots of capital since a year ago or two years ago. In terms of how we think about it, obviously, with the tender, we did something different than we've done in the past. We haven't retired debt in the past before maturing. In terms of share repurchases, that's just an ongoing practice for us, and we don't forecast how much we're going to do in the future. It's just part and parcel of our practice and you sort of see it as we report it monthly. I think what we have tried to do, at least with the variable dividend component, to actually make it a little bit more meaningful component of our capital management practices.
If you were to look back four or five years ago, our dividend was a relatively small component of any capital return, relatively small. The extraordinary dividend was obviously a very big event. Now the variable dividend is a pretty significant return of capital in that form when and if it's prudent. We're actually very pleased with that activity and the structure of that plan because in 2008, it wasn't part of the program because of the constraints we put in place. In terms of prioritization, it's hard for me to say this is one, two, and three. We have a full suite that we can participate in. Obviously, as I've said, share repurchases has been the biggest in the past, and you see our actions of what's going on with the debt repurchase and the variable dividend component.
The magnitude of how much it will be. Certainly, we're very pleased that we have more capital, and we're very pleased with that capitalization. I think certainly the activities in the investment marketplace in 2008 and 2009, if nothing else, gave us a little bit of cause and caution for how much capital do we retain and at what pace do we return it, right? I think if nothing else, that at least gave us pause to think about how fast and at what pace our actions are.
Just to be clear, versus a couple of years ago, you're going to be more cautious and more careful, more probably slower in returning capital than you were in the past, given obviously the crisis. Past performance does not necessarily predict the future.
I think we try to learn from the past. Let's just say that, and it influences. The extraordinary dividend that we did in 2007, right now we're not thinking of that as part of the next play. We will continue to reevaluate as time goes on, and I hope what really happens is we generate more and more capital and it continues to be a topic of discussion of how we return it.
Just to put a slight caveat on what appeared to be a characterization that wouldn't be a lot. Our tender is up to $350. Could go higher. We also have very clearly signaled by then that we will have removed a restrictive covenant, assuming that that happens. It's absolutely critical from our point of view that we're very transparent about everything we're doing, but that gives us an opportunity to maybe buy debt beyond the tender too. If you take what the potential could be for the variable dividend, take stock repurchase, and a significant amount of tender. I don't know that any of those necessarily will happen that way. I think we're talking about a substantial amount of money. I know it would be great from your perspective to have greater certainty. That's not how it works. I know you know that.
Take us at our word when we say, when we feel we have excess capital, we return it to shareholders. We do what we say.
We're running a little bit long, understand if folks need to leave. I thought maybe we could take another question.
Sure.
I was wondering if you could give us an update on your Australian venture and how that's going and
Sure. This is probably one of the updates that's a little hard to give. We have something in the less than 1,000 customers category. I don't generally use it as a primary talking point. However, what we did in Australia, let me be very clear, we have taken a relatively small amount of our resource and capital and asked ourselves the question, can we build a model of insurance distribution for the future? It's an internet-only model. Everything is served on the internet. We run the internet out of Mayfield Village, Ohio. Maybe we should make more of the fact that we're outsourcing to the United States. It's going to be quite a challenge to develop a brand from scratch in a new country. There are pieces of the infrastructure claims specifically where we're not going to overrun our ability.
We want to make sure that we can get the claims done. I would tell you, think about our Australia as an option for the company in the future as a process option. Don't be putting it in your model as a line item. It's not going to work on either the revenue or profit, which I don't think we'll see for some time. Having said all that, I actually had a chance, personal visit, because we can't afford me as an expense on that P&L to go to Australia, and I had a chance to sort of do some press there, and the press are very, very welcoming of us. I think we're very welcome to that environment as an alternative option for consumers.
I spent a fair amount of time with the regulators there who are exceptionally proud that they didn't step down a well during the financial meltdown. They were very picky in who they let come into the country, and I think we had very constructive dialogue. All of those things cosmetically are working well in Australia, and we're just going to find out what it's like to be a small guy again because we've got to act and think like an entrepreneur, and ask the question in a couple of years. I think the answer will be a little more insightful.
I think we have one more question over here.
It's Harry Fong from Roth MKM. In the illustrations regarding lifetime combined ratio, could you interject how retention affects those numbers? How did you arrive at the 4-year timeframe for calculating lifetime combined ratios? I suspect you have different retention ratios for different products.
Right.
How all of that might come into play.
Right. Yeah, just so everybody's clear, that was just an illustrative example. It wasn't the true policy life expectancies. To your question, if the retention improves, which is why we continue to say let's continue to improve, increase our retention, and assuming the margins on renewals are such that they are, it either lets us do one of two things. One, increase the amount that we could be willing to spend for new business acquisition, which can drive future growth. We could increase our target combined ratio for new business if our policies extend longer. We could price our renewals to a higher combined ratio. You could do either of them. So far, one of the things that we have really done is more focused on trying to generate more new business.
Policy life expectancy enables us to increase the new business combined ratio target, therefore spend more. You're right, we do it at a segment level or product level, et cetera, and the policy life expectancies do vary quite a bit by segment.
I'm being told that's it. Simply thank you for allowing us to share a little bit of the company. There's a gentleman standing at the back there that many of you know, but more by voice than by sight, Pat Callahan. Patrick has answered a lot of the questions that probably I managed to confuse you on at some point on a conference call, and you've had to call back Patrick for clarification. Patrick will be moving on to a product management role in Progressive. That's his farewell in that responsibility. Right behind him is Clark Khayat will take over your follow-up calls on a going-forward basis. Just a quick note to recognize there's a changing of the baton there as Patrick goes on to some new responsibilities in our commercial