The Progressive Corporation
Q4 2017 Earnings Call Transcript

Published: 2 Mar 2018

[Julia Hornack:]

Thank you, Latoya, and good afternoon to all. Today, we will begin with a presentation about efficient marketing to maintain a leading brand. Our presentation will last approximately 45 minutes and be followed by Q&A with our CEO, Tricia Griffith; our CFO, John Sauerland; and our guest speakers, Dan Witalec and Cat Kolodij. Our Chief Investment Officer, Bill Cody, will also join us by phone for Q&A. As always, discussions in this event may include forward-looking statements. These statements are based on management's current expectations and are subject to many risks and uncertainties that could cause actual events and results to differ materially from those discussed during the event. Additional information concerning those risks and uncertainties is available in our 2017 annual report on Form 10-K, where you will find discussions of the risk factors affecting our businesses, Safe Harbor statements related to forward-looking statements and other discussions of the challenges we face. These documents can be found via the Investors page of our website, Progressive.com. It is now my pleasure to introduce our CEO, Tricia Griffith.

[Tricia Griffith:]

Thanks, Julia, and welcome to Progressive's Fourth Quarter Webcast. During the last few webcasts, we talked a lot about top line growth as far as revenue. We really outlined on the addressable market on both the auto and home side. It's almost $300 billion of which we have about $30 billion, so a lot of runway, and we're very excited to continue to execute on that plan and continue to gain market share. For today and the next webcast, we're going to talk a little bit more about bottom line growth, specifically efficiency. Today, we will be on brand efficiency and in creative. But before I go into that, I want to talk about how we think about winning at Progressive and winning in the right way. And we use what we call our four cornerstones

[Dan Witalec:]

Thank you, Tricia. So why are we talking about marketing spend today, as you likely know, we have seen a real explosion in industry, insurance industry, advertising spend over the last two decades. This chart shows U.S. P&C insurance advertising from 1996 through 2016, the last few that we had, industry spend available. And you can see, there’s been about an 11% average annual increase in advertising spend and the insurance industry. When you compare that to overall advertising spend in the U.S. only going up about 2% or 3% per year, you can really see why insurance has become one of the most heavily advertised industries across the entire country. The rest of it’s certainly been a big part of that, and increase it's spend along the way actually slightly outpacing industry spend during this time. So that’s why we’re here with all of this spend going on, we want to make the case if you’re spending it efficiently. And we will do that by hitting on these four different topics. So I am going to talk a little bit about the discipline controls that we have around our media spending, then I am also going to talk about our media mix, and then I am going to turn it over to Cat Kolodij to talk about some of our efforts beyond auto and our creative network. So let’s start on discipline, the natural question is, are we getting a good return for all of that spend that we have going on? So one simple way to look at that is, are we getting incremental sales as we spend more? Here is a chart that shows a scatter-plot that has media spend on the horizontal axis and direct auto unit sales on the vertical axis Each dot here is a year, 2001 is in the lower left hand corner, 2017 in the upper right hand corner. So the great news for us here is that in general we see a linear relationship. We have seen and we continue to see generally a linear relationship between spend and sales. So as we spend more, we are getting incremental sales. And in fact, if you look at 2017 in particular we had a large increase in our media, but we’ve got those incremental sales to go along with it, even slightly above the line here. So we generally feel great about that incremental benefit we’re getting. So what are some of the controls that we have to ensure that we are spending efficiently? So at a high level, our most important single control on our spending is looking at our cost per sale and ensuring that that is less than or equal to our target acquisition cost. Let me explain each of those. Our cost per sale is simply our total acquisition cost which is mostly media, but also includes other items like the cost to generate new creative, and we divide that by the direct auto unit sales in a given time period. So, we want our cost per sale again to be less than or equal our target acquisition cost. I think about our target acquisition cost simply as our maximum allowable spent we have to acquire a new customer. It’s also how much we recover for acquisition expenses over the life of a policy. So it is fundamentally tied to how we price and that is obviously very important for us, but this all fits into our overall pricing and economics. Now I am going to show some data at a high level in aggregate for our business on cost per sale versus target acquisition costs. But note that we actually have the data and make decisions at a very refined segment level. So for example, a auto and home bundled customer and a high premium state is going to have a very different cost per sale and target acquisition cost than a auto insurance customer who rent in a low premium state So let’s look at some data on cost per sale. So this is our cost per sale from 2010 through 2017. You can see it’s actually been pretty flat over that time period despite some pretty significant increases in spend. This is actually one chart of my presentation I don’t want to rock it up until the right. We are actually pretty excited that we've been able to keep cost per sale relatively flat over this time period. Perhaps the bigger story here is that we've actually seen a nice increase in our target acquisition cost over the last several years. Now remember, we generally want our cost per sale to be less than or equal to our target acquisition cost, so we actually feel great about that gap. And I think it's really a testament to many of the Destination Era strategies we've been talking about, playing out. So what's driving up that target acquisition cost? Probably, most notably, our retention has been increasing significantly. This has long been a goal of ours, and we are really seeing some nice benefits over the last several years. In fact, our retention is up about 15% relative to 2013. When you have customers stay with you longer, their lifetime profit increases. And so you can spend more to acquire new customers, and that is certainly what we're seeing with retention increasing. On a related note, we are also seeing a nice shift towards longer-retaining customers. So at the last investor call, Heather Day talked a little bit about our Robinsons segment. That is our home and auto bundlers. Now they are our smallest segment right now, but they are growing the fastest, and that's a segment that retains the longest. So you can see on this chart that Robinsons segment is growing really quickly. They retain longer as they become a bigger portion of our book. That is certainly contributing to our retention as well. So another way to look at our marketing spend efficiency is to look at our total acquisition expenses in a given year and compare that to the projected lifetime premium of a new cohort of customers coming in, in that given year. If you look at that over time, it's actually been coming down for us. So again, we feel great about this, a nice testament to the fact that our Destination Era strategies are really paying off for us. While cost per sale relative to target acquisition cost is our single most important control. We look at several other things that could potentially constrain our spending. So first of all, we need to make sure that our direct auto lifetime combined ratio is at or below a 96%. So cost per sale and target acquisition cost refers to our acquisition expenses, but we also want to make sure our other parts of the combined ratio still all fit together to be at a 96% or less. And we feel great about that. We also need to be able to service our customers well. It's rare, but there have been times where we're growing so quickly in a state that we don't feel good about the staffing levels. And we have to pull back on our growth by pulling back on media spending. This hasn't been the case recently, but we always put customers first, so we are willing to do that if we need to. And then the final constraint here is we need to make sure, as you well know, we have an aggregate, company-wide, calendar year combined ratio at 96% or less. In 2016 in particular, we were worried about our calendar year 96%, and so we did pull back on media spend that year because of that. So those are 3 key constraints. I also want to talk about a fourth one and that is really thinking about our incremental cost per sale and ensuring that, that's sufficient. When I talk about incremental cost per sale, I'm really thinking about the last dollar we spend, making sure that, that is a good use of funds relative to alternative uses of those funds. This is a harder thing to measure, but really critical for us to ensure that we are being efficient overall. So as I talk about incremental cost per sale, that's a natural, segue into our media mix. So media mix is simply of our total marketing spend, which particular media types are we spending those dollars on. Is it TV, digital, direct-mail, or other media types. So any presentation on media mix has to start with this quote attributed to John Wanamaker, many-many years ago half the money I spend on advertising is wasted. The trouble is I don't know which half. Our goal really is to try to understand every dollar and make sure we have little to no wasted media spend. So how do we accomplish that? Our main tactic in trying to really understand our incremental cost for sale and making sure that we're efficient is focusing on randomized controlled tasks or AB test. So what's that AB test in the media space? It simply means that we will have a group of consumers call them the test group or the B group that will receive one of our ads and we'll compare the results of the effect of that ad to a control group -- the A group that doesn't receive that ad. So the difference between those is the lift of that ad. Now AB testing is not new as long as we've had Progressive.com or auto quoting funnel, we've been doing AB testing and we continue to do them heavily to this day. In media however it's traditionally been quite a bit harder to actually do AB testing robustly, especially in our some of our off-line media. So I'm going to talk about that in a bit, but let me first just give you an example of why we think AB testing and incrementality is important for us. So here is a fictional car shopping site where you can see a Progressive ad in the lower right-hand corner. So traditionally how we measure the effectiveness of that digital ad, we can actually see when or customer or when a consumer goes to the site and sees our ad and then comes to Progressive.com and actually buys an auto insurance policy from us. So when we measured in that way in the past, we actually felt really great about these car shopping sites and they're effective at digital ad because we saw a lot of sales after somebody visited a car shopping site and saw our ad. However, several years ago as we thought more about incrementality, we added a new layer to this test. We had a holdout group or a control group that saw a public service announcement ad had, like this ad here for the American Red Cross, and that group we looked at how often they actually bought Progressive insurance from Progressive.com when they saw this public service announcement ad, and we found that there actually wasn't that much of a difference in terms of sales for people who saw the Progressive ad on that site, versus those that saw a public service announcement ad. So this was a great insight for us, it led us to advertise that much less on these car shopping sites because we saw that we didn't get true incremental sales from that advertising. Now the main point of this isn't specifically for this example, I think this example is unique to Progressive's particular situation and the power of our brand and our other advertising, the fact that a lot of these people who came to the car shopping site, came to Progressive.com, regardless of whether we had a digital ad. The bigger point here is that having that system in place is the best way for us to really understand the effectiveness of our ads. So I guess the other thing I'll say on here, you've likely heard various news articles over the last several years about how problems in digital media reporting, so there’s some thought or other reporting issues that may cause wild miscounts of impressions or clips. This is a big problem in the digital advertising industry. We are less vulnerable to those types of issues because we are measuring incremental sales directly and we get that data on incremental sales without having to depend as much on our partners’ site, so just another reason why incremental media is really important for us and incremental sales are really important for us. So let me now turn to TV, so like you've also heard a lot about TV and various things going on there. Here’s some data that we look at on what’s going on with TV. So if you look -- this data shows total hours per week of viewing, a traditional TV in the blue line versus non-traditional TV in the orange line. And for non-traditional TV, think of, Netflix or Amazon Prime Video. If you add those two lines up, you see that total viewing has actually been relatively constant over this timeframe, from 2002 to 2017. But we are seeing a clear shift from traditional TV to non-traditional TV. Now as an advertiser the issue with that is that non-traditional TV and again think Netflix and Amazon Prime Video, in particular generally has less advertising available than traditional TV. So that means that there are fewer impressions available in the marketplace, the supply is down. Demand for this content is actually up or at least constant which means that the cost of this media is going up and we may see that typically in CPMs or cost per thousand impressions. That’s a difficult situation to be in as an advertiser. If the cost of your media is fundamentally going up and in some cases pretty significantly, but we still have to hit our three year cost per sale targets. So how do we deal with that? We have really responded by trying to be smarter about how we buy TV and building a dataset to help us to do that. So let me take you through the dataset that we are building out and how we are approaching this problem. So we have always known which Progressive ads are on which shows. Obviously we buy the media directly so we have a good view of that. The new dataset that we have is from set-top box data or think of digital video recording devices DVRs from your cable companies, they enable us, they have data that enables us to see which individual household are watching which shows. Now we’ve always known that in aggregate across the population but we haven’t known it previously at a household level. So if you combine those two datasets, you actually know which households are seeing which Progressive ads. We also have a great view of which households are buying Progressive policies. So if you combine that Progressive sales data with the other data I talked about through a third-party matching service. So we actually don’t see any of the individual household data, that’s all anonymous to us, but we are able to wending, which Progressive ads are linked to actual Progressive sales. That may seem like a pretty simple thing and honestly in the digital space we’ve had this for a while but in TV this is a real step forward for us. This type of data allows us to do things like look at a frequency versus response rate curve in TV and frequency -- and let me just explain real quick, that’s not how many accidents you get into in a given year, in this context it is how many Progressive TV impressions, TV ad impressions based household sees in a given week. So you can see, as they see more Progressive ads, the good news is they're more likely to buy a Progressive policy. But there is a diminishing returns nature to that curve. So that first Progressive TV ad impression is worth a lot more to us. It's much more likely to get a boost in Progressive sales than the 9th or 10th ad that a viewer may see in a given week. So the fact that that's the diminishing returns curve isn't new news, but having the specific quantification of it is very valuable for us. We also can now see for individual programs, a frequency histogram. So we know what percentage of viewers are seeing their first Progressive ad on that TV show in a given week versus their second, third, fourth, fifth, or sixth. Let me share a quick example of how we actually use this to make decisions. So let's take a given program that we're advertising on with a particular Progressive creative. Let's say it has 1 million impressions, a cost of $15,000. So, it CPM is $15. We've always had that information. The new information is we now have a good view of the frequency histogram for that particular show. So we know how many viewers are seeing that ad for the first time in a week versus the 10th time. And then based on the frequency response curve I shared on the prior slide, you can actually calculate the incremental sales from that TV ad and then the incremental cost per sale as well. So let's compare that program A to a different program. So consider it maybe a high-profile award show or something like that, that might be a little bit more expensive. So this case -- in this case, program B has a CPM of $18, so more expensive than program A. But the frequency histogram is very different in this example. You have a lot more viewers who are seeing their first or second Progressive ad of the week on that program and remember. Those first impressions are so much more valuable to us. So we're actually able to put a value on those and actually generate more incremental sales from this program than program A and, actually, despite the higher cost, have a lower incremental cost per sale from program B relative to program A. So this is a really powerful data set for us and really shows how we're able to tie some great external data to our Progressive internal data and knowing our economics to make the right business call not just on cost of impressions, but really cost of incremental sales. We actually buy a lot of our media in-house. And having this type of data available to our buyers is hugely important. In many cases, as the buyers are negotiating with networks, it may mean that they simply walk away from deals if we can't hit our incremental cost per sale target. In other cases, it means we're able to negotiate harder knowing exactly what we need in terms of a price to be able to hit our goals, and it's been a really powerful benefit for us. So that's a TV example. To be clear, digital is really important to us as well and, in fact, has been growing quickly as a part of our overall media mix. In fact, digital is our single largest media spend category right now, and there are several kind of established platforms that we're on. We're also constantly testing at newer platforms. So video, we referenced before, certainly, YouTube, Hulu, others. We are advertising on and continuing to test there. Mobile is a huge portion of the new quotes that we get. Actually start on at mobile device, and so we are advertising heavily on mobile. Social is certainly another big category especially as we go into homeowners insurance. We found some new social networks that are particularly interesting to us in that arena. Games are not just for teenage boys anymore especially in kind of the mobile social era, games are huge we advertise within games but we've also created several gains from Flo or Superstore that our customers can play and are natural advertising tool for us. And then certainly voice as that really explodes recently we are testing with voice assistance and in podcast and other areas so a lot going on for us in the digital arena. So in my portion of the presentation I really focus on incremental cost per sale or the acquisition economics to bring in new customers. What's was really exciting for us as we look forward though we have new data sources a new information to make some of our media and marketing decisions. We for example, now can often understand not only the impact of an individual media type on acquiring new customers, but also its impact on retaining existing customers. I shared earlier the importance of retention in terms of driving our overall budget and our allowable target acquisition costs. Obviously, if you have an ad that can increase retention it becomes that much more valuable to us. In addition, increasingly in some media we're able to tie a specific media type to individual customers that that media type is bringing in so we can then look at the profitability of those specific customers by media type and really factor that in along with retention and the acquisition economics to a unified lifetime value model that helps us make smarter decisions in media. So for example, we may have had a media type previously that had a great low incremental cost per sale, but then we may have found the customers that we were bringing in from that media type at a high loss ratio or low retention which certainly would make us feel very differently about how much we're going to spend for that media. Where we're really going is expanding the lead that we've always had on the pricing side of our business and segmentation and really applying that to the marketing arena. I started at Progressive as a state product manager and segmentation the importance of segmentation was drilled into my head you know I think we have a ton of opportunity to really take those learnings on the pricing side of our business into marketing and ultimately get to the right message for the right person at the right time. That's not an easy task and we're not completely there yet, but we have the right team in the right systems in place to really get there over the next several years. I couldn't be more excited about the opportunities ahead of us. Now the marketing and media engine that I talked about wouldn't work for us if we didn't have really effective creative that resonate with our target consumers. So I'm going to turn it over Cat Kolodij to talk about that.

[Cat Kolodij:]

Well, thank you, Dan, I think we've made it pretty clear line why marketing loves to work with acquisition, in fact Dan did a really job of explaining how we measure and distribute our marketing message is critical in today's highly competitive environment, now we're proud to be able to do so efficiently and effectively because of this disciplined approach to media planning and buying. As we flex our marketing agent for the Destination Era, what we say and how we say it is going to become as important as where we say it. I am thrilled to be able to take some time with you today in order to talk about how we are expanding the brand so that we are known for more than just auto, and also, how we are evolving our creative platform so that we really can manage our network just more likely to connect with more types of people. Now with everything that progresses, we’ve taken disciplined approach to both message development and creative management in order to generate growth. Let’s talk a bit about how. Let’s begin actually by talking about how we are efficiently expanding our brand beyond being known as an auto insurer. Now a moment ago, Dan talked about the importance of driving shift in our customer mix. In order to attract additional customers segments, we actually need to improve how we talk about our products and the kinds of services that we provide in order to meet the new and emerging needs of these segments and specifically, meeting these as they evolve over their lifetime. This means we need to expand what our brand is known for, which means we have to begin with the customer segmentation. Now as a reminder, you’ve probably seen us before, if you've actually been participating some of our webcasts, but we have four major consumer segments. The first two Sam and Diane have simple insurance needs. Sam is our frequent auto insurance shopper. There often driven by price. Diane are our biggest customer segment. They also are interested in price but for very different reason. Because Dianes are trying to save money in order to fund major life events. So both Sam and Diane are interesting in our price base message but for very different reasons. And both Sam and Diane will always be our core audiences. They remain vital to our growth and in fact most of our broad-based marketing is targeted to their needs. Now emerging segments are our homeowner segments, Robinsons who bundle their home and their auto together with a single insurance company and Wrights who are auto owners but they actually have insurance for their homes with a different carrier. Now homeowners as a group tend to really not want to shop a whole lot. So they are really great for our customer attention but not really good for our customer acquisition, cooperated broadcast. We’ve actually talked a lot about our strategy to focus on the younger segments and we continue to do so today. Because of the inert nature of homeowners we focus on acquisition efforts on a group that we call Future Robinsons. With a more favorable opinion of Progressive and a more active shopping behavior we actually think that this segment is perfect for us and shows a great promise. Now our strategy is to attract them and retain them so that as their needs change, we actually congratulate our Future Robinsons into Robinsons. But to do this we have to first understand who are they and what do they need and in particular what are the Robinsons need? So who are the Robinsons? Representing 41% of the auto and owner households in America and 7% of our customers, now that’s the percentage of households that we have, not necessarily the percentage of premiums, the Robinsons segment bundles their auto and home insurance as a single provider. Now savings do remain important to them and is a key driver of bundling. Savings is not the only they bundle. In fact many of Robinsons bundle because they do have complex insurance needs. And by keeping all of their insurance with one insurer, it actually just makes things a whole lot easier. What is interesting is many Robinsons need to insure much more than just their auto or their home. In fact, our research indicates that up to 16% of Robinsons have other things in their garage, other things that belong in our special lines product portfolio, like motorcycles or boats or RVs. The thing is that, given our leadership position in special lines, we feel really good about our ability to cover all of their needs, whether it's auto, home or even more. Now as you might imagine, it's going to take a little bit of time for us to build our reputation among the more mature Robinsons. However, the same isn't true for their younger generations. In fact, if you take a look at the future Robinsons, their needs are much more aligned to what we can offer today. Let's take a look at the chart on the right. If you take a look at the red dots with the dark outlines, those are the Robinsons. You can see their needs diverge greatly from that of the other segments. Now if you take a look at the red dot with the hash, that's actually our future Robinsons. Their needs are much more closely aligned to the other segments that we've always been very successful with. And we think because of that, we're going to be building on a solid foundation with our future Robinsons. Actually, let's take a closer look. You probably already understand that Progressive has always been known as a value brand. This is a position that we really do want to maintain in the future years. Historically we've actually put a lot of money into our savings or price-based messaging in order to build up this value reputation. But the thing about value is that it actually has two parts. Price is very important, but so too is quality. Now we know our reputation as a quality provider will be increasingly important to us as we grow. After all, I just got through telling you that quality is actually quite important for our Robinsons. Here, you'll see a chart illustrating the degree to which people believe that Progressive is actually a quality brand. Here, quality is a 10 attribute number that include statements like

[Julia Hornack:]

[Operator Instructions] Before taking our first participant from the conference call line, Tricia and John would like to answer a question that is likely on the top of many minds.

[Tricia Griffith:]

So John and I are were talking a couple of weeks ago and considering a lot of the questions we're seeing around the Tax and Jobs Act of 2017, so we thought we'd get out in front of it and how we think about it really from all three constituencies, which all we are both aware both customers, employees and stockholders. So we're going to answer sort of a high level. If you have other questions, we can answer them. We also have our resident expert, Jim Cruzman in the audience, if you want a deep dive into anything tax-related. So let's think about investors. So as a growing company, we need more capital for more growth and to satisfy our regulators. So any profits will help with that. It's always our first choice to reinvest in the Company. In addition, our annual variable dividend considers the tax rate. So all things equal, we'll be about 21% higher. So that's really how we think about it from an investor perspective, growing the Company and our annual dividend. When we think about customers, we sort of put customers and communities together. So we have a Progressive foundation. It's been in force for over 15 years, and our employees really like it because they're able to give and get a match up to $3,000. In fact, the average amount we paid out of the foundation over the last five years has been $4 million a year, and they can give us as long as it's a not for profit of 501(c)(3) to either national or local communities, a lot of people love giving to the local communities, which, of course, where our customers are, to churches, to schools. And so, we continue to fund the foundation. And because it's based on underwriting income it will, also all things equal, increase about 21%. In addition, it is the Tax and Jobs Act. And so since the beginning of 2016 to today, we have hired over 13,000 new external hires. So having those jobs, and they're not just in Cleveland, they're around the country, specifically in the claims organization and the CRM organization, having those robust jobs, we believe, really help the communities that we serve. And lastly our employees, so we are market-based for both compensation and benefits and will continue to be market based. As the market changes, we will shift and we'll watch any inflationary trends. We'll look for key indicators like high turnover in an area of people not accepting our jobs. We no longer ask for your prior compensation, but people might give us anecdotal information, so we are prepared to shift as the market shifts. In addition and this is related to the tax plan, but how Progressive has always thought about employees is really to our gain share plan. So, I know a lot of companies gave one-time bonuses which I think is great. We've had a gain share plan for over 25 years and in fact last year the gain share of 1.79 our medium bonus was $8,100. So our philosophy has always been when we gain we share with our employees and our shareholders. So that gives you a kind of an overview of how we've been thinking about the tax changes and now if you want to open it up for other questions that would be great.

[Julia Hornack:]

Great, Latoya, we will now take our first question from the conference call lines.

[Operator:]

The first question comes from Ian Gutterman of Balyasny. Mr. Gutterman, your line is now is now open.

[Ian Gutterman:]

So I guess first just to follow up on what you're talking about on the employee growth. I think there was a newspaper article around you and they suggested, you were hiring 7,500 new people, which is I guess a 20 plus percent growth rate and given there's quite a little bit of wage in there too. Should I -- does that put pressure on expense ratio or I'm reading too much into that? Is that a gross number, not a net number? There's attrition that offsets that just how should I think about that?

[Tricia Griffith:]

Yes, Ian, there will some attrition to offset that. What we've been trying to do is grow in customer facing organizations, we've been trying to keep our non-servicing headcount fairly flat and add when we need to. But again as we grow the ratio of our premium, so as we've grown and added people, it’s worked out. So it hasn't put the pressure you would think on our expense ratio or LAE.

[Ian Gutterman:]

So then just one other topic real quick is on capital, I think there was some language in the K about essentially the strain from some growth, and you mentioned maybe having to raise some debt. I guess I was hoping you could give a little more detail on that. I guess as I was -- I was just taking sort of consensus numbers just as a proxy and it looks like, if I look at where premium growth in street models that requires adding near 3 to 1, as much as a 1.5 billion capital for growth, which is greater than your dividend capacity. So when I look through that, it seems that I shouldn't expect much of a stack dividend this year. Is that the right way to think about it that essentially earnings growth is going to fund the growth and this have to be all the kind of small and the debt will pay the corporate dividend, the debt raise?

[Tricia Griffith:]

Here's how I would think of it, and John you can weigh in. Earnings are material source of our capital and when that's not enough than you can expect that we'd go to the capital markets for more. So we do have a lot of earnings coming in, but that's how we think of it. And literally John and I, and Bill Cody from our Capital Management, and Katherine Brennan our treasurer, we talk about capital all the time in terms of how we can continue our growth knowing the regulatory and constraints you put on the 3 to 1 on the auto. And so, we think about that all the time, but earnings will be a material source of our growth. But again, we can expect to go to the capital markets, if in fact we needed more.

[John Sauerland:]

And to that, you mentioned our dividend, our dividend is an annual variable dividend and is tied to our gain share score, which is a function of growth and profit across our business lines, multiplied times a third of after-tax abstract underwriting income. So, we have established that dividend program for the year, and we would use any capital beyond that obviously as efficiently as possible and today that means reinvesting in the business.

[Operator:]

The next question will come from of Gary Ransom of Dowling & Partners. Mr. Ransom, your line is now open.

[Gary Ransom:]

I had a question on the announcement that came across recently about Uber and your relationship with Uber, this has expanded. I know you went into Texas a couple of years ago, but you haven’t talked about it recently and it seems like it’s an important source of additional data that can be used in commercial lines and may be there’s some overlap with what you’ve learned in Snapshot with that. And I wondered if you could just update on, how you are looking at that program? What it might mean for your data analytics in your long run? And any other thoughts you have on that program?

[Tricia Griffith:]

Thanks, Gary. Yes, so we started with Uber in Texas in 2015 when it was a pilot on the commercial side, and we’ve been learning a lot and of course this moved from pilot last year to full board and that was added on as of today Arizona, Colorado and Florida. I wouldn’t say it’s necessarily related to an UBI, I would kind of bifurcate though. But I would say that we’re learning a lot about the transportation industry and the T&C industry. And so, we are really excited to continue to work with Uber and add on three more states. Again because it’s new to us, we are going to take a measured approach. And I think we are really excited about that. But I wouldn’t necessarily correlate it to our usage-based insurance.

[Gary Ransom:]

Maybe you could talk about a little bit -- then about the usage-based. Is there anything that you’ve learned or discovered, just update of how powerful you think your Snapshot program has become?

[Tricia Griffith:]

Yes, what I would say is we are really happy about the mobile device specifically on the direct side and the take rate on the mobile device. And so, we are learning more and more and I would say we are learning more and more about distracted driving. We are not ready to use that data in ringing but we are learning a lot, so more to come on that. It’s going to take a while before you gather in and our customers still have the option to have the dongle or the mobile. But we are finding, we are gathering a lot of data on the mobile, and we are learning I would say very interesting segmentation things around distracted driving.

[Julia Hornack:]

Right, Latoya, I am actually going to take a question from the webcast now. The next question is frequency versus surprising benefit in 2017, and one that you probably didn’t plan for in fact I think we’ve said we didn’t plan for. How do you view giving back some of -- the possibility of giving back some of that critical "excess" profit to further accelerate growth?

[Tricia Griffith:]

Yes, so when we think about -- yes, one, it’s really hard to understand and predict frequency and in fact -- so over the last trailing 12, we have been -- we had a frequency much more negative than the competition. Third quarter 2017, the rest of the industry kind of changed a little bit as we saw compete fair results. So we are seeing as an industry. Again, we don’t necessarily bake that into our rate. We don’t see where we are at as of the end of 2017 or even January in excess profits. We think about it as continuing to invest in the business. And again, this is very capital intensive when you think of the regulatory capital we have to have. But in addition, I might see differently if we weren’t growing. So we have that balance of growth and profit. And so I believe when you’re in a really sweet spot and able to roll really fast and make at least $0.04, remember it’s at least $0.04, we want to do both. And if either one of those change, as you know from 2016, we will monitor that and react to that. But we're going to need capital to grow.

[Julia Hornack:]

Thank you. Latoya, if we could take the next caller from the line please?

[Operator:]

Yes, the next question comes from the line of Brian Meredith at UBS. Mr. Meredith, your line is now open.

[Brian Meredith:]

Tricia, I'm just curious, as I look at your loss ratios and loss ratios going forward. Are you seeing a benefit from the Robinsons to on your loss ratios kind of increasing the mix? And is that perhaps maybe mitigating some of the kind 10-year effect that you typically see with the growth you're putting on?

[Tricia Griffith:]

We would say that our Robinsons are considered lower pure premium, in that they're less likely to get into a loss. The hard part is we look at everything so granularly in terms of states, channels, the products, and I think you would see -- what you'll see probably more movement on, when you think about ratios, would be more on the LAE or NAER side. Do you want to add anything to that?

[John Sauerland:]

Yes, I think you start hitting on the fact that we priced all segments to a common combined ratio target. And the preferred segment, we look at both the loss side as well as the expense side, so both of those are considered in the pricing. There are some segments that drive higher expenses. We build that into the prices as well. So you will find segments for our business that have actually lower loss ratios, higher expense ratios than perhaps the Robinsons in aggregate. But the point is, is that we are looking very granularly at each segment and making sure that we're pricing to the same lifetime combined ratio target.

[Brian Meredith:]

And then my second question, I'm just curious, going back to the tax stuff, and thanks for the -- your answers there on that one. But my question is more from a regulatory perspective. How do you deal with regulatory potential pushback? Do you expect it, particularly when you make rate filings? You've obviously got very attractive margins and returns right now. Are we going to see pushback? And how do you respond to that?

[Tricia Griffith:]

Yes. I haven't -- it's too soon to tell would be the answer. I haven't really seen -- we have good relationships in the 51 jurisdictions that we're in. Really, a regulator's job is to make sure that we are not excessive, inadequate or unfairly discriminatory. And if they look at that and they -- we have actuarial justification that should really be the job of the regulator. And so again, it's too soon to tell, but we have not had any pushback at this juncture.

[Julia Hornack:]

Latoya, can we take the next caller please?

[Operator:]

The next question comes from the line of Yaron Kinar of Goldman Sachs. Mr. Kinar, your line is open.

[YaronKinar:]

I had a couple of questions. First, on the Snapshot mobile device, Tricia, it sounds like one of the aspects of the device or the app -- sorry, may give you access to is distracted driving behavior. Are there any other real significant differences between the data collected on the mobile app as opposed to the dongle?

[Tricia Griffith:]

I would say it likes it's same as the dongle, and then we're able to see if you are -- if it's handheld, a call or an app and hands-free. So we're able to see that, and not sort of using that for rating but just understanding that, as well as location. So those are the things. Again, we're not rating that. We're just gathering data to kind of understand if it correlates to different losses depending on what we see.

[YaronKinar:]

And then the other question I had was more on the agency side. So I think you had added about 1,500 new agencies in 2017. Can you maybe talk about, what the main drivers for that were? And then is there any difference in the profile of these agencies.

[Tricia Griffith:]

No, I wouldn't say so, I mean we have a very broad distribution of agents and if we believe they follow our values and they will sell our products then we will likely have them be a part of the Progressive family. So they're kind of straight across the country 35,000 so, some come and go but I wouldn't say they have anything necessary. The different agents that we talked about in the past really around are platinum agents which we give a different commission base to and access to annual policies on the auto side as well as the home and the ability to earn more depending on how many Robinsons that they get.

[Operator:]

Thank you. The next question will come from the line of Meyer Shields of KBW. Mr. Shields your line is open.

[Meyer Shields:]

John, I want to sneak in a little bit to one of your early responses where you talked about pricing to lifetime profits. Does that imply a bigger initial I'm going to call it new business penalty on the Robinsons because they're expected to stick around for a longer time?

[John Sauerland:]

Great question, it depends on the channel, it depends on the segments. Certainly in the direct channel, we have a material "new business penalty." We don't think of it as a penalty we think of that as the cost of acquiring customers that we're going to keep for a long time. And we again are pricing in the target acquisition costs that Dan was talking about across the life of that policy holders. So even though we will run well above a 100 combined ratio for a new direct customer we are going to recoup that over the life of the customer. So, the Robinsons into that spectrum certainly is generally speaking a longer retaining end of the spectrum, so we have longer to recoup those costs. We've normally then would be able to have a smaller load if you will on each policy term, but again the new business combined ratio for direct is going to be far higher, in agency channel there's far less disparity across expenses for new and renewal customers and in the Robinson the spectrum is actually less disparity on the lost performance as well. Relative to sort of the same end of the spectrum so there's certainly a new business penalty for growing we think the lifetime combines for all those customers are well within our targets so we're very excited to be growing at the pace we are and delivering the calendar year combined ratio results we have as well.

[Meyer Shields:]

Okay and I guess, that's very helpful, second question is. Are there any calendar year constraints that you're enduring right now?

[John Sauerland:]

Right now calendar, our calendar year constraint is the same it’s always been, it's 96 calendar year growth fast as we can as long as we can service our customers. So when we were in a similar position a couple years ago with being able to get out in front of hiring and make sure that we could service our customers and we we're had this competitive price and a great product our 8 to 4 product on the street. We really wanted to make sure to capture as much of the market as we could. Of course, you remember in the fall of 2016, we pulled back on advertising because of the CAT losses and that's always our constraint. Every calendar year, our constraints are going to be 96 that is a constant its part of our culture, it's not a sale for, and so that will obviously be a constraint. Clearly, right now, one month in its results. We are not seeing that as a constraint. But again we’ve got one month of results in. It’s a long year. We don’t know it’s going to happen with another nature. I wouldn’t have thought the 2017 would have been a more difficult CAT year than 2016, but of course it was. So that’s really our biggest constraint is our profit goal and making sure we can take care of our customers. And we have been very, very happy. I wrote in my letter of -- what I call our recruiting machine. Not just the amount of people we have been able to hire but the caliber of people. When I am ended doing hire classes, I am amazed. And one of the ways we look at that especially in the claims organization is how their accuracy is going. So that’s a large part of what we pay out in loss cause and our accuracy continues to be at record level. So that’s really what excites us about having few constraints right now. It’s a good time here.

[Operator:]

The next question comes from the line of Mike Zaremski of Credit Suisse. Mr. Zaremski, your line is open.

[Michael Zaremski:]

I had a follow-up to the earlier ride-share question but along different lines. So specifically what’s in the personal auto segment, are you able to comment on whether you feel ride-share services are impacting the loss cause? And I guess one of the reasons I’d ask is because an increasing number of insured drivers are making ride-share services a part of their lifestyle, and I’m also pretty sure your client base is relatively more tech savvy than many peers as well?

[Tricia Griffith:]

Yes, we have a personal auto endorsement in 22 of our states. So we always offer them and ask when we are getting navigation as well as a loss if you’re an Uber or Lyft driver. And so we do strive to flush out, and it helps to understand if we need to charge them differently, add an endorsement that is covered. It’s really hard to understand fraud that could happen to people on it, it’s actually why we are offering endorsement. And so I think we believe you’re asking the right questions to the right people at the right time.

[John Sauerland:]

It’s very hard I think to tease out what you are getting at. We do generally younger clientele than some of our competitors certainly on the direct side of business. We think we skew more towards urban areas where we would expect there would be higher penetration of T&C users. But it’s very hard to definitively understand, if and to what degree people using Uber, Lyft et cetera at different times of the evening or weekend are helping a loss experience at all. But it certainly is a great hypothesis, but one that we really can’t answer with any surety.

[Operator:]

The next question will come from the line of Elyse Greenspan of Wells Fargo. Ms. Greenspan, your line is now open.

[Elyse Greenspan:]

My first question was related on, in your Tricia you mentioned an internal strategy council that you guys are forming on. I just was hoping to get some additional color there. Is this in response to autonomous vehicle? Is there something else in the industry? And how would that potentially include Progressive potentially expanding and to underwriting its small commercial policies on to our own balance sheet or potentially include additional M&A?

[Tricia Griffith:]

Okay, so if you recall a couple of webcast ago, I went over sort of our Horizon concepts that we have Horizon 1, which we talked a lot about today, execute Horizon 2 which was Expand and also we talked a lot about the commercial that you talked about a little bit Elyse with Bob in small business. What I am having the strategy council really focused on is Horizon 2, which is explored. So you can think of it as assumed there's a day where all these vehicles are autonomous or half of that et cetera, so we're modeling out opportunities for us. And so right now, the Strategy Council is fairly new. John and I meet with them very regularly. It's a group of 10 folks within the Company. Andrew Quigg, who I think you have met, is doing it as a side to his job of retention and PLE. I mean, we have other people taking some of his work. He's doing a tremendous job. And the team around him, we picked because they had a diversity of a careers for Progressive, and so we're able to leverage those careers to kind of think about what are things that we do really well and if we want to either grow our revenue or replace revenue; if the market gets smaller, what are opportunities we can do. So we're excited about this council. Again, we're pretty new there, 90 days in. So I'll have more to come as we flesh out more. But we're really excited, and I've been very impressed with their work today.

[Elyse Greenspan:]

Okay. And my second question, you guys have been growing a lot over the past couple of years. As some of your peers have been retrenching, taking a lot more rate as they've seen pretty high frequency and, in some cases, severity trends. As we think about the overall, it seems like most players in the space are more or less at peak rate taking levels will take less rate and look to grow in '18. As you envision that environment, how does that play into how you think about continuing to go after the Robinson cohort and growing your policies in force?

[Tricia Griffith:]

Yes, I mean, so we've always talked about rate in terms of not having to rate shock our customers. When we've had to do that, you see PLE decline. So we've got out ahead of rates, probably 2015. And we've often said, it's better to have smaller bites of the apple. So 3 1 is better than 1 3 in terms of increase, and so we're really comfortable with our rate right now. And we have watched our competition, and that's what happens. And that's why shopping occurs. This hard market, we've been privileged to get those customers while they're shopping. And because we have a competitive rate, we've been able to grow. So at this juncture, I don't know, obviously, what all the competition is doing. Very few companies when we look at statutory data, actually, I think as of Q3, only 2 companies were growing profitably. And so I imagine those companies are going to want to get out in front of rate. And hopefully, as people shop, we will continue to be the recipient of those customers, Robinsons, of course, but actually, every customer. We want every Sam, Diane Wright, Robinson we can get as long as we can service them. We have obviously talked about the Robinsons a lot, because that's really the customer that we want based on our acquisition of ASI on the agency channel and working with other affiliated – unaffiliated companies on the direct side. So we'll continue to push those. But I think Cat talked a little bit about our changing creative. And really when you think about the Parentamorphosis and owning your first home, that's really addressed to the Robinsons. And so we are seeing -- we're seeing that happen. So we're going to continue with more creative around trying to retain more Robinsons.

[Julia Hornack:]

We actually only have one more caller in the queue, which is perfect given our time right now. So, Latoya, can you introduce last caller please?

[Operator:]

Yes. The next question comes from the line of Jeff Schmitt at William Blair. Mr. Schmitt, your line is now open.

[Jeff Schmitt:]

Just a quick question on some of the comments from state insurance commissioners, recently about passing on the benefits of the new tax rate or the lower tax rate on to consumers; do you have any view on the likelihood of that or a sense on what that impact may be?

[Tricia Griffith:]

It's hard to say -- it's hard to say, there's a few states out there that will likely do that and we'll react to that. Remember, our combined ratio goal is pre-tax, so our CR496 vehicle is pre-tax. So we will work with the department to make sure we have a product that the citizens of the states. Again, our hope is that once regulators regulate in terms of what I talked about in terms of not being -- not having our rates inadequate, excessive or unfairly discriminatory. But again we're nimble and we'll work with it to make sure that we provide a good rate for the customers and not change our stands on our 96 combined ration goal.

[Jeff Schmitt:]

Okay, and then I'm not sure if you mentioned it, but did you say what the growth of the platinum agents was in '17 and what percentage of total agents is that now?

[Tricia Griffith:]

It sells us small percent, we didn't say, it grew must say like 2000ish.

[John Sauerland:]

In terms of agents it agent headcounts, that's a reasonable number, Yes.

[Tricia Griffith:]

Yes, it's around 2000-2500 somewhere in there. We sort of stopped mentioning it just because we're really focusing on the agents that we have and making sure they have all the tools necessary to get those Robinsons. Again when we rolled out the Robinson model, we really rolled it out as a scarcity model because we wanted to be able to have these agents really wanted to have access to these customers. So they access that was great and then did they want to have us be number one in their shops. So we're really working now, especially as we integrated the company's, really working on making sure they had everything they need to have us be the number one choice in their agencies. So it's less about adding so many more, I'm sure we'll add more this year but it is really more about the ones that we have really making sure that they have what they need to provide more Robinsons for us.

[Julia Hornack:]

Well, we've actually exhausted our scheduled time, so that concludes the event.