Carvana (CVNA) Presents at Deutsche Bank Technology Conference (Transcript)

Carvana Co. (NYSE:CVNA) Deutsche Bank Technology Conference Call September 12, 2017 4:20 PM ET

Executives

Ernie Garcia - President, CEO and Chairman

Analysts

Mike Levin - Deutsche Bank

Mike Levin

Good afternoon everybody, I would like to welcome Carvana, which is an online retailer of used vehicles. While used auto retailers a traditional business, few companies have attempted to sell 100% online and virtually none have succeeded. This is very difficult problem to solve requiring know how in inventory acquisition, reconditioning, finance and particularly logistics. We believe that Carvana has actually shown the deep expertise in all of these areas and has a compelling offering that is resonating with consumers.

To learn more about the company, we have with us Ernie Garcia, the CEO, and Mark Jenkins the CFO is also in attendance.

I will now turn it over to Ernie to go through a few slides.

Ernie Garcia

Thank you. So I think the plan here is to go through a quick presentation to get an overview of the company and then we will jump into Q&A, so I will try and go as fast as I can.

First of all this presentation will include forward-looking statements, Safe Harbor, we undertake no obligation to update those, please forgive us for having to spend a slide on that.

So really quick high level I think Mike set it up, but our mission is to change the way people buy cars. The general idea behind Carvana is that if we can give people a very simple way to transact online if we can reduce costs and by reducing costs we can give customers lower prices and better experiences. And then we have to build an entire operational chain to support that we will talk about all that as we go through the presentation.

Really quick kind of context, the industry is really, really large, it’s a -- I stand here just directional it’s a $1.1 trillion industry in total, used automotive alone is $700 billion, it’s enormously fragmented, which has a lot of implications for the competitive environment. The largest player has a 1.6% market share, the large 100 players combined have a 7% market share and then 81% of consumers say they do not enjoy the car buying process. So we think that that’s a really good backdrop for innovation and disruption.

Our general kind of strategic lens is that we want to give customers the best experience possible in buying a car, we want to give them the biggest selection process we can and then we want to give them the best value and everything that we do aims for those three attributes. So best experience, customers can buy a car in as little as 10 minutes going through our process that includes getting approved for and selecting financing, getting a trade in value applying a third financing signing a contract online and then scheduling either delivery or pickup. We deliver the car as soon as the next day and then we give the seven day return policy.

Best selection, we will spend more time on this later, but our model because we hold all of our inventory in several locations and then we connected to a logistics network to every market where customers are. All of our inventories available to customers anywhere, which is significantly different dynamic than traditional dealerships that generally have a frontline about a 100 cars and basis have a local inventories servicing local demand.

We've got a global inventory servicing global demand and that’s significantly number of ways, but the most obvious is selection, the next is best value. Dealership spend a lot of money on a lot of things, but the biggest areas of costs are one basically labor per transaction which takes the form of sales people, salesman, general managers, F&I managers all the other employees of the dealership. Two is the big dealership itself and then three the customers acquisition cost.

Our view is by moving everything to a self-service platform it can reduce labor by moving fulfillments to logistics and then to a much kind of lower cost vending machine versus dealership. We can reduce kind of physical overhead and then by providing a differentiated offering we can ultimately also reduce customers acquisition cost that’s how we think about providing best value. And then we pass those SG&A savings on to customers in the form of lower prices which reduces GPU always constant, gross profit.

Okay, this is slide, I think in many ways is our report card, the whole in the market in our view is giving customers a very simple experience when buying a car. And so we like to look at how customers are rating us, we now have over 10,000 reviews up online, our average ratings is 4.8 starts, 95% of customers said they would recommend us to a friend. We think that this is really essentially important it’s a number that we watch very carefully and somewhat interestingly we think that this metric actually matters more of automotive retail than many other businesses. Not just because customers bring relatively low expectations in, but also because customers are buying a car once every five years.

And so the experience that you give a customer gives them a certain quality of experience they are going to tell their friends about, but then the amount of time it takes for that kind of referral to show up in incremental sales is much longer when the transaction is like was five years then we have a shorter transaction cycle.

So, we think there is kind of more latency and momentum in a referral curve because of kind of that lag time than there would be in many other products. And so we think this is also more indicative of the future than it might be in other businesses where you instantly see those increases in sales. And so we already are showing up in sales when you get high quality reviews.

I think this slide is largely kind of the most important slide in the whole deck. If you look at the graph on the right that basically shows market share in each of our markets, the X axis this quarter since inception, the Y axis is total market share. The important things to note here overall this as of Q4 of 2016 Atlanta and Nashville were both about 1.1% market share. That doesn’t sound like a lot, but when you compare to the largest player having 1.6% market share nationwide, that's actually very significant, those are meaningful market shares. But you can see they’re still growing very quickly in the last 12 months ended December 31, 2016.

Atlanta grew about 65%, Nashville grew by 156%. So, even those individual underlying curves the most mature curves are still growing very, very quickly. And then if you look at all the other curves for the most part, they’re ramping faster and they are following very similar trends. Worth noting we lunched about quarter seven for Nashville, we launched the vending marching, you can see that kind of the curve separated there clearly led to an increase in sales and we've seen consistent response in vending machine launches in other market as well.

And so we think that's a really good investment we’ll probably talk about that a bit more. What is it take for us to open a market if you look up at the top left, there is basically four steps to opening a market for us, it's much, much lower cost and much kind of capital later than the traditional automotive business model.

All we need to do is first we connect with logistics network that means we spend about $250,000 on a 10 car hauler. Those are the big trucks you see on the freeway moving cars around. We connect to that from the nearest node in our existing logistics network up to the new market that we want to open. And every day it runs back and forth that effectively connects our 7,000 car inventory to customers in that new market.

And then they can see, when they go to the website, they can see that if I buy this car, buy today at 10 it can be deliver by tomorrow at noon. And that was the first process that truck just runs back and forth every day. We have a team of expansion advocates. All of our customer facing employees we call advocates, there is basically two forms of that, one are those that answer the phone are at home office in Phoenix and then one are the employees that are out in the field that handle last mile delivery to customers door.

Those are the employees that we need to hire to open a new market. We have a team of expansion advocates that travel from market-to-market go and hire those employees and train them up, and then we open a new market, they move on to the next one. We've to turn on marketing, our biggest expense by a pretty significant amount in any given market is basically is turning on local marketing channels.

So, we do have home office, it scales very well, but that is to these variable expense in markets. And then we buy several single car haulers that cost about $75,000 a piece and we build out a little office space to house the employees that are handling last mile delivery. So, because of that it's very, very easy for us to open up market. So it doesn't require a ton of people, it doesn't require ton of real estate, and it doesn't require a ton of capital. So we have been accelerating our market openings, in 2015 we opened six, in ‘16 we opened 12 and then in the first half of the year we opened nine markets in 2017.

This slide gives a sense of our logistics network, so all of those blue dots are locations where we are currently selling cars, the stars are our inspection centers. So we currently have one just outside of Philadelphia, one just outside of Atlanta and then one just outside of Dallas. These are generally 40 acre ballpark facilities that have room to park thousands of cars, hundreds of lift technicians that are repairing those cars and getting them ready for the customer. We also photograph at that location and then put them on the website and they are transported away from those inspection centers after a customer purchase them.

All of those other dots are connected by our logistics network we don’t have that illustrated here it would made the graph a little difficult to follow, but that’s how that all works. And then we have announced that we are working on our fourth reconditioning center out in Phoenix, we recently opened in Phoenix market a couple of months ago and then I believe a week and a half ago or so we opened up LA. So it was a big expansion for us to connect the logistics network out to the West Coast.

We have been growing very quickly, so retail units sold grew 145% year-over-year in the second quarter, revenue grew by 142%. So very rapid growth, it’s been fairly consistent growth and we feel like it’s relatively straight forward growth to forecast because it’s effectively just layering all those different markets curves on top of each other. We opened markets and then each of the markets mature and the underlying markets themselves are also growing at very high rates of speed.

So our growth is not driven just by new markets again to reiterate Atlanta in the 12 months ended December 31, 2016 grew by 65% and I believe that was our slowest market growth of all the underlying curve. So, lots of growth.

Our total gross profit per unit is also growing very rapidly, we lost $200 per unit sold in 2014, we made $215 in 2016, we got that up to about $1,000 and then as of Q2 2017 we are up to about $1,500. So we are making a lot of progress there and expect to continue to do so.

We think there is a lot of room for growth I won’t labor this too much, but we can open new markets. We can have those markets continue to catch up to existing markets. We built out a lot of technology that separates us from just about anyone else out there as it relates to the customer experience on the website supporting logistics, valuing trade inside unseen, there is still a lot of technology that we continue to build out and we will continue to do that and we think that there is room there.

Improving brand awareness, I think it’s very early days for us getting our name out into the broader world, we have just started to test national marketing over the last quarter, we will continue to do that overtime, national marketing has been very successful on automotive retail for several large kind of listing sites and branding sites. And then also for some of the largest automotive retail brands.

I also think something that’s interesting here is, we like to use this metaphor of online dating, I think the online dating 10 years ago was very, very strange, I think buying a car online today to most customers still feels very strange, but once they’re friend or neighbor or family member has bought a car from us saved $1,500 got the car first in 15 minutes, had a huge selection, had it delivered to them got a seven day return, we think there is a lot of opportunity to penetrate deeper in to the population and we are so far as that behavior normalizes.

That’s the presentation. So with that, I will sit down and we can start the questions.

Question-and-Answer Session

Q - Mike Levin

Great, maybe just starting with kind of the core principles you are talking about in terms of best selection, best value and best consumer experience. So just starting with value, today you sell for about $1,400, $1,500 below Kelley Blue Book retail. I think we think that comes down overtime, is the idea that that kind of moves down in the next couple of years with the strength of your brand. And then you maintain some sort of discount in the long run, is that the right way to think about that?

Ernie Garcia

Yes, so I think I would start with, we think our cost structure is different and we think that the pricing that we are offering customers today is basically taking our SG&A savings and passing it on to customers in the form of lower prices. And we think that’s a really powerful lever.

Now what are we going to do overtime, I think there have been a number of very well-known automotive retail brands that have started as big discounters and overtime migrated up in price and kind of down in discount. I think that opportunity is likely to exist for us, we pay a lot of attention to price elasticities both on our site and with our own data and extrapolating from some others out there and we don’t feel like there is any price split at any point in time.

That said, we think the opportunity is really, really large, we're drilling very, very quickly. We said that in Atlanta in March we sold 1,000 cars we're now in 37 markets. It's just simple math there right would get you to a company that's 10 times as larger as we are today. So I think we also want to be very thoughtful about how we roll those things out and make sure that we're giving customers a great experience and a clear value prop that they understand and we think that the opportunities from there are very, very large.

Mike Levin

And then on the consumer experience front, obviously, at a very high level you guys have done really great, which is if anybody hasn't bought car recently is really saying something on auto retail. But when we just kind of like browse through the bad reviews, it seems like a lot of things have to do with either unmet or improperly kind of set expectations and like you said kind of online dating it seems kind of like an awkward stage. I mean what are some of the pieces of the puzzle of kind of getting that right and communicating that and improving the consumer experience to making it feel less awkward or is it just a matter of time?

Ernie Garcia

First of all I wish you’d raise the good ones.

Mike Levin

I did that too.

Ernie Garcia

I mean, I think we're not perfect, and I think a big part of our brand is trying to be transparent across the entirety of transaction and that includes putting our once our reviews up on the websites so that customer can go and see like what the risks are. Now across all of those reviews average rating is 4.7 to 4.8 stars. So on average it's very, very good.

I do think if we miss customer expectations I think whether it's an online transaction or an in person transaction generally kind of the car not meeting the customers' expectation is the most likely culprit there. And we try to once that and continually get better. But like I said it's important to us the brand put that out there we think it's very valuable for customers to see that we're not perfect and we're not perfect we learn from those mistakes we try to fix it. And we're very, very happy with the overall response we got from customers.

Mike Levin

Great. And kind of one of the most essential metrics to ensuring this business really works is your days to sale. And right now you're running about 90 days give or take we think that breaks down to about 30 days between auction and listing and about 60 days from listing to sale. Can you kind of give some of the smaller pieces of what that process breaks down as and then how you get that down to at least 45 days and then over what time period that kind happens.

Ernie Garcia

Yes, so I mean at the highest level let’s talk about how it kind of get pushed down first and we can kind of break into steps. I mean one thing that’s very nice about our kind of pooled inventory model is we had to take demand from all the different markets and push it on a single pool of inventory and that's really, really different. So if you start with kind of the simplest model you can, if we sell a 1,000 cars in a month in Atlanta and then you open up another Atlanta and then another Atlanta, another Atlanta. Those customers in Atlanta are facing 7,000 cars in selection today. It stands a reason that if you open up equivalent markets you can get another 1,000 sales of the exact same cars.

And so I think being able to take all those different demand pockets and push them onto a single inventory pool is a powerful attribute of the model that we think is fairly straightforward to understand and overtime will cause us to push turn time down. Given that it make sense to invest in a lot of inventory upfront if you feel very confident that you can push your turn times down it makes a lot of sense to invest in inventory because if you are spending money to drive customers to the website their likelihood to purchase is largely a function of how many cars you have and whether or not you're able to meet their goals based on what they're looking for.

So we invest in a lot of inventory upfront historically going from the company’s kind of inception today we've largely grown inventory with sales and just grown a bigger and bigger selection starting in Q4 of 2016 we basically stopped growing inventory we've grown sales by 91% since then with the flat inventory. So I think it's reflective of kind of the underlying fundamentals there. And we feel good about our ability to continue to push that down.

And then because car is a depreciating asset without changing the customer offering at all, that kind of mechanically leads to higher margins. And so we think that's a really important attribute of the model. Now to your point about stages, I think there is kind of four stages that are worth noting in a car's life. Stage one is sort of when you purchase it and then you transport it from wherever you purchase it to the reconditioning center. The way that we buy cars we take into account how much it's going to cost, and how long it's going to take to ship a car from where we brought it from to the reconditioning center.

So there is time that elapses there and there is also an expense to ship it. But those two things are taking into account in our bid. And so in effect they’re sort of almost free it’s like the cost that those the form of that cost takes is our likely of winning cars that are further away just decreases because we're bidding lower amounts. So in a way that doesn’t necessarily directly impact kind of the margins.

The next step is the time to the reconditioning center, that’s purely the faster you can move the better you are because the car is depreciating and you need to get it up to retail standards and that’s an important step that we think we are very good at and we continually get better at all the time.

The next step is when it’s up on the website and a customer is looking to purchase it, that’s the step that most mechanically reduces by just putting more eyeballs on the website and that’s by far in a way the longest step in this entire chain. And so we want to continue put more eyeballs on the website and push that down.

And then the last step is once any given customer locks that car and to purchase it at that point the car stops depreciating. So they may shop for a couple of days in our purchase process some customers buy a car in 10 minutes and some take a week and half, but that’s not -- we’re not really bearing the cost of that, but does show up in our turn time. So I think those are the different steps in the purchase process.

Mike Levin

Got you. And I mean one of the good -- I mean you alluded to it, one of the good charts you showed in the Q2 letter was basically that you have been growing sales at a pretty high clip without increasing inventory and clearly that should kind of continue as your days to sale is coming down, but once you start getting to your target rates for days to sale, how should we think about inventory growth moving forward, should it still be kind of sub-linear to your market unit growth?

Ernie Garcia

So great question and I am going to comment here it to make a couple points. So the way that we think about kind of our inventory strategy that has three phases and the phases are not discrete or perfect, but I think they are important to note. So Phase 1 was basically the inventory build phase where we are growing inventory with sales linearly. In that phase you have feedback as you kind of grow sales you are able to just buy more inventory as you carry more inventory you increase sales in all of your markets and so there is this nice positive feedback.

Phase 2 is when you say okay, now what I want to do is I want to hold inventory flat, take all those incremental eyeballs, push turn time down, push GPU up that’s going to create better cash flows for the business. But that’s kind of a transitory phase. The problem with that phase is you stop growing inventory. And so you give up this tailwind and you actually also take on it a saddle headwind, because you have given that the number of cars that are washed out in a purchase process are proportional to the number of cars that you are selling. If you are holding total inventory flat, you are actually slightly decreasing the number of cars that customers can purchase.

And so in this phase that we have been in for last six months, we are facing a sales headwind, we have kind of given up the network effects and kind of the positive feedback loop, yet we grew sales by 91% in that two quarter period. So there is clearly a lot of underlying growth.

Phase 3, I think to your point is once you get to a turn time that you like and a GPU that you like now you start just growing roughly speaking there is many ways to do it, but roughly speaking you grow inventory with sales and now you just get the feedbacks like we’re going to get. And so I think from a growth perspective we are in the hardest, the Phase 1 and Phase 3 are the easiest from a growth perspective, Phase 2 is the hardest growth phase and we are in the middle of that right now and we are growing very quickly through it. So I think we are very excited about that.

Mike Levin

And then I mean kind of with that right now you are buying about 90% of your inventory at auction, other retailers are self-sufficient to the tune of 40 to upwards of 75 at a franchise dealer. Where do you guys see that kind of getting to in your long-term plan how much more profitable are those versus an auction unit just kind of your thoughts on how your strategy on sourcing move forward?

Ernie Garcia

So I think it’s a huge opportunity and we plan to invest more in it and continue to make progress there. I think it’s important to put that in context. I think if you talk to many people in automotive retail you ask them what matters, what’s the most important thing, how do you differentiate yourself and how you kind of win this game. 99 out of 100 dealers will tell you it’s all about vehicle acquisition right, if source cars right, you make your money on the purchase and then you selling the customers that’s more commoditized and it’s harder to make a huge difference in the actual transaction.

I think that’s the answer you will get from anyone and I think sometimes we get pushed back, it says okay well that’s true why you are buying your car from auction, right. I think the reason that many retailers believe that in my view and this is speculation to some degree, but my view would be the retail businesses they run are very, very undifferentiated and because of that the area of greatest possible differentiation is vehicle acquisition.

I think our business model is all about differentiating kind of the retail pipe it’s about giving customers a lower costs simpler experience in the retail pipe and we think that’s the most important. So we want to focus on that first and that’s we are differentiating today. I think as we built that brand out clearly there is opportunity to go buy more cars directly from consumers and many big brands have been very successful at doing that. We absolutely intent to do that, it’s a matter of priorities and I do think that that’s actually simpler to do than redoing the entire kind of retail pipe. But that's an opportunity that we'll look to expand overtime.

Mike Levin

Okay. And so currently we're about $1,500 in total gross profit per unit about half of that is on the retail side, and you’ve talk about getting to about 3,000 sort of a midterm goal. What are kind of the components of that bridge from here, what are the biggest pieces what are we kind of looking out for the near-term and over the next couple of years?

Ernie Garcia

Yes, so, I like to break that into two major buckets, so one bucket from $1,500 and $3,000 is basically scale driven and then one is product driven. On the scale driven side, the biggest single driver is basically just increasing the turn of vehicles and therefore decreasing depreciation, increasing GPU that's definitely the biggest single driver in the entire walk, but also the biggest driver in kind of the slightly larger than half of the walk they goes towards scale.

The other two major steps in scale are basically pushing more throughput to the logistics network and then through the reconditioning centers. Those are also fairly mechanical if you use trucks that are already running back and forth more that lowers cost. If you use inspection centers that are already built more and you run double shifts et cetera that lowers cost. So, those largely will happen just in kind of scale as we grow sales those will show up.

There is some seasonality in kind of depreciation rates that will cause that to move up and down seasonally, but in general we expect progress there to be kind of very steady and it’s been very steady for the last several quarters. The other bucket is basically product enhancement so how do we monetize finance better, how do we monetize warranty better, how do we built out more buy from public and be able to purchase more cars at a greater margin. We announced in Q2 that we are starting to test gap insurance, we'll continue roll that out over time. So, that bucket is driven by kind of product enhancement, which should be chunkier overtime.

Mike Levin

So, I mean are those product and so maybe just backing out the step, you guys kind of midterm target of about $3,000 GPU putting that against CarMax your nearest comparison is about $4,000 and the biggest difference there is really just your selling price and how you are structuring getting to about the same EBIT per car. But your long-term EBITDA margin targets are about 7 to 11.5. CarMax is at about 8. So, what are the kind of the biggest pieces that you guys see as potential for getting beyond CarMax?

Ernie Garcia

Yes. So, I think at a very high level maybe easy way to answer this question is that we believe on a variable expense basis we can save about $1,500 per transaction versus more traditional retailers give or take. So we think that's kind of the savings in SG&A then the question is okay what happens up in GPU and to your point we're shooting for a $1,000 back of CarMax. So, if you’re $1,000 back in GPU, but you’ve got $1,500 lower SG&A there is extra room there.

I think there is a question of okay what’s like the of all the components of GPU, where do you differ from someone like CarMax and I think price is by far in a way the major difference and I think there are some other differences that are largely offsetting. They do very, very well on the wholesale side of the business we do pretty well on the finance side of the business and understand that pretty well.

So I think there is a couple of other components that are offsetting but high level I think if we pass the SG&A saving on to customers in the form of lower prices, we should have similar bottom-line and if we don't pass all of that we should have better bottom-line.

Mike Levin

Got it, okay. And maybe just before we kind of open it up, maybe just talk a little bit about how you see the competitive landscape here a lot of people kind of think cars.com and these software providers or lifting aggregators or competition, we kind of view them more as a platform for market. So just traditional players the kind of emerging players that you see as direct competition and maybe the looming 800 pound gorillas Amazon has begun efforts in Europe and is probably eventually coming back to the U.S.

Ernie Garcia

So, here is how I think we like to think about competition. I think one way to think about competition is that someone else creates a customer offering for any given customer set of preferences that's better than yours and therefore consumer start to choose something different. I think that's kind of the standard structure of thinking about competition. In alternative to that is what happens to your own product offering.

And to be candid I am more concerned in this market with us to make and sure that as we scale we keep our same high levels customer product offering. I think if we do that, like I think it is very, very hard for anyone to scale fast enough to dent us in a market this size. Just given how fragment it is given all of the difficulties in scaling there are significant operational hurdle to scaling. I think it’s really hard for anyone to dent us.

So to me when we have companies about competition, I more think about it as how is your offering compared to everyone else’s. And I think it’s just important that we stay very focused on as we scale very quickly we keep giving customers the same quality offering we have in the past. And I think if we do that, it’s just in our control. I think this is an enormous market where the largest player has a 1.6% market share there are just aren't that many players that really have the resources to be able to invest in a fixed cost structure like we've been able to do.

There are several big players out there that I think could enter the market, but it's $1 trillion market. I just think that this is not the type of market where a competitor or two can come and really change the dynamics very quickly.

Mike Levin

Why don't we open it up if anybody else questions in the audience.

Ernie Garcia

Come on someone ask something.

Unidentified Analyst

How do you position yourself given the dislocation in Florida and Texas to respond to that given the flexibility of the cost structure and infrastructure?

Ernie Garcia

Yes so we’re thinking about the hurricanes in sort of three phases. So Phase 1 was the storm phase which was obviously a huge event for many people. It dislocated a lot of lives and put people in a tough spot for a while and that clearly disrupted car sales significantly. I think Phase 2 is sort of replacement cycle, which traditionally has taken on the order of kind of four to six months as people go and replace cars. I think in the case of Houston I have seen scrap adjustments between 500,000 or 1 million cars will need to be replaced give or take. Very high level rough estimate, there is around kind of 80,000 used cars sold per month in Houston.

So that gives you a sense of kind of the magnitude of replacement that will happen relative to an average month sales. I think that could be significant it's very early to know exactly how that's going to unfold. Most importantly I think to us given where we are in our growth cycle Phase 3 is going to be what's the brand that we help to build ourselves by treating customers really fairly in Phase 2 whether they're replacing their cars, because that's going to be over four to six months and then it's going to be back to kind of normal life again. And did we do a good job giving customers a great offering in this time when many people were buying cars and talking about replacing their cars. We think that that's really important.

And so we're really trying to put our best foot forward to try to make our inventory available quickly and easily. We're giving customers in these affected markets discounts to kind of get their life back so they can buy a car. We're trying to make sure that we treat customers well in this period. So we think that it's an important time to make sure that we're building our brand.

Mike Levin

What percent of your transactions come directly to your website as appose to a third party? And does that number need to come down dramatically for the long-term mode around your business to be proven true?

Ernie Garcia

So, numbers that we've said in the past. We spent about half of our marketing dollars on TV and then the other half is made up of many direct channels and then many sort of brand channels. The majority of our customers are organic. The majority of our traffic is organic. And we think that that's really important, because we want to build a brand. We also see interesting dynamics where in a market where you build the brands direct channels tend to get more efficient.

So we will tend to invest more in direct channels in markets where we already have a brand versus in markets where we don't have a brands. But I think by far and away the most important thing that we can do is continue to give customers very high quality experiences over and over again maintain our differentiated offering and build brand. And I think that's what really going to matter to the business over the next several years.

Mike Levin

Maybe just one more. So clearly you guys have already shown a lot of operating leverage just as you've been growing over the last couple of quarters. And one thing we're kind of thinking about for the long-term is what is the ultimate kind of customer acquisition cost. I mean right now it's pretty easy getting up that penetration curve with normal TV advertising spending and some other channels. But as things become a little bit more mature, your brand becomes bigger. Just how are you thinking about what the customer acquisition cost kind of look like overtime, because if that increase as competition increases the operating leverage kind of looks a little different.

Ernie Garcia

Yes, so I think first of all to answer that is just by looking at kind of the data that was in our S1. So we put out kind of customer acquisition cost by market in our S1. And you can kind a see that it's clearly falling very quickly in any given market as the ages I believe that Atlanta in Q4 was in kind of the mid-500 give or take something like that and that market was still growing very quickly and marketing budgets were relatively flat and that market continue to grow since. So, we're making lot of progress there across the whole company was about $1,100 last quarter give or take a little bit, but that's basically made up of some market that are much lower and then new market that are much higher.

I think given that the company is a little less than five years old and transaction cycle for automotive retail is about five years. We don't -- we haven’t really made into the lead of repeat customers yet, we're still normalizing behavior, we're still building brand, our awareness levels in these markets are still very low. I feel very good about where those customer acquisition cost are going to go overtime.

And I think they’re already to levels where given kind of the magnitude of things we're talking about $1,500 of SG&A savings and $1400 of lower prices. The gap between us and the best in customer acquisition cost is already small enough to where I don't think that that's going to be a huge part of the story. I think the numbers are big in new markets where we are ramping up, but we already can kind of see where it’s headed.

And then as it relates to a couple of competitors coming in and driving up customer acquisition cost. I would answer that the same way that answer the broader competition question, I just think it’s hard to come in and impact things enough in a market of this size and if two or three competitors are fighting that hard, I think those two or three competitors are very, very excited.

Mike Levin

Got it. So I mean this seems to lend itself very well to say Super Bowl ad is that something we have to look forward to or might that drive too much traffic?

Ernie Garcia

We are constantly testing many, many marketing channels as I said half is TV and half is every other channel you can virtually think of. We've tested local Super Bowl ads in the past we've a good sense of how those work. I could imagine that unfolding overtime; we take a very, very quantitative approach to those things. And your point is right, when you air Super Bowl ad you spend a lot of money, you need to think about it like a branding event because kind of the instantaneous traffic burst that you get is just so significant that it’s hard to handle we'll be thoughtful about that in all of the marketing channels overtime.

Mike Levin

That’s our time for the moment. Thank you.

Ernie Garcia

Thank you.