Carvana Co (CVNA) 2019 Q1 法說會逐字稿

Good afternoon, and welcome to the Carvana First Quarter 2019 Earnings Conference Call.

(Operator Instructions) Please note, this event is being recorded.

I would now like to turn the conference over to Mike Levin, Vice President of Investor Relations.

Please go ahead.

Thank you, Andrea.

Good afternoon, ladies and gentlemen.

Thank you for joining us on Carvana's first quarter 2019 earnings conference call.

Please note that this call will be simultaneously webcast on the Investor Relations section of the company's corporate website at investors.carvana.com.

The first quarter shareholder letter is also posted on the IR website.

In addition, following the completion of our first auto loan securitization in Q1, we posted a detailed 101 on the transaction under resources investor materials on the IR website.

Joining me on the call today are Ernie Garcia, Chief Executive Officer; and Mark Jenkins, Chief Financial Officer.

Before we start, I would like to remind you that the following discussion contains forward-looking statements within the meaning of the federal securities laws, including, but not limited to, Carvana's market opportunities and future financial results that involve risks and uncertainties that may cause actual results to differ materially from those discussed here.

A detailed discussion of the material factors that cause actual results to differ from forward-looking statements can be found in the Risk Factors section of Carvana's most recent Form 10-K and Form 10-Q.

The forward-looking statements and risks in this conference call are based on current expectations as of today and Carvana assumes no obligation to update or revise them whether as a result of new developments or otherwise.

Our commentary today will include non-GAAP financial measures, including, but not limited to, ex-Gift measures, that exclude the impact of the 100,000 Milestone Gift to our employees.

Reconciliations between GAAP and non-GAAP metrics for our reported results can be found in our shareholder letter issued today, a copy of which can be found on our Investor Relations website.

Please note that all gross profit, SG&A and EBITDA metrics mentioned by us on the call today are on an ex-Gift basis.

And now with that said, I'd like to turn the call over to Ernie Garcia.

Ernie?

Thank you, Mike, and thanks, everyone, for joining our call.

The first quarter was a spectacular quarter for us in just about every respect and was full of milestones on our path to fulfilling our mission of changing the way people buy cars.

In the quarter, we opened our hundredth market and sold our 200,000th car.

It took us 22 quarters to sell our first hundred thousand cars and open our first 50 markets and only 3 quarters to double those numbers.

Those are fun visible milestones, but they tell a story of many harder sea forces at work.

We have incredible people who care about what we were building together and those incredible people are delivering incredible customer experiences.

Every bit as exciting as that, the speed of our improvement is accelerating.

We're continuing to invest heavily across every facet of the business, getting a little better every single day.

The selection, fees, fun and fairness we're able to deliver to our customers today is far better than the status quo, but is still a far cry from where we want it to be.

And that opportunity to improve is what motivates us.

These quarterly calls offer an opportunity to step back, pick our heads up from the daily grind of trying to improve and to look back at how all those improvements are adding up and translating into numbers.

They are translating.

We grew units and revenue at 99% and 110%, respectively, reflecting continued demand growth across cohorts and exceptional execution.

We also had a phenomenal quarter of GPU, posting over $2,400.

On our first public conference call in June 2017, we were coming off a year of $1,000 GPU and announced our mid-term goal of getting to $3,000.

At that time, $3,000 looked pretty far away, but we knew the path was there and execution was all that was in our way.

Over the last 2 years, we've executed quarter after quarter and today that goal is plainly in sight.

In fact, there's clear visibility beyond it.

We continue to believe execution is the only thing that stands in our way, and we plan to continue executing.

Our EBITDA margin was another high point for the quarter.

In 2016, heading into our IPO, we had an EBITDA margin loss of about 23%.

Since then, we have methodically marched that number down to about 17% in 2017, about 10% in 2018 and 7.4% in the first quarter of 2019.

That's a lot of progress, made very quickly with consistency and it speaks to the efforts of our people and the leverage in our model.

We've continued to invest heavily in the business.

Since day 1, our primary focus has always been on delivering the best customer experience we possibly can and investing in all the areas that enable it.

As we've gotten larger and grown faster, areas of investment that have received elevated attention are creating more customer self-service functionality, automating more parts of the transactions and producing more high-quality internal tooling, all of which allows us to further simplify our customer experience and to build a more scalable and efficient business simultaneously.

These investments have taken and continue to take many forms, and we're pleased with the progress they are enabling.

We broke our record for sequential unit growth in the first quarter nearly doubling our previous best.

Along with the effort required to handle that increase in sales, we also added 24 markets of vending machine in Pittsburgh and opened our sixth IRC in Cleveland in the first quarter.

Despite all of that growth and all of that investment in future growth, we were able to lower our EBITDA dollar losses quarter-over-quarter.

In the first quarter, we also completed our first securitization.

It's a significant accomplishment that required a lot of work and preparation.

All the hurdles were cleared.

The deal was completed laying out a template for similar deals in the future and lighting the path for significant progress in the finance contribution to our total GPU.

The calculations we've done historically to determine what we believe is ultimately achievable for our finance programs simply rely on the way other mature issuers in the market are executing similar transactions today.

So in a sense completing our securitization doesn't really impact our view of where we're headed at all.

That said, it does mark a significant milestone and provides us with a nice empirical data point with our own loans that cements our view of what is possible.

We also continued making exciting progress in our business of buying cars from our customers where we grew units by 230% year-over-year, which now represents us buying 40% as many cars from our customers as we're selling.

The experience we offer our customers when we buy a car from them is very simple.

They go to carvana.com, get a value, pick a time for us to pick up their car and we put money in their account.

That customer-facing simplicity requires complex technology and operations capabilities in the background, and we're very excited to see it generating the same kinds of reviews from our customers that we've seen on the retail side of the business, which we expect to fuel the same kind of explosive growth.

It's early days for Carvana in general, but it's even earlier for our business of buying cars from customers.

We'll take the same long-term approach developing this business that we've taken on the retail side: constantly experimenting, not been afraid to take risks and always searching for creative ways to give our customers even better experiences.

Our results in the first quarter are very exciting, getting us to a $3 billion run rate, but it's important to put that scale into the context of our enormous opportunity.

In the first quarter, we estimate we had approximately 0.35% market share.

That means only about 1 in 300 customers who bought a used car in the U.S. in the first quarter bought it from us.

In our most mature market of Atlanta where we've been operating for just over 6 years, we have approximately 7% unaided awareness.

That means only about 1 out of 14 people in our most mature market were even able to list us when asked to name a business where they can buy a car.

There are a lot of ways to read into those numbers.

Our view that the most important by far is that we're just getting started and our runway is long.

Our customer offering is desirable, highly differentiated and rapidly improving.

We have strong leaders and passionate teams across Carvana working at a fever pace to make us better at every aspect of our business.

We have our eyes firmly fixed on our goals of selling 2 million plus units per year and becoming the most profitable automotive retailer, and we're pleased to report another quarter of successfully marching along that path.

Mark?

Thank you, Ernie, and thank you all for joining us today.

Unless otherwise noted, all comparisons are on a year-over-year basis.

We're pleased to report another quarter of exceptional growth in both retail units sold and revenue.

Retail units sold totaled 36,766 in Q1, an increase of 99%.

Total revenue was $755.2 million, an increase of 110%, marking our 21st consecutive quarter of triple-digit revenue growth.

Our growth this quarter reflects a strong tax season and the many investments we made last year and continue to make to further improve scalability and satisfy increasing demand.

Total gross profit per unit in Q1 was $2,429, an increase of $575.

Our growth in GPU was broad-based and reflected gains in retail, strong monetization of our finance platform, progress in buying cars from customers and higher attachment rates on ancillary products.

EBITDA margin was negative 7.4% in Q1, an improvement of approximately 5%, reflecting meaningful operating leverage across the entire business.

SG&A levered by 2.8%, and advertising levered by 1.7%, benefiting from declining CAC in existing markets and lower initial CAC in new markets both due to continued increases in awareness of our brand.

While we have shown rapid EBITDA margin leverage over the past 2 years, we have now started to see margin improvements show up in smaller sequential dollar losses.

The first quarter marked a significant milestone in this respect.

Over the past 6 years, we have shown a sequential reduction in EBITDA dollar losses in 2 quarters, both heading from Q1 to Q2 in 2017 and 2018.

This year, for the first time, we showed a sequential reduction in EBITDA dollar losses heading from the fourth quarter to the first.

This is a meaningful achievement and carries a lot of positive implications about where we're headed.

We ended the quarter with $342 million in total liquidity resources and more than $50 million in additional real estate and securities on our balance sheet.

Giving us substantial flexibility to execute our plan.

We opened 24 markets in Q1 and have opened another 15 so far in the second quarter, bringing our current total to 124, up from 85 at year-end.

We now expect to open 55 to 60 markets in 2019, bringing our end-of-year total to 140 to 145 markets.

In Q1, we also started production of our sixth inspection and reconditioning center in Cleveland, Ohio.

And following quarter end, we started production at our seventh inspection and reconditioning center in Nashville, Tennessee.

This raises our production capacity to approximately 300,000 vehicles per year at full utilization.

Q1 marked another exciting milestone for our finance platform.

On March 29, we closed our first auto loan securitization, successfully adding a new large and diverse monetization channel.

We sold 350 million of principal balance of loans in the securitization to approximately 30 investors, who were new to the Carvana platform.

We're excited about what this means for our finance platform and expect to recognize additional gains on our way toward our long-term financial model.

In terms of outlook, we're raising our full year guidance for retail units sold to 165,000 to 170,000 and total revenue to $3.5 billion to $3.6 billion and reiterating our guidance for GPU and EBITDA margin reflecting strong execution by our team.

As we look forward to the remainder of 2019, we're excited about our progress toward our long-term financial goals.

Thank you for your attention.

We'll now take questions.

(Operator Instructions) And our first question comes from Zack Fadem of Wells Fargo.

First question here on new market openings, 24 in Q1, your largest ever in terms of size, but as you start to enter markets of smaller average populations, curious if you can talk about strategy here first in terms of market entry costs whether there's any leverage from closer existing markets and then second in terms of performance and how you think about these cohort curves will compare relative to the more densely populated areas that you've entered over the years?

Sure.

So as we kind of said in our last quarter, I think our plan for this year is mostly driven by filling in markets that are nearby existing markets and kind of filling out what already we can reach in logistics networking, so the expenses of opening these incremental markets is definitely less since we generally don't have to add long legs to the logistics network.

The largest expense in opening these markets is generally just turning on local marketing.

And then given where GPU is today and how quickly these markets are ramping, the payback on these new markets is very, very fast.

So the economics of them look great.

I think we're very happy with the pace which we're opening markets, we're currently on pace almost for a 100, but we're guiding to 55 to 60 per year because we feel like that's a comfortable pace that makes sense for the business.

And then as it relates to how these curves are performing, we continue to see the same trends that we've seen in the past across all of our cohorts.

Our oldest cohorts are still growing very quickly.

Newer cohorts are generally growing faster than previous cohorts.

And then smaller markets are also growing faster.

And so these are generally new markets that are smaller markets as well.

So there's 2 reasons for them to grow fast, and we're seeing them grow very quickly as a result.

Got it.

And then it looks you sold more than 5,000 vehicles that you sourced from customers in the quarter.

So as you continue to get better there, curious how or when you think this translates to a more meaningful reduction in vehicle costs?

And then second, curious if you could talk about how the attachment rates are for customers who are both buying and selling a car from you at the same time?

Yes, I think the way that we try to think about that, the math of buying cars from customers, is we try to collapse that into the single number, how many cars are we buying from customers relative to the cars that we're selling to customers.

And so that number is 40% this quarter, which was up from 36% in the previous quarter.

So we continue to make really nice progress there that we're very excited about.

From there, those units then get forked.

Some go to wholesale markets, where we sell them at auction and we earn a margin and we had a great margin this quarter, we're very happy with that.

And then some go to retail and are sold to other customers and then contribute effectively to incremental margin because we're able to buy them less expensively.

I think the way that we think about that is, in general, is that you're the market leader in that business is buying roughly as many cars from customers as they're selling and is able to achieve roughly $900 to a $1,000 on the wholesale side, which is visible in their financials for every car that they originate.

And so we think that the total GPU opportunity is about that 100% times $900 to $1,000, and we're obviously achieving a fraction of that today where we're buying 40% as many cars as we're selling and on the wholesale side in the quarter, we had about $450 of wholesale margin.

So I think there's a lot of opportunity left there.

We think the most important thing is that we continue to grow it methodically, we continue to make sure we're delivering great experiences.

And there's clear areas to improve the economics over time that we're aware of, and so I think that's how we'll think about that business.

Mark, do you want to jump in on the second question?

About attachment points or attachments for customers that buy and sell.

Any meaningful differences there?

Sure.

So I think are you asking about trade-ins on that side?

Yes, yes.

Yes.

Sure.

So a component of our business of buying cars from customers is trade-ins and a portion is buying cars directly from customers without necessarily a retail sale.

I think historically on the trade-in, attachment, we've bounced around in the 15 to 20 range, and so, yes, hopefully that answers your question?

It does.

I really appreciate it.

Our next question comes from Chris Bottiglieri of Wolfe Research.

So a 2 part.

I guess first in context of this very robust level of growth that you're seeing, particularly relative to the kind of how you ended the year and how you were guiding, was wondering if you could maybe attribute or talk just through some of the levers that you think led to this like pretty enormous level of growth?

Sure.

I think there's maybe 3 things we call out, 2 on the demand side and one on the supply side.

So on the demand side I think, first and foremost, we continue to see really robust growth across cohorts.

Demand just feels very strong and we're excited about that because it carries a lot of great long-term implications.

I think on a more transitory note, we do believe that we had a great tax season.

Total tax dollars returned to customers -- our best numbers we've seen suggest they were down about 2% to 3% year-over-year, but we believe that more of the dollars took the form of child tax credit dollars, which are more efficient in terms of driving customer car purchases.

So we think that the tax season was also very strong, and we think we benefited from that a bit.

And then I think those 2 things led to higher levels of sales and then importantly, the business held up very well, the operations teams executing incredibly well.

And so I think we executed and were able to handle the growth.

That's helpful.

And then your average monthly website traffic really spiked this quarter.

Can you talk about the growth you're seeing there and where you are in terms of the search engine optimization and customer acquisition?

And then lastly, you highlighted a new Board member addition.

I could see how someone coming from a background in all my travel agency space would help, but maybe just kind of talk about this holistically in where you are and how this all fits together?

Sure.

I think as I said we've seen robust demand growth whether we define it as sales or website visits or kind of any stage of the funnel.

We think there's probably a lot of drivers to that.

We think we continue to have the best offering available to customers who want to buy a car and get a simple experience that's fair and have it delivered directly to their door, and you have a 7-day return policy, we think that's a very, very high-quality offering.

And then we think that we're also starting to see some of these relevant facts that we've talked about over time showing up.

So we've always kind of talked about 3 network effects in the business.

One is, as we grow markets and grow inventory, that should increase conversion across markets.

One is, as we put more inspection centers closer to customers, that reduces delivery time and that should increase conversion in the markets near those inspection centers.

One is, as we continue to invest more in national marketing and then we open more markets and therefore have more ad dollars put to work and are able to pour those into national marketing channels that also should feed back positively.

And I think we're seeing a lot of evidence of that.

An example is the Indy IRC, which we opened in the first quarter.

We just started ramping up inventory there in the first quarter and we saw our conversion rates for all of our Midwest markets bump up considerably, and to a degree that we would have expected, given all of our models that we kind of had before, but it's nice to see that working.

And then I think maybe the national brand is starting to take hold.

I think as I said in my prepared remarks, we still have a lot of room to grow in awareness, but I think that we're also a lot stronger than we have ever been from an awareness perspective.

And so we think that's really exciting.

On the SEO/SEM front, we're constantly moving forward there and I think our foundation is great because we've got a really high-quality customer experience.

We've got lots of cars on our website.

We've got a deep experience that allows customers to spend a lot of time on our website, which is helpful in terms of driving the quality of experience they get and the way that Google and other search engines would measure quality.

So I think that's great.

We're definitely seeing early signs of progress in SEO, and we continue to see significant gains in SEM.

So I think that's exciting as a testament to all the work that our team is doing.

And then as it relates to bringing Neha on board, which we're really excited about.

She is spectacular, and she got our first -- her first Board meeting a couple of weeks ago.

She definitely has an experience set that we didn't necessarily have before with everything that she'd done at Hotwire and everything she's seen there on the customer acquisition front and just generally has a measured perspective that we think is additive to the group.

So that's how that all fits in and, as I said, we're excited to have her, we're excited about all the progress we're making, generally.

Our next question comes from Sharon Zackfia of William Blair.

A couple of questions.

So I think in the text of the shareholder letter you talked about initial deployment of the AI tech that you purchased.

So if could you give us any update on what you're seeing there?

And how we should expect that to influence the business going forward and at what pace?

And then secondarily, Ernie, you talked about a time for delivery.

Are there any metrics you can share for us on what the average time is for delivery today versus a year ago for consumers at this point?

Sure.

So first, yes, we completed this acquisition of Propel about 6 months ago, give or take, and then we're starting to implement that, what we're calling our next-generation customer communication platform.

That includes AI chat, it includes basically dynamic FAQs and our support menu that based on customer click-through activity put the right questions in front of the right customer at the right time.

We also have short-form videos where we do the same, they're constantly updating and learning to make sure that we're answering customer questions in automated ways and trying to kind of clear that work off our advocate's plate.

And I think you can think of AI chat as being effectively the same thing.

It's a way for a customer to come in and in natural language let us know what they're concerned about or trying to understand in that moment, and then make sure that we can put the highest quality content right in front of them at that time.

We've completed several of our initial rollouts in our next-generation communication platform.

And currently those are being tested to different size of populations depending on which part we're talking about.

And I think we're seeing meaningful contributions that are generally taking the form today of basically fewer calls into advocates and customers being able to make it deeper into the funnel without having to have their answers -- or their questions answered by a person, which we think is great.

And so I think there's a lot of opportunity there.

It's still very, very early.

The foundation that we're building on top of we feel really, really good about, and we feel that we're just scratching the surface.

The quality of results we're seeing right now is definitely material, but we think that there's a lot more room to improve.

So I think that's something that is a several-year road map that we've got laid out in front of us that we'll be continually working on.

And then as it relates to delivery time, we definitely handled the tax season sales burst more gracefully this year than last.

We've talked a lot over the last couple of quarters about investments we made in scalability, which includes things like we just discussed on website customer communication, but also includes enhancements to our scheduler, the way that we're handling customer verifications, just preparing to have more people and more trucks ready to see the growth that we expected through tax season.

So we're definitely better off than we were a year ago to the tune of probably a day or 2, give or take.

Our next question comes from Armintas Sinkevicius of Morgan Stanley.

If I look at the outlook for the rest of the year here, it's about negative $100 million of EBITDA, negative $45 million of cash interest and then call it minus $100 million of CapEx.

So that's minus $245 million.

And then meanwhile, you have $85 million of cash and $112 million of real estate.

So call it about $200 million there.

If the floor plan covers the inventory and the working capital, there's a bit of a gap there between the cash burn and the liquidity available.

Anything I'm missing there or if you can walk me through how you plan to fund the business into the end of the year?

Sure.

Yes, I think there's a major thing you're missing as a starting point, which is we have a significant amount of immediately available liquidity on our floor plan line of about $182 million.

So to look at the complete picture, I think, we had about $342 million in liquidity resources at the end of the quarter.

As you mentioned we had a little over $100 million in real estate and another $20 million or so in securities on our balance sheet.

All of that leading to significant flexibility for us as we continue to march down the path on our plan.

I think the other kind of key components on cash usage over time, our EBITDA and CapEx are 2 you pointed to.

I think you can see from our guidance and also from our commentary about the trajectory that EBITDA is headed that we expect declining EBITDA losses in the back half of the year or back 3 quarters of the year, I should say.

And then I would also just reiterate the point, which you did allude to, we've had great success financing our CapEx historically -- a great example of that is in Q1 we completed our first sale of leaseback transaction for an IRC, completed a sale leaseback of the Indy IRC, which came online around the end of 2018.

And all of that points to significant flexibility.

Okay.

I can appreciate the floor plan angle.

I just thought the floor plan was tied to inventory.

So is the floor plan being used to finance anything?

It's inventory that we already have on balance sheet.

Okay.

So you're going to effectively -- okay, I got it.

And then the other one is with the securitization -- first of all, congratulations.

And the question I have there is, how close does it get you to that 500 to 700 target that you laid out at the Investor Day late last year?

Sure.

So I think, it's a starting point, and I also agree, I think, the securitization was a big success I think for the whole team and the whole company.

We're very excited about it to be able to bring our offering to a much, much broader market.

I think some key numbers there to keep in mind, I think, before securitization deal fees and expenses, which were significant, on this inaugural deal in particular, we did about $707 of total finance GPU that's -- a meaningfully step forward from where we were at Analyst Day, but only a partial step toward our long-term goals.

We do see the inaugural securitization as opening up a new market to us, which we believe will be a key player in allowing us to accomplish those long-term goals over time.

Our next question comes from Nick Jones of Citi.

Just one question.

Is your approach 2 million units over time and that's kind of long-term goal, I guess presumably you'll have to go for older cars, how does the reconditioning and that kind of pan out?

Can you do more disclosures on the website or through your messaging app where you can maybe do less reconditioning but give more information to the consumer or will it just cost more to recondition?

Sure.

So I think -- let me break that in 2 pieces.

First, I think I just want to make sure we're being clear about the size of the market and the size of the market in cars that are less than 5 years old.

The best numbers that I can remember and I could be off these a little bit is that around 50% of the market, give or take, is cars that are approximately 5 years old or less.

And so we're talking about 20 million units per year there.

So I think there's plenty of room to grow inside of less than 5-year-old cars, for example.

And I think also if you look at Atlanta's market share in Q4, which was nearing 2% and you extrapolate that nationwide given the cars that we're selling today and then just kind of add a little bit of growth underneath it, you can get to pretty big numbers pretty fast.

So I don't think we necessarily feel like we would need to go to much older cars to be able to sell too many cars per year.

With that said, we do believe that we are able to go to older cars, and we see great evidence of that.

We've sold cars up to 9 years old so far, and we don't see materially worse reviews.

We don't see materially higher return rates.

So we don't see a lot of evidence of there being customer experience friction, and we do see evidence of our ability to execute on those cars.

So we think that there's absolutely opportunity to continue to move to older cars with higher mileage, and we think there's opportunity to move upmarket to more expensive cars as well.

So over time we expect to continue to migrate in all those different directions and the way our system is built.

It's largely kind of a pull system so if we can find and attract those customers who want those cars, then our kind of inventory algorithms will automatically go out and look for those cars and we'll start to buy more of them and so we would expect to be growing both within 0- to 5-year old cars and across car types over time.

Our next question comes from Ron Josey of JMP Securities.

Just wanted to follow-up on 2 things: one is the new IRC planned for Cleveland.

Can you just talk about the approach just overall about IRC expansion as you get to that 5% share that we've been talking about longer term?

And then, Ernie, when thinking about new markets, I mean the 24 in the quarter is pretty impressive and things are clearly going great, but it wasn't too long ago where there were growing pains in terms of things that need to be fixed, make things more efficient, more electronic?

Can you just talk about how like the process for growing these new markets is going?

I'm assuming things are going really well, but any ways that you can improve this process and obviously hard to answer that question given 24 launched, but I had to ask.

Yes, of course, I mean in terms of the IRC plan, our long-term plan is roughly to open 40 inspection centers that are able to produce about 50,000 cars per year around the country near 40 large population centers that enable us to deliver cars very quickly.

That's kind of the long-term plan.

Cleveland, Nashville fit into that plan very well, and then we'll continue to add cities over time.

I think as a bit of an aside, but a nice data point.

One of our inspection centers in the last week actually had its first week where it produced 1,000 cars in a week, which is a nice operational data point.

There's clear evidence that we're able to produce 50,000 cars per year out of these facilities.

That's the first one that's got ramped up all the way and has been able to hit that milestone.

So that was a big one for us that we're excited about.

I think on the market side, you're absolutely right.

A year ago, we had a great Q1, and we found ourselves a little bit constrained and struggling under the weight of the growth in many respects and then I think when we're trying to size our market openings at any point in time, we need to be thoughtful about that.

We think that opening markets is not operationally difficult for us at this point.

As I said, we're -- we are on pace in the first quarter to open nearly 100, and I think from a market opening ops perspective, that's the pace that we probably would be able to handle.

It also economically makes a lot of sense because we've got higher GPUs and the markets are ramping faster and there's not a ton of expense to open them up.

So that makes a lot of sense.

The questions around how many markets to open and how we should think about that, really do boil down to kind of the extra operational strain that it puts on the rest of the system, the shared parts of the system, the logistics network, the inspection centers, the inside advocates.

Are we able to handle that or not, how kind of quickly do we feel like we can ramp that up if more demand shows up than we initially forecast?

I think that drives it and then the other driver of that decision is looking at all these markets and saying okay, as we open these markets, what's our service quality going to be when we open these markets?

Do we have cars that are nearby?

Can we deliver them quickly?

Or are we going to open up and have 18,000 cars at our site, but they're all going to be 5 or 6 days away because they're all on the opposite side of the country.

And so we're trying to be thoughtful about that as well.

And that kind of interacts again with our inspection centers and where we're opening them and where we're putting inventory to enable us to open up more markets.

So there's a lot of considerations there.

I think if we're doing something smart then that means there's a good argument we should be ramping faster and a good argument we should be ramping slower and we're trying to find the balance between all those arguments as best as we can and feel good about where we are.

Makes a lot of sense.

And real quick just that 1,000 IRC goal you hit, is that a new IRC or existing ones?

That was in our Phoenix IRC, which was the first one that we built from scratch.

Our next question comes from Steve Dyer of Craig-Hallum.

Revenue for the quarter, I think, your guidance -- I'm sorry, for the balance of the year was up $100 million, EBITDA, absolute EBITDA when you flow it through sort of the midpoint of the guided range yields EBITDA on an absolute basis of about $5 million less, so sort of still implies negative incremental margins.

Is there a level, whether it be units or revenue per quarter, where you see those incrementals start to turn positive?

Sure.

Yes, I think this quarter is a good example of positive incremental margins.

Obviously it's declining sequential EBITDA dollar losses going from Q4 to Q1.

This quarter represented the biggest improvement in EBITDA dollars that we have seen in our history as a company.

It's also the only time we've ever shown positive EBITDA going from Q4 to Q1, which is -- has some seasonal elements in it.

And so I think we're feeling really good to your question on showing positive contribution sequentially, which is what we just did, and I think you can see from our guidance how we think about that on a year-over-year basis as well.

So I think, to round that out a bit more, our guidance for the year is 3.5% to 5.5% EBITDA margin, that's a significant increase or improvement from the prior year.

That significant improvement comes from all types of sources.

So it's broad-based gains in GPU, broad-based gains in SG&A leverage and we think places us on a very clear path of profitability.

Yes, Steve, all I would add to that is, in order to offset kind of the implied increase in EBITDA dollar losses there, we would have lowered our EBITDA margin guidance by 0.1%.

I mean that would have offset it.

So it's just, I would say, we're in kind of rounding errors at that point, and I think we're feeling really good about where the business is headed, but I wouldn't read too much into it.

Got it.

Okay.

And then just on the cars that you're buying back from consumers, you're obviously starting to see a lot of traction there.

Is there any way to sort of quantify or bucket sort of what is coming through your website versus the kinds of stuff that it's literally boots on the ground going through Craigslist, going through ads, those types of things?

It's -- the vast majority of it's just on the website, people going through, getting a trade-in, getting a value on their car.

We kind of called out in our shareholder letter, this quarter was the first time that we did a TV advertising campaign specifically pointed at driving customers that are going to buy a car from us.

And we saw great results from that, that we're very excited about.

I think hidden in that story is another actually exciting customer acquisition cost story.

Mark talked about the leverage that we saw in customer acquisition costs that leveraged 21% year-over-year.

I think something that makes that even a little bit better is that a lot of our advertising dollars or at least a meaningful portion of our advertising dollars this quarter were pointed at building a brand that we buy cars from customers as well.

And that's definitely less efficient at driving retail sales, but despite that we still saw the growth that we saw in the retail side and we still saw the leverage that we see in customer acquisition costs, which is effectively taking all ad spend dollars and dividing by retail sales and therefore, kind of missing some of the benefits of the brand building that we're doing there.

So we feel really good about all that.

Our next question comes from Seth Basham of Wedbush.

My question is around retail GPU.

Could you just give us a little bit more color on the drivers for the sequential improvement, that would be really helpful?

Sure.

Yes, so there's a number of drivers of sequential improvement in GPU.

I think one is we do this Cyber Monday promotion every year in Q4.

That's a $1,000 discount for customers who take advantage of the Cyber Monday promotion, which has a sizable impact on retail GPU.

Seasonally, Q4 has historically been a relatively weak quarter for depreciation rates and that's industry-wide.

And so that has an impact on Q4 GPU as well.

Q1 is not a strong quarter for depreciation rates, but Q4 is particularly bad.

I think that, there's some other small sequential gains.

I think we made some little gains in recon for example, getting a little bit of leverage through the reconditioning centers but the seasonal ones would be the sequentially would be some of the bigger ones that I'd point to.

Okay.

Thinking about the balance of the year and thinking about things like inventory days, do you expect a lot of improvement in that factor, and do you think that's going to be a contributor to retail GPU improvement?

Sure.

So average days of sales is where we've made tremendous progress over the last couple of years.

For those of you that have been following us for a while, we were 91 days for the year in 2017.

Now we moved that down into the high 50s, low 60s range.

I think that's obviously had a positive impact on retail GPU other things equal.

I think as we look forward, we expect that to stay in a stable reasonable range.

We feel pretty comfortable at the level we're at.

We may see that bounce around but feel pretty good about where we are today.

In the long term, I think, that's a big opportunity to march it down steadily because of the nature of our model, which centralizes inventory and puts a lot of eyeballs on a fixed set of cars, opportunities to increase efficiency in the recon centers, et cetera, but in the near term, we're pretty comfortable with our range.

Our next question comes from Sameet Sinha of B. Riley FBR.

Couple of questions here.

So in terms of the IRC, the Indy IRC, can you help us think about what sort of a deleveraging impact that had in Q1 and the reason I'm asking for it is primarily because as you seemed to have increased the level of rollout for IRCs -- Cleveland, Nashville and then another eighth one coming on this year.

Just wanted to kind of get an assessment of how much -- what's the dampening effect?

Secondly, Ernie, I think you might have addressed it but from your guidance to the new market opens, you just have 21 markets left from now to year end, so is the plan to hit that and kind of just slow down that for a while, while you ramp up your IRCs or do you think there's a potential that you could increase your new market opens if you get to that number early on?

Yes, so on deleveraging from the Indy IRC in Q1, I would not think of that as being a particularly large effect.

I think Indy is now 1 out of 6 IRCs and so any individual IRC coming online will have some effect because new IRCs are underutilized relative to the utilization of other IRCs.

They may also have some services outsourced initially that we bring in-house over time but I wouldn't view that as a particularly large contributor to our GPU in Q1.

Yes, Sameet, really quick just to jump on your second question.

I think as I tried to outline before, I think we feel like we can open a lot of markets, we feel like that's a choice and the considerations are kind of as-discussed.

So we do plan to, in the back half, have slower market openings for all the reasons discussed and again, that's a choice.

We feel like it's the right choice for the business.

Our next question comes from Rick Nelson of Stephens.

So a follow-up on the cars that you're sort of seeing from customers, how would that retail GPU compare to those that you would source from auction?

Sure.

So I think first order, I think a good way to think about that is that the cars that we're selling in retail should have similar incremental margin to the wholesale margin that we get when we sell cars at wholesale.

The reason for that is basically, if we're selling a car at wholesale and we're earning a profit, you could imagine us buying it from ourselves on the other side, and so that kind of wholesale margin is the same kind of incremental margin that we get for a car effectively when we sell at retail.

So it's a first order -- it's very similar.

And that's why we like to think about the business of buying cars from customers holistically and just how many cars are we buying relative to the cars that we're selling and then how much are we able to earn incrementally on those cars whether through wholesale or retail.

Those numbers won't always be perfectly consistent but they are conceptually at least somewhat tied to one another, and ends up how we think about the business.

And if you could comment on the brand awareness where it sits today, and how it's grown?

Sure.

Yes.

I mean in my prepared remarks, I talked about our brand awareness, or our unaided awareness.

I just want make sure we're using the precise language here being about 7% in Atlanta.

We view that as really, really exciting.

So I -- and let me just try to give you our framework for how we think about the potential of the business over the long term.

Generally, we like to separate what we think growth can kind of -- the places growth can come from in terms of conversion, which is basically changes that we make to the experience customers that are already on our website get.

So if we buy more cars and we have more cars on our website, that leads to higher conversion.

You're just converting more of the same customers you already have.

If you can deliver cars faster, you're converting more of the customers you already have.

If you make product improvements that enable you to be more efficient and communicate better to customers, you convert more of the customers you already have.

Then you can always go out and acquire more customers and so when you're thinking about getting more customers, I think, there's kind of 3 steps before a customer is psychologically ready to buy a car from Carvana.

Step 1 is obviously they have to know that we exist at all.

And so that's this awareness measure.

We can't sell a car to somebody that doesn't know we exist.

Step 2 is they have to understand what buying a car online means.

If they know what Carvana is in concept, we sell cars -- they don't really even know what buying a car online truly means, they are not really a candidate for buying a car from us.

And then 3, they have to trust Carvana.

Even if they know what buying a car online means, they know what Carvana is, they have to believe that the car we deliver is high-quality, the 7-day return policy is real.

And so those are all the kind of vectors for growth to isolate into the awareness vector.

We think the fact that we're 7% unaided awareness in Atlanta is conceptually extremely exciting.

I mean that suggests lots and lots of upside and then I think there's this question about how quantitatively can you look at that?

It's pretty clear that directionally there's a lot of upside there, how quantitatively can you look at that and what can we take from it?

And I think that it's probably dangerous to just do super simple extrapolations or implied extrapolations off of that because we don't quite know how that's going to unfold over time, but we do know that 3 years ago, our unaided awareness in a very similar study that we did in Atlanta was about 2%, and so we've seen our unaided awareness over the last 3 years grow by about 3.5x.

Depending on what quarter you're looking at, we've seen growth in Atlanta in units grow by 2.75 to maybe 3x.

So there's, at least, been a pretty tight relationship there that we think is exciting.

And so I think, virtually, any math you do off of that suggests very exciting things, and I think that's just more evidence that points to the direction of our view that the demand for this product is there.

We need to execute.

If we keep rolling this out and executing period after period delivering great experience to customers, making sure we're able to handle the volume, we feel like we've got an absolutely enormous opportunity in front of us.

And we that's the real risk.

(Operator Instructions) And our next question will come from Gary Prestopino of Barrington Research.

Just wanted to ask a question in terms of have you seen any changes particularly in some of your older cohorts?

In competitive responses, what you guys are doing out there in the market?

Is anybody -- some of these rooftop dealers trying to replicate what you guys have or is it even possible to do that?

Sure.

So well, let's go with the meat of that question.

We still have not seen an impact that is measurable in any of our markets that we would assign to competition.

That doesn't mean that maybe there's little things happening from a competitive perspective that we're not seeing or they're so small we're not noticing the kind of visible impact.

But nothing that kind of hits our radar that we feel like is measurable have we noticed in any of our markets at all.

And I think that fits kind of our mental model, which is this is a very, very large extremely fragmented market.

As I said in my prepared remarks, we have 0.35% market share nationwide.

In our most mature market, Atlanta, we have a little less than 2% market share.

It's just a really, really big market, and it's hard to imagine that anything that any other competitor does is likely to have too big of an impact on us relative to us just making sure we continue to execute into the huge demands that we see and the huge opportunity that we see.

So we're really focused on execution and believe that's the risk.

I think even when we try to think longer term, about what does competition really mean for us, I don't even know which direction we're rooting there because we do think obviously there's negatives to having more competitors out there that are doing something similar to you.

We think there's also positives.

In that kind of stream of awareness, understanding, trust.

A part of understanding is if other companies are out there telling people to buy a car online or telling people how buying a car online works.

That's actually helpful to us.

If other companies are out there telling people that buying a car online is a normal thing to do and that's helping to establish trust, that's actually helping us.

So we're not even really sure which direction that cuts and that's why we're just going to stay focused on our customers and on execution.

And then last point you made, I wouldn't want to guess at what other people can and can't do.

But we definitely think that this business model is very difficult to replicate.

It takes expertise in a lot of different areas.

You have to be able to buy cars intelligently, to recondition cars, to ship cars around, you have to have an expertise in building out a great e-commerce experience.

You need to be able to handle financing seamlessly because a customer's not going to separate all these different parts of the transaction.

They're going to -- however, the whole transaction feels is what their Carvana experience is going to be.

And so you need to be good at a lot of different things.

You have to give them last mile delivery with a smiling face who is as excited for their delivery as they are.

That's hard to build.

And so we think, like I said, we're not going to say what other people can or can't do, but it's hard.

You should also say imaging the cars too because you do a good job at that.

And then lastly, I couldn't find it but what were your day sales this quarter?

I don't have the Q. I don't have...

They were 63 days, down from 70 in Q1 last year.

Our last question will come from Derek Glynn of Consumer Edge Research.

How much in funds or liquidity did you unlock from the sale leaseback of the Indy IRC?

And as a follow-up to that the $75 million in availability under the sale leaseback agreements mentioned in the letter, does that include the IRCs or is that just vending machines?

So we haven't disclosed the specific value of the Indy IRC.

We've given it some color -- directional color on how much it costs us per line in the previous shareholder letter and that was on the order of $10 million, $12 million per line.

So to answer the first part of the question.

And then to the second part of your question, the master sale leaseback agreement that we have in place applies to both vending machines and IRCs, the real estate we have on the balance sheet is mostly vending machines.

Got it.

Understood.

And just secondly, in terms of demographics, have you seen any wider adoption of your services from an older demographic as you've increased your brand awareness and expanded to newer markets?

We continue to see somewhat similar, I would say, demographic trends.

I think the right first-order mental model for that is that we're selling through the distribution of used car buyers across the country in just about every demographic break.

I think there's been a couple interesting trends that we do view as positive, we continue to see a subtle migration in the direction of millennials.

We continue to see a migration in the direction of mobile.

We have more of our persons coming from mobile than ever before.

We see a migration in the direction of women, which we think is also very interesting.

In general, women make more purchase decisions per households in the U.S. and in automotive, that's not always the case.

So we think that migration is also positive.

So we view those subtle migrations as indications that we're tapping into the growing part of the population so we think that's exciting, but in general, we have seen similar trends that we have seen in the past.

This concludes our question-and-answer session.

I would like to turn the conference back over to management for any closing remarks.

Thank you for joining the call.

Thanks, everyone, on the Carvana team for all the efforts you put in this last quarter, they're definitely showing up in the results.

Please keep working hard.

We're on a great path and you should all be very proud of what you're doing.

Thank you.

The conference has now concluded.

Thank you for attending today's presentation.

You may now disconnect.