Wayfair Inc. (W) Earnings Call Transcript & Summary

Good morning, everyone. I'm Michael Lasser, the hardline, broadline and food retail analyst from UBS. We could not be more excited to have the team from Wayfair with us. Niraj Shah has -- in a very short period of time has established himself as a legend, starting Wayfair in his mid-20s with his partner, Steve Conine.

Yes, late 20s.

Late 20s, growing it to a multi multi-billion dollar business, which if you have not heard how I built this, it's a fascinating story. I will save that for another day. With us as well from the company is Michael Fleisher, the well-regarded CFO from the company. Jane and Arjun run the Investor Relations effort. And we have a bunch of questions. There's also an app. If you go to www.ubs.involved.events, you can put in questions that we can include during our conversation.

With that, an area I want to start is the key investment debate, as you know, is on if the company can maintain its sales growth while making progress towards its path to profitability, what drives your confidence, which you've articulated a lot of conviction that you can get there, but how are you going to maintain this balance between sales and profitability?

Yes. So one of the things we've always done is taken a long-term view of the business potential, where we have a $400 billion home market in North America, $400 billion home market in Europe, $800 billion market, which is increasingly moving online at a heavy pace. And we really focused on how do we ultimately become the home specialist winner in that online opportunity. And so, fundamentally, building the European business, building the logistics business, a lot of the investments we've been making in North America around imaging and the merchandising of categories, these are all oriented around getting to that position, which would be quite a large, but also quite a profitable business. And what we've tried to do is make sure that the unit economics along the way are very sound, but also, we're taking a long-term perspective in terms of how we think about the investments versus just saying, hey, how do we get x amount of profitability this quarter or next quarter? I think when we were -- so let me just give you kind of a rough way to think about it. 5 years ago, when we embarked on building the logistics business and building the business in Europe, we're about $1 billion in revenue. And that's roughly when we went public, 2014. And at the time, the things we wanted to do, they outpaced the ability for the business to generate enough contribution margin to be both profitable and invest against the long term. And then as we -- we're making these investments and building the brand, building the business, we're seeing kind of sustained revenue growth, where every quarter since we went public to now, so in those 5 years, we've grown at a heavy rate because the repeat business grows faster than our total growth. So we spent lot of advertising money to get new customers, but then the experience brought us back. If you look at the repeat order growth, it's been higher than our average. So the total growth is pulled up by the repeat growth, which is an outcome of the experience. Well, during that period, we grew from $1 billion to 2 years ago, we were around about $5 billion. And at $5 billion, you obviously have a lot more contribution. Our contribution margin runs around about 20%, right? So at $5 billion you have $1 billion in contribution margin, whereas at $1 billion, you'll only have $200 million. And we think the embedded profitability in the business is more than 10%. But just for round numbers, say it's 10%. So when we were $1 billion, we would only have $100 million of profits that if we were just focused on today, we could keep, but we wouldn't really be able to take these long-term investments, or we could invest. Well $100 million was not enough relative to what we want to do. So we invested that and then some operated at a loss. When we get to $5 billion, right, you have $500 million, and at that point, we still want to invest in Europe, that's a couple -- $200 million, $300 million; logistics, $200 million, $300 million; the U.S. investments in these categories, merchandisers and other, couple of million dollars. So we're still operating at a loss. Well one thing we found just where we are today, we're now around about $10 billion in revenue. And our unit economics have continued to get stronger. So let's just stick with that 20%, though. At 20%, you would then -- you'd have the $2 billion in contribution margin and the embedded profit, we think, around about 10%. You have $1 billion of embedded profit. Well, you would then, in order to operate at a loss, you'd have to spend more than $1 billion. So last year, this is roughly what we did. And what we ended up finding was that while the long-term aspiration we have still the correct one and while the individual efforts all were sensible, in aggregate, the sheer amount of things we're trying to do was causing our own kind of execution inefficiencies. Just in a given area, there's clearly priority A things to do and priority B things to do, priority C things to do. Well that area just adds enough people and adds enough resources to do all 3. And you say, well, that's great, you're going to do all 3. And as long as they have good ROI, that would be great. The problem comes in the Tier 3 things distract whoever is leading it from the Tier 1 and Tier 2 things. The Tier 3 things need resources from other groups, maybe some help from one of our logistics groups or help from one of our merchandising groups, so on and so forth. And so the sheer number of things that need help from different groups makes each group needing to do this many things to this main things. And you get to a place where all of a sudden, you're getting things done less quickly than you were before, even though you have a lot more resource than you did before. So that's obviously not what you want to do. And so when we clearly saw this in the fall -- we do business reviews twice a year. So we saw this in the fall, what we said is, "Well hey, we really want to get done." It's all these Tier 1, Tier 2 things were at the top of the list. Every group knew what those were. They make a big list of all these things. It's quite a lot. Everything we ever talked to you guys about on stage is already on that list. We never talk to you about the Tier 3 things. So you've never even heard about any of the things that we'd say were the less important things.

No better time to tell them than right now, maybe.

Yes, we're still running through the list, yes. What we found is when you say, "Hey, I want to get these hundred things done, but I'm going to make sure I get these hundred things done by not doing these other things. And just telling everyone, this is what we're focusing on. And so every group knows what's expected of them by every other group. It's very clear what the priorities are in the company. You organize all the resources against those things, so that nothing is short of any resources. You eliminate resource contention. All of a sudden, what's interesting is those things can move more quickly and, frankly, you're not -- you don't need to spend $1 billion on those things. And what we've realized is that you can only grow the resource pool at a certain rate and still absorb the people and have them be productive and absorb the new initiatives and have them be productive. So what we think the irony of where we are today is, as we run through time, as we run through this year, we've talked about how OpEx as we run through this year and you get -- later in this year, you're going to see OpEx as a percentage of revenue lever. We said ad costs, you're going to see that lever. And I can explain some of the things we've done to make sure that ad cost levers. We've just talked about how gross margin will expand. And we're not saying that will happen next year or that will happen the year after. That will happen this year. And we talked about how the U.S. will get to sustain profitability next year. And that's something we generally only guide the current quarter, but Michael said it in his guidance, and Michael's typically somewhat conservative. So when the thesis is that's going to happen in 2021, we obviously believe we have a clean path to march there at a really good speed. Well, the reason we believe that is not actually because we're cutting efforts that we think give us a long-term gain. It's actually the opposite. We believe that because the things we're doing are what are going to let us actually be really effective at these big investments we're making. They happen to also have significant ramifications on reducing cost. And so that's why we feel very confident about it. That's why we think it will play out very quickly. And that's also why we don't think it undermines our long-term outcome and why we think revenue growth actually will be faster in the future than it is actually even right now.

You mentioned that this was a result of a business review that you do twice a year. What prompted you to be able to see this now? Because presumably, that clutter, that stickiness that was glueing down some of your progress had always existed in the business. It just seems that you're able to visualize it today.

So we ramped the headcount aggressively in 2018, 2019. Over those 2 years, we went from 4,000 people in OpEx to 8,000 people in OpEx. In 2018, we realized a number of initiatives we had started to be so large, we needed a systematic way to review them. So we created this context of these individual teams around each initiative with the structure of reviewing them twice a year. We created that at the end of '18. First time we did those reviews was in May of '19, second time was in November of '19. The third time is coming up in May this year. In the May reviews, it was the first time doing them where we did all -- they're basically like 95 of these things, we do them all in a relatively condensed period of time. The beauty of that, you're not so much trying to decide if the initiative makes sense, and you're not trying to actually -- there's someone already in charge of it, and you're assuming that they're doing a good job running it. But what you're trying to do is identify patterns that emerge across all these different things. And the pattern in May, it's our first time doing them, and generally, the plan seemed thoughtful. We're sort of digesting kind of how these reviews should work. You're basically making sure -- one of the goals is that -- we have 15 people in the company who sit in all of them. And for most of the 15, they really don't know about a whole bunch of them because they're really focused on running a very smaller subset of them, so it's really also a way to really get the cross-company interdependencies there. Well, in November, what happened is there was a pretty easy theme, I think, for everyone involved to see that the resource contention -- like, what you're seeing is we have more resource than ever going after these things. And the thematic thing on time lines are getting delayed and slowed down because of interdependencies that are causing these things from a prioritization standpoint. And so the theme comes up, it doesn't tell you the why. Then you go figure out the why. The why is what we -- we have too many. We have too many. You kind of might have had a supposition of that, but then you dive in, it becomes very clear. Then you basically say, okay, well, let's tackle that. And you also realize, well, the underlying way you got there is there wasn't enough scarcity constantly being imposed to basically say, well, the overall resource growth can only be x at most. And if the ROIs in there should be less than X, but it can't be more than x. That then sets you off on this journey.

So the narrative coming out of the last couple of months and some of the news coming out of Wayfair has been, oh, the company's pivoted towards profitability and maybe away from growth. And it doesn't seem like that's the right message. It's more so we're coming up on more efficient growth than we've had in the past, is that right?

So the way -- yes, so the way I would describe it is we're focused on being really efficient. We think that's why we succeeded in the past, and I'm being very long-term oriented. However, today, we think we can also commit being on a pretty aggressive ramp on profitability at the same time, because we don't think they need to be mutually inconsistent anymore because of the scale we've gotten to. So we think we can commit to seeing profitability ramp in the near term.

In your example of a department having A, B, C, D priorities, and D getting in the way because it requires help from logistics. Now that D is removed, how quickly do you think that change can manifest to these outcomes that you're anticipating, which is having good sales growth and working towards the path to profitability?

Yes, so -- yes, so a couple of thoughts there. So the sales growth, the other thing I'd comment on sales growth. So in our guidance for the first quarter, Michael mentioned that growth quarter-to-date was running, I think, the quote was a little less than 20% or just under 20%. I forget...

Any updates, right now? No, I'm just kidding. If you want.

I'd like Michael to continue to be the CFO. If I start giving guidance, you'll no longer be interested. So I'm going to stay away from guidance. What I will say, though, is that growth is an outcome of a period of time where we think the efficiency of execution was not high. And frankly, where we've added an additional constraint on our advertising. Instead of just advertising being constrained at the channel level within a brand, we've actually added an additional constraint at the brand level to reflect the fact that repeat business is much more efficient on the ad cost side and saying that we would expect each brand to get leverage year-over-year due to that. And so we're in a period of time where we've constrained traffic even more than the new efficiency targets, frankly, would allow, and we certainly have the team and the marketing side distracted from figuring out how to grow traffic profitably by their goal of currently getting to a different efficiency target and then doing that. So I would say that you're going to see -- as you run forward in time, you're going to see revenue accelerate not just because of the productivity of the team, focusing on A, B and C instead of D, but also because of how traffic plays out as you go through cycles like this, which we've gone through in the past. And so there's a lot of reasons why that will play out. And then on profitability, we -- the profitability is not just driven by the headcount. So I think -- while I think we're in the news a few weeks ago, primarily because we've reduced headcount in the OpEx side by 550 people. So that's the headline that gets out there. That obviously impacts the cost structure in a positive way, but frankly, there's a lot of other costs we incur in our business. And so a lot of the opportunities around "efficiency" actually unlock other costs. So an example I would give, for example, back when we did studio photography, we only spent so much on studio photography because it's quite expensive. Then when we figured out how to create 3D models, and we got the quality of rendered photography to a point where it was indistinguishable between -- humans couldn't tell the difference between what we rendered versus what was done in the studio. And the order of magnitude is roughly a 10x. It was roughly 10x cheaper. It was 1/10 of the cost. That let us explode the amount of photography and imagery work we did and we have a lot of different use cases. So the imagery for Perigold or lifestyle brands, or imagery for the sale of lens on Wayfair, imagery for our house brands. We've always had our use case for imagery, imagery at the item level. Then when you get teams looking and say, "Hey, is everything you're spending money on highly productive?" The amount we spend on imagery is sizable, it's tens and tens and tens of millions of dollars. You look at it, and a whole bunch of like every use case I just read out loud is actually highly valuable, highly productive. But you start finding there are some items where we're putting imagery support behind them, but these items are not good sellers yet. They're not -- they're too speculative. You say, well, why are -- you look at the biggest -- why are we doing this? Is this right, is this right? So that team itself looks at and says, "Hey, we probably over expanded what we're doing in imagery. We're finding there's a -- I'm making this up, but 1/4 of the budget, 1/3 of the -- not budget but the spend or whatever, is actually allocated to a really low-return area. We should just cut that and just say, "Hey, we're going to continue to support items with great imagery, but once it shows x amount of traction or whatever." All of a sudden, you can free up significant amounts of money. And that's just one example. You find that in a lot of cases. And so this is the outcome of, frankly, having not imposed enough scarcity as we scale these teams. The irony is, it gives you a lot of opportunity to get tighter. And you sort of shed a lot of money fairly quickly from your cost structure. So Michael's guidance about '21 profitability, I think you would know, Michael would guide somewhat conservatively. So would we expect to see a march towards that well before January 1 of '21? I would think so.

Yes. I want to get back to your point about marketing because that does seem like a critical lever towards getting to this goal of profitability. And if we were to simplify the explanation around what was -- what's been happening with marketing and what some of the changes are, would it be that up until now, Wayfair had very stringent return expectations for each channel of marketing? And so you -- but whereas there might have been some overlap and redundancy, like you're advertising online, you're advertising on television. And then so now, you're looking at it in total. And so maybe you can take down the online marketing a piece because you're -- what you're -- the halo effect that you're getting from advertising?

I'll describe it slightly differently. So we've always had these channel-level constraints for payback. And then each channel is for each brand. So like Wayfair Canada would be a brand, wayfair.com would be a brand, Joss & Main would be a brand. So display -- I'm going to make it up -- say, display advertising had a 1-year payback, okay? So we'd expect display within Joss & Main to be a 1-year payback, or display within wayfair.ca to be a 1-year payback. And then, say, a more transactional channel, like Google product listing ads, maybe has x month payback, shorter than a year because it's more transactional. And so we've always managed within that. Well then what happened over time is -- and brands were at different stages in the journey, but brands get to a place in the journey, where a significant portion of their business comes from repeat customers. And so what you actually want to do is within the channel, you want the channel targets to start to vary based on how much is influencing new versus repeat. Because repeat are largely driven by the fact that the person has experiences with Wayfair, and the quality of those experiences drive their future propensity to buy and loyalty and so on and so forth. And so within each brand, we basically created an additional constraint saying this brand should be showing X amount of leverage. And these channels need to stay within this payback, and it's the more stringent of the 2. Because what you don't want is you don't want to recycle the money that basically is being allocated to repeat and just spend it even if you don't need to. You want to basically show that you want the repeat efficiency as repeats now, it's 69% -- it's ticked up to be 69% of the orders. It keeps ticking up. And I mentioned repeat has been growing faster than our average revenue growth every quarter since we went public. So it's kind of always been that case, of it ticking up. Well, it's ticked up substantially if you followed that journey over 5 years. So it now has a material swing on certain channels. And so it's just an additional efficiency constraint to just help us make sure that the spend is highly efficient. The overall ad cost will show some leverage because of that. But then, frankly, we think we'll be able to keep spending more and more money on advertising. It will just be more efficient. And the reason you'll be able spend more and more money is that you use targeting, you use the revenue attribution, which we keep making better and better, you use creative optimization of what the ad units you use are to keep driving up performance gains that unlock additional types of inventory. But if you look at the ad -- efficiency of the ad cost as a percentage of revenue, you would see a bit more from them.

There's 2 elements to this discussion, I guess. The internal one, which is the current constraints that the company set, the return requirements. And then there's the external side, which is there's a lot of money that's going to acquire customers within the home furnishing sector. It's the overall noise that needs to be cut through. So based on all that, is it becoming just more prohibitively expensive to acquire new customers in the categories that you serve?

So the way I would describe that. So we basically define the return threshold we want. We don't have a budget, and then you spend the budget and you see what return you got on the budget. Instead, we say there's basically an infinite budget, but the return needs to be X. By locking the return, then you may only be able to spend x amount but they need to say, well, if I want to spend more than the x amount I have to unlock that inventory within this return threshold. How do I do that? Well, maybe a better ad unit on the creative side or maybe better targeting or maybe better revenue attribution that lets me find where my spend is inefficient or efficient. And so you use all the things you keep unlocking more and more. So the way I would say is we've seen ad cost inflation, but what we've been able to more than offset it with is leverage because of those things that I just mentioned that drive gains and we've kept the return threshold the same. What I just described, as we said, well, we've tightened up the return threshold a little bit. And we would expect to continue to see inflation, and we'd continue to expect to offset it more than that by gains in those types of methods I just described, thus allowing us to keep getting more and more traffic. And that's been sort of the history, we've been able to unlock more and more traffic through all these things while keeping the ROI. We would expect to keep unlocking more and more traffic within the ROI threshold. Oh, but by the way, we just tightened up the ROI threshold a little bit based on having the repeat new mix kind of calculating how the leverage ratio was.

And as you get further into your growth curve, are new customers you're acquiring today as profitable as they might have been in the past, if you look at them holistically?

Yes. So they're more profitable. So if you actually -- we debated whether we should include this. So we had this chart we put in our quarterly investor deck every quarter from the time we went public, it was cohorts by year. And we update it every quarter. And then finally, after like a few years, we said, okay, we're going to stop updating that, we're going to go other metrics. And then for a year, we feel like, oh, we really love that chart.

Love that chart.

So then we added it at the end of last year. We said, all right, we added it. We're never going to do it again. Then this year, we had a bit of a debate, and basically, I thought we shouldn't add it. I lost. We added it. So we added it again. So if you look at the last investor presentation, just the one we just -- the quarter we just, the year-end quarter, you actually see the cohort chart, current, updated through January, whatever the last month we had the date on when we did earnings. And you'll actually see the annual cohorts each year is successively higher than the prior year, higher on the repeat. Now the lines that are less dense because -- the lines -- data point shows up once you have 3 months, but until it's more than -- until it has a full 12, the line moves around a little bit. Once it has 12 it doesn't move because it has the full year in it. But you basically see them just be higher and higher and higher. So each customer cohort by year that we've acquired is stronger than the year before. And that's continued to be the case. We would expect, at some point, that to diminish because we have so many customers, and like I said, it won't necessarily inherently be valuable. But what's happened is the experience keeps getting better and better. So a customer who shows up, and this is their first taste, their experience -- their entire historical experience average is higher. Right? Whereas the customer you got years ago, you keep making their experience better. And so you actually see the tails of those cohorts actually turn up and they give us more of their spend, but they're starting from a much lower base.

And one of the reasons why they have a good experience is because the investments that Wayfair has made over the years, notably in the area of logistics, and that's a big advantage for the company. So can you compare Wayfair's advantage of shipping economics versus its competition? It would be helpful to put it in terms of if you took a chair, how much would it cost Wayfair to get it from its central pool of inventory to its customer versus a traditional player? How much would it cost for that traditional player to get to its customer?

Yes. So let me talk about the shipping economics, they are slightly different way than you posit it. But here's the way to think about it. Of every revenue dollar, roughly $0.20 of that revenue dollar is spent on logistics. That logistics includes everything from -- majority of the items come into the U.S. from Asia. So it includes ocean freight, includes drayage, includes over-the-road transportation, includes warehousing, includes last-mile delivery. So you have all those different pieces, adds up to $0.20 of the revenue dollar. So $0.55 of the revenue dollar is the product itself without logistics, another $0.20 is basically the logistics associated with it. So it's a pretty big expense. The reason it's so big is if you just think about the goods we sell, they're pretty low-dollar value per cubic foot. They're generally large and bulky and often heavy for a relatively low dollar value. So just the logistics cost as a percentage, it makes sense that it would be higher. So it's basically 20%, $0.20. So what we've done is we've built a network to optimize that. Well, what does that mean? Well, for the goods coming in from Asia, we have, basically, contracts where we have a nonvessel ocean operating carrier license with the U.S. Maritime Commission and the Chinese Maritime Commission. So we basically are -- we don't need to go through a freight forwarder. We can directly negotiate with Maersk and Hapag-Lloyd and the container lines. We can then pass those rates on to our suppliers, most of our suppliers' small to medium-sized businesses. So typically, going through a freight broker and their access to ocean freight rates is as competitive as what we can get amassed, if not the same. We then can manage that flow. We can manage the drayage operation. So that's the inbound trucking to the warehouse. Our warehouse, if you look at the network of warehouses we built out, they're really dense along port-based locations where there's high population, right? So we have a new warehouse in Northern California, which is just a little ways out of -- outside of Oakland. We have 2 large warehouses in Southern California. We have one in Savanna, we're opening one in Jacksonville, one in Atlanta. We have a complex of a couple of warehouses in New Jersey. And so these are all near ports, but near population. So you can picture these goods, the container comes in, you don't really have much in transportation cost because you're not moving it inland a lot other than our warehouses in Kentucky, Toronto, sort of a different situation because it's a bit of a twofer, it kind of has both. And Dallas, we don't really have inland warehouses. So you only move things inland if they're basically going to service just that geography, you never would move it in to move it back out, right? And so we have efficient flow in. Well, that takes some cost out. We use the buying power to get better pricing. That takes some cost out. We then have it in that local area. And so the last mile delivery, if you don't have to move it that far, you can save money there. And so on the 2 main deliveries, which we do ourselves. We try to position goods in that geography. And so if an item, large item comes into the East Coast and stays on the East Coast, you save $60 than if you basically bring it into Southern California, which is the cheapest inbound point, but then you truck it across the country to deliver it on the East Coast, it costs you $60 more in total. Now 70% of the population is on the East Coast, and yet almost all of our suppliers on a drop ship basis, but their warehouse is in Southern California. Why? Because it's the cheapest inbound point. And they're not worried about the total cost, they're just focused on their cost, right? But if you think about it, together, between us and our supplier, we should care about the total cost because the total cost, you figure out how you split that money. But at the end of the day, 60 extra dollars in transportation costs adds value to who, right? Well, maybe it adds value to the shipping carrier, but it doesn't add value to the customer, it doesn't add value to us, doesn't add value to the supplier. So all these things are optimized. We do a lot of sortation in these buildings so that for the small parcel carriers, their hubs get congested, so we can deliver to forward hubs or one of the things we'll be able to do this year with the volume, because the cascade volume has grown 48% year-over-year, we'll be able to induct into their terminals, even skipping their hubs, which will get congested during peak times. So you basically -- it's not like one thing on the logistics side takes your costs down materially. It's actually you can take the cost out materially because you shave this, you shave that, you shave that. The customer, in the meantime, gets a faster delivery, which you can promise them and badge that fast delivery speed, they love that, so it ups your conversion, which gives you economics on the advertising side. And then the other thing you get out of it is, frankly, you lower the damage rate. Because these items are prone to damage. And every additional handling touch drives up the damage rate. So when you take out touches, you basically drive down the damage rate.

That's helpful. Is there a way you could contrast that to what somewhat an organization that's clearly not as efficient might look like? And where would it go in a steady state environment, if it's $0.20 in a dollar today, can that get down to $0.10 to $0.15. And a competitor is still going to be at $0.25 to $0.30?

So I think most of our competitors, they have a very narrow set of SKUs that they have in their own logistics operation, and their logistics operations are generally not optimized for the big large items. So those warehouses tend to be suboptimized for them. And they typically sell other items, like they might sell tires or snow blowers and they mix those into those buildings, so they can't control for the damage as well. And then they have a larger amount of selection that they drop ship. And when they drop ship it, they're shipping it basically from California to wherever the customer is, which is basically the population. And so I believe our network is really the only one that's basically tackling those pieces, and there is significant margin gains you take out. In terms of what the $0.20 goes to, I mean, the truth is it'll keep going down as you get more and more scale in these operations. But I'll get in trouble if I tell you what I think it gets.

You won't get in trouble. Speaking of supply chain, what are you seeing today in terms of disruption coming out of Asia? How long do you expect this to last? And then I'll take piece of it.

So the supply side. So it used to be 60%, 6-0 percent of the goods, both that we sold, but also in just the categories we're in for the whole U.S. market, came out of China. Almost all of those goods were on the list of tariffs that started in 2018 at the 10%, and then went up to 25% in 2019. So suppliers for 1.5 years, creating alternative supply chains to basically try to get away from paying those tariffs because the tariff -- 25% tariff's pretty meaningful on these items. So that 60 dropped to like a little over 50, but it's still a pretty big percentage of the total. So what happens -- the suppliers in our categories typically turn goods about 4 times a year. So we typically have about 3 months of inventory on hand. And everyone knew Chinese New Year was in January, you'd have no production in. So you more or less have inventory that gets you to April, April or May. So the disruption that's happened over the course of February into March basically hits you more in the April, May time frame than now. So today, you don't have the supply disruption evident yet, but you know it's coming. What we've seen in the last few weeks is we've seen China production be ramping up at a pretty fast rate. So we think it's back up to around 70%. And we think over the next 4 to 6 weeks, it will get from 70% into the 90s somewhere in production. So we see an end to the production challenges. But what we think is the different suppliers at the supplier level and at the item level, there'll be shortages as you get to the spring into early summer. And it'll take another mid- to late summer before the supply side is back to normal. And so what we do, we don't buy inventory, but we sell a lot of products. So we generally do understand what suppliers have for inbound volumes. We generally know what's available, what's not available. So what we've been doing is making sure we really understand where suppliers' impacts are. And it's in a class of goods. So if in accent chairs, we think we're going to be overly impacted, well, let's figure out where to source additional supply from. But by and large, we have so much selection available because we have all these different suppliers with all of their inventory available. The vast majority of what will happen on our platform is if a given supplier runs out of goods, their volume will obviously suffer, but there's a significant amount of competitive goods that are offered by others that will effectively take the share. And we've seen that happen at different points in time.

And do you expect any bottlenecks or supply constraints, just from a lot of product that could be coming all at the same time. Might some product have to be airfreighted that normally would go on a boat. Are those...

Not -- yes. It's not super economical to airfreight these items, so I don't think you'll see a lot in airfreight. I think what you'll see in these categories, though, there's -- the ocean freight capacity, like a lot of shipping lanes reduced capacity because volumes reduced. But what we're starting to see is volumes are increasing, they are increasing capacity. So I think there'll be some crunches, but I think there'll be -- that will be relatively modest. It's more a question of you're ramping up production, but you have this lost period of time. So it takes a little while before you net or caught back up.

One of the features of Wayfair is it takes a very scientific and data-oriented approach to just about everything. So as you look through the patterns of data that you see on the demand side, has there been any observations? And if we go through a period of extended consumer hibernation, how does Wayfair position itself to benefit from that?

Yes. So it's interesting because, so obviously, in the U.S., you're seeing like in Seattle, Washington State, perhaps, be the most acute location in the U.S. where there's a school district in North of Seattle, which just announced a suspended school for 2 weeks. And you're starting to see some of this happen. In Europe, you've seen some of this start a little earlier. In Europe, we only operate in the U.K. and Germany. We see no demand hit there, and we've seen no demand hit so far in Seattle, Washington State. So that's not to say that under no circumstance could there be a demand hit. It's more to say that we haven't seen that yet. And you have all these puts and takes like if you're sequestered in your house, would you shop online more or less? If you're sequestered in your house and you're panicking, would you just shop in general more or less if you're -- by the time you play this out, if the local stores don't have inventory of as many items because the supplier they bought from doesn't have inventory, does that drive more online or not? So it's very hard to like figure how these things net. So we're kind of watching it. But so far, we've not seen that impact. And it's a little hard with any confidence to handicap what you think will happen next.

Does it change the potential trial for Wayfair? Do you -- can you do anything proactively to say if someone's going to be at home for 2 weeks and you are in the market for a new accent chair, this is -- we are as good of a solution for you as any, especially when you're not going to a store?

I mean fundamentally, if people are sequestered in their house more, they're going to spend more time online. It seems highly likely. Are they actively shopping? Or people -- what we saw in 2007 to 2010, at that time, we had about 75% exposure to furniture. Today, it's less than 40%. And furniture after travel is the second hardest hit category in times like acute financial crisis because, frankly, it's easy to postpone travel and it's easy to postpone buying furniture. And what have the overall U.S. market went from over $100 billion of retail and furniture, 3 years later, it was down to $70 billion. The contraction was on the -- is very close to the amount of contraction during the 3 years of the great depression. We actually saw demand dry up for just a few months, and then demand rose, and we actually grew meaningfully through that period. And part of it was people were interested in exploring online just because there's more values, they were willing to try it even though they hadn't before. And then we were able to show more value because suppliers had excess inventory, which they translated into lower pricing. And then our 2 biggest input costs after the product itself, are basically, I mentioned transportation. Well, transportation carriers are fixed cost networks. They lower price dramatically if they needed to cover off their baseload. And the third one is advertising, which is basically you're bidding against everyone else for advertisers when demand drops down if it gets cheaper. So we grew quite nicely through that period on good economics. This scenario is not identical to that, but I do think there's a bunch of scenarios you could posit where online could benefit, but you could, of course, create scenarios where online and, perhaps all retail, suffers.

Now if we were to take the other side of that with respect to stores, Wayfair's been testing a physical location. What have been the observations from that? Could you see a day where this becomes a bigger part of the way you go to market, having physical locations, because it just...

It's relatively easy to see it being a bigger part when we're doing $10 billion revenue, and we have one small store.

Yes. It could only go up.

It could only go up.

It could only go up.

On a relative basis, yes, I could see it being a much bigger part. But I think of it as a channel. Like, I mean, we have television channel. We have a huge direct mail channel. We have significant online channels. A physical brick-and-mortar channel could make sense, too. It would be part of the total mix. So I don't think it's sort of -- those customers would never shop online, the online customer would never go to the store. If you say things like the online customer may still appreciate a catalog, and the online customer, we may still be able to incrementally increase our share of wallet from the online customer by the way we run television ads that tell them about new categories, so on and so forth. So it will be part of that mix, and it would need to have that economic payback. We're still in the early days of figuring that out.

Yes. How quickly, once you do figure it out, would you pursue that opportunity, is a plethora of real estate available?

Well, I think the way we're thinking about it is, we really need to figure it out before we'd want multiple stores. And what you then do is you'd want to have multiple stores and prove out that it works before you'd really start scaling them. So this is a long-cycle activity.

Understood. And coming back to the profitability in the minute or 2 that we have left, Wayfair has noted that it can realize significant margin gains from a variety of areas, including emerging categories. What are the other big drivers of gross margin expansion from here? If you...

Yes, I rattled off a whole bunch on the earnings call. So obviously, new categories, as we scale and then the margin profile then gets a lot better. A lot of what we're doing with merchandising, creating our own collections of items -- we're not buying the inventory, we're just curating items in, but then we're creating the merchandising assets. We're finding that, that gives us more pricing power. On the operations side, when we talked about shipping, one of the things with the shipping, economics getting better and better, that increases your gross margin. Another thing is we're offering our suppliers more and more services. I mentioned the logistics network. If the suppliers pay us for use of that logistics network, and when they buy ocean freight from us, we make a margin on that. So these things show up in the form of gross margin because anything our suppliers buy from us is effectively accounted as contra COGS. And so it increases the gross margin. It just nets against what we pay the supplier. And so we have certain advertising products that are oriented to help them launch new items and be -- really take the time for a new item to succeed down from 9 months to 1 to 2 months, which really helps their cash flow dynamic and their business progress. We have the logistics services. We have other services we're building for suppliers that each would lift gross margins. So we have a series of things that all will accrue to positive gross margin.

And it seemed like, during the earlier part of our conversation, that you talked about some of the -- tightening some of the constraints, the return constraints on marketing as a way to improve the profitability. Has there been any similar pivots on the gross margin side to realize some of these opportunities in the foreseeable future than --

Yes, I'd say, significant. But the reason I'd say significant is like when I talked about going after inefficiency, inefficiency doesn't mean inefficiency in the marketing spend or inefficiency in headcount. It's really, in general, in the business, we went from a $5 billion business to a $10 billion business over 2 years. We had -- we went from 4,000 to 8,000 people in OpEx headcount, we went from the logistics network probably over doubled, it's 15 million square feet. I don't know exactly what it was 2 years ago, but let's for argument's sake, let's say, it was 6 million, 7 million square feet. When you ramp things that fast, it's unlikely that you did it all surgically. What this has done is give us enough pause to identify certain pockets are inefficient. You then sort of go look everywhere. And what you do is you do find opportunities everywhere. So you say, hey, let's take a deep breath, and on this next cycle, let's make sure we're not just focusing on scaling, but we're focused on scaling efficiently. Let's root out a bunch of what's gotten in there. We don't want it to become embedded. We want to unlock it, and that's what we're doing. And you could say this is either good news or bad news, depending on your perspective, but there's a tremendous amount of it.

I think it's good news. And we look forward to watching your progress against this opportunity. Please join me in thanking Niraj for a great conversation, and --

Thank you.

Thank you very much.