Good morning and welcome to Groupon's First Quarter 2019 Financial Results Conference Call. On the call today are our CEO, Rich Williams; and CFO, Mike Randolfi.

The following discussion and responses to your questions reflect management's views as of today, May 1, 2019 only, and will include forward-looking statements. Actual results may differ materially from those expressed or implied in our forward-looking statements.

Additional information about risks and other factors that could potentially impact our financial results is included in today's press release and our filings with the SEC, including our Form 10-K for the year ended December 31, 2018 and Form 10-Q for the quarter ended March 31, 2019. We encourage investors to use our Investor Relations website at investor.groupon.com as a way of easily finding information about the company.

Groupon promptly makes available on this website the reports that the company files or furnishes with the SEC, corporate governance information, our quarterly stockholder letter and select press releases and social media postings. On the call today, we will also discuss the following non-GAAP financial measures: adjusted EBITDA, FX-neutral results and free cash flow.

In our press release and our filings with the SEC, each of which is posted on our Investor Relations website, you will find additional disclosures regarding the non-GAAP measures, including reconciliation of these measures to the most comparable measures under U.S. GAAP.

As we discuss our results during the call, note that all comparisons, unless otherwise stated, refer to year-over-year growth as reported. All gross profit comparisons are FX-neutral with the exception of gross profit per customer, which are as reported.

And with that, I'm happy to turn the call over to Rich.

Thanks, Heather. And thanks to everyone for joining our quarterly conference call.

We look forward to discussing our results and taking your questions. Hopefully, everyone has had the opportunity to read our quarterly letter to stockholders, which we posted yesterday on our Investor Relations site.

It contains a thorough discussion of our first quarter results as well as our 2019 strategy. With that, I'll hand the call over to Mike for some additional color, and then we'll take some questions.

Thanks, Rich. Adjusted EBITDA in the first quarter of $47 million was better than our expectations for the quarter but down 11% compared to last year.

In Q1, we generated gross profit of $306 million, which was higher than anticipated, driven by North America customers shifting to higher average order value and higher-margin offerings. However, in North America, gross profit was $210 million, down $10 million or 4% year-over-year.

Q1 North America Local gross profit was $161 million, down $6 million or 3%. Q1 North America Goods gross profit was $33 million, down $4 million or 9%.

In Q1, North America net customers declined 940,000, which reflects the continued optimization of marketing spend toward customers that have a higher potential for long-term gross profit generation and ongoing headwinds from e-mail and SEO traffic. In total, gross profit per customer for Q1 on a trailing 12-month basis in North America was $29.72, up 5% versus the prior year period.

Turning to International. Gross profit was $96 million, down $1 million or 1%.

Q1 International Local gross profit was $69 million, up $4 million or 6%. International Goods gross profit was $19 million, down $3 million or 15% due to pricing and promotion strategies to stimulate demand in light of weakening consumer sentiment across Europe and particularly in the U.K.

These efforts helped hold International Goods units near flat. International customers declined slightly in the first quarter.

Q1 gross profit per International customer on a trailing 12-month basis was $24, down 3% as reported. In the first quarter, on a consolidated basis, marketing expense was $93 million, down $6 million or 6%.

In North America, Q1 marketing expense was down $12 million as we continue to refine our marketing spending toward high-value customers and less higher offline spending last year. In International, Q1 marketing was up $6 million or 21% as we continue to invest in the long-term potential of the International business, such as conducting city-specific marketing campaigns.

SG&A for the first quarter was $210 million, down $12 million or 5% due to our ongoing efficiency efforts. During the first quarter, we repurchased 4.4 million shares for $15 million, ending the quarter with $275 million remaining on our share repurchase authorization.

As of March 31, 2019, we had $646 million in cash and $250 million on our undrawn revolver, which we are currently in the process of refinancing and expect to do so on equal or better terms during the second quarter. Before we close, we want to remind you, we are focused on building Groupon for the long term while creating shareholder value.

We plan to continue to discuss annual guidance for adjusted EBITDA each quarter in order to emphasize the long-term drivers of the business. However, we do not intend to provide specific financial measures for the applicable quarter.

We'll still give commentary on trends that are relevant to our outlook and answer your questions in Q&A. For 2019, our continued expectation is for approximately $270 million in adjusted EBITDA.

With that, let's open the call for questions.

I just have a few, if I could. First, I was wondering if you could help us unpack some of the gross profit beat within North America.

What drove that average order value? And did any of the gross margin improvement, was that driven by the progress that you guys have been making in cardless?

Sure. So first, we're very pleased with the progress that we've continued to make this quarter with regard to our path to cardless.

I mean we've continued to increase the number of cards linked. This quarter, we launched our card-linked version of our classic Groupon offers, x for y.

It's live now in Chicago and New York and L.A. and San Francisco.

So we're very excited about the progress there. And we see customers responding well to that.

We've also launched our card-linked loyalty feature in our health, beauty and wellness offering. So very, very pleased with that.

And so really pleased with the progress and really pushing hard as we move into the second quarter and into the back part of the year. In terms of our gross profit beat above our expectations for North America, keep in mind, we continue to emphasize maximizing gross profit over an extended period of time.

And as part of that, we continue to work to optimize what to surface to a consumer in a way that maximizes gross profit and generates more gross profit on a per unit basis. Within that, we've continued to add over time high-quality, high-end inventory particularly on our health, beauty and wellness segment.

And that helps steer our customers towards higher average order values as well as higher margins. And so between that and us, through our relevance algorithms steering customers towards that which is going to maximize GP, but also that which is a good customer proposition for the consumer, that was -- that's what led to our beat versus our expectations in North America GP.

And if I could just follow up on that answer, which is very helpful. I think last quarter, you called out some sizable investments that you guys were making for card-linked and partnerships, particularly in the first half, which would pressure gross profit.

Did those investments materialize in the quarter? Or will there be -- or will they be more pronounced later in the year?

And then I just have one on International.

Sure. So those investments materialized and the headwinds associated with those investments materialized during the quarter.

As we move into the second quarter, we intend to continue to reinvest our outperformance. And what you're going to see us do is continue to push really hard on card linking and bookability, continue to increase the pace of experiments around mobile.

And with that, as you look to the second quarter in particular, what I would expect is that the trends you saw, whether it be revenue or gross profit in the first quarter, are similar trends in the second quarter. And with us continuing to step up our investment in card linking, in bookability, in our mobile experiments, I would expect that in Q2, gross profit would be somewhat lower in Q2 than in Q1.

And with that, I would expect adjusted EBITDA to be somewhat lower in Q2 than Q1. Now consistent with what we've discussed in our prior earnings call, we view the front part of this year as a period where we are accelerating very important strategic investments and ones that we believe build Groupon for the long run.

And with that, we expect that we'll begin further reaping the benefits of those investments as the year progresses, and we would expect trends to continue to build and continue to improve as the year progresses.

Great. And maybe just one quick one on International.

Do you still expect International customers to grow this year?

Well, what I would say is -- on International is, over the long run, we think that's a region that has tremendous potential. I mean I think about many of the markets that we're in, they're some of the largest economies in the world.

And you think about like the U.K., France, Italy, Germany, Japan. So I think there's the potential there for substantial growth, keeping in mind that today, the number of customers we have in those markets is roughly half what we have in North America despite the addressable customer base being roughly twice as large.

So I think there's definitely opportunity for a longer-run growth. In terms of the short run, I'd say as I look closer in, I'd expect the trends in Q1 to continue into Q2, particularly as a result of what we're seeing in the U.K.

where there's a softening consumer. If you exclude the U.K., we actually have growth in customers in our international region in the first quarter.

And I would expect that we would continue to see that in the second quarter. So it's just tough to tell in the short term how consumer sentiment influences the U.K.

in particular. But I'm still very optimistic about the potential to grow customers internationally over time.

Can you elaborate on the opportunity you see in these distribution partnerships? Specifically, I guess, what are you doing with Augeo?

How big is the universe of potential partners you see? And is the opportunity here more on the customer deals side?

And anything you can share about the economics?

Sure. So thanks for that.

And as I mentioned in my letter, it's a place where we're really excited about the opportunity. It is a new opportunity space for Groupon, and it's an opportunity space that's been enabled by our moves into voucherless.

And why I say that is that when you're working within a voucher, that voucher is effectively a gift card. And that gift card is really, in many ways, in a way, it's associated with Groupon directly.

And in the distribution space, you're -- I think in order to be effective, you have to have seamless products that can be branded the way the partner's distribution network wants to see them branded. And the voucher makes that really hard.

Things like card-linked inventory, in particular our card-linked supply, doesn't require that. It can be completely white-labeled, and it can be used as a tool by a partner to increase their own value prop.

And we see a lot of demand for that kind of content that's hyperlocal, that's value-oriented and savings-oriented in a way that can be completely white-labeled and completely seamless for a consumer to use. So we see that addressable market to be really big.

If you just think about all of the space that people typically play in with fee-based services to add value to their either membership products, their card products, et cetera, that space to me is wide open for us now, that we're starting to build real voucherless inventory scale. Augeo is an example.

It's a platform that administers card-linked programs for U.S. banks and quite a bit of -- quite a few U.S.

banks. So we'll be an inventory partner for them.

We'll distributing our card-linked inventory through their network. And that's just, I think, the first step for us in what can be a much bigger piece of our business over time.

But again, this is our first distribution deal that we've signed. I would expect more of them to come as we go throughout the year.

And we're going to have to learn how big it can really be. At this point, it's still too early to put a marker on that.

But again, I think it's -- there's no shortage of opportunity for us because there's no shortage of other brands that are looking for ways to increase their own strength of value proposition. And we're now a seamless partner in being able to do that in a really potent way.

So what does that mean for us in a bunch of ways? It's your question.

Is it customer? Is it deal?

It's both. For us, it's a way for us to reach much bigger populations that are not dependent on our e-mail channels and not dependent on things like search engine optimization.

So I think broadening our reach and horizons on that side is increasingly powerful. It also matters to our merchants.

If you think about the number one dissatisfaction point for our merchants, even though our merchant satisfaction is very high, the #1 reason merchants aren't happy is they want to sell more units. And this is an opportunity for us to increase distribution and reach, not only for us and the consumer front, but for merchants to sell more units, to get more customers through their front doors and to help them make more successful.

So we see that as an opportunity on both sides of our marketplace. And as far as the deal economics, I would generally think of them looking pretty similar to our third-party marketplace business and local in general and well within the kind of norm for distribution in the online commerce space period.

So I can't give you specifics there, but I'd say there's nothing we're doing that's outside of the ordinary.

A couple of questions here. So Rich, you indicated that you reduced the spend in offline marketing.

Does that not curtail your efforts to kind of refresh the Groupon brand? Or is this kind of an initiative where, "Let's get the product in a much nicer initiative in place before we kind of restart that campaign"?

Secondly, if you can talk about the new desktop and app experience. When do we get to see that?

And I'm specifically interested in kind of the new experience which optimizes the online real estate. And last question is, can you give us any insight into the uptake of Select?

Sure. So there's a lot there, Sameet.

Thanks for those questions. Our marketing spend, as we've discussed it in the past, all of it is tied with the same fundamental philosophy of return on our marketing investment in terms of incremental gross profit generation over time.

So our 12- to 18-month ROI thresholds apply to offline as well as every other really category and channel we work in. So when we move some things down or move some things up, which happens constantly in our spend, it's largely because of performance.

Now in the case of offline here, recently, we've been doing some 2 things really specifically. One is some market mix model-testing that's really based on improving the total performance and productivity from our market investment.

And the second piece is exactly what you mentioned, which we really need to continue to scale the product and merchandising pieces before we start to lean really heavily into more brand reeducation, at least more so than what we're doing. And you can see that in our offline campaign.

They're still very much rooted in savings in the classic Groupon with a little sprinkling of new product experience. That one is going to -- you're going to see us continue to spend and optimize along those campaigns.

They work well. But before we really lean hard into the rest of it, we need to see broader reach and exposure of our core new experiences around bookability, card-linked, voucherless, et cetera.

So we have some work to do on that. It's why we're leaning so hard into accelerating those in the first half in particular.

And I would expect you're going to see more of our marketing start to tilt towards education and brand development as we move into the second half. As far as the desktop and app experiences, when you should see those, there's like one -- so depending on what browser you're going to use and if you clear your cookies, you may -- if you go to mobile web as an example, you may already start to see some pretty dramatic version -- changes to versions of the mobile web experience.

And you're seeing everything from automatic rotating slots to horizontal scroll and a lot of different merchandising slots coming in just to improve search and browse discovery. So we have pretty significant changes that are already rotating in market and they're live.

Just whether you're in a test-sell or not, it's hard to say. But you're starting to see those already.

But what you should think about there is that we don't -- unless we -- one of those pieces that we're testing is just an outright winner, I would not expect a big bang change. In general, I think you've seen historically, when brands that have significant consumer populations decide to make big bang changes, even if it's generally good and for the right reasons, it takes consumers a while to get used to it.

And those can be pretty cold starts for consumers that are well trained in your existing UI. Instead, I would expect us not to do that.

I wouldn't expect us not to do the big bang but more to feather in our winning elements that we see consumers are interacting really productively with. And you're already seeing some of that happen in our core app.

We have new filters. We have additional filter metrics and a bunch of other things that are making it easier to -- for consumers to find what they're looking for.

But we're feathering those in as they win and as we go. And I think that's just a healthier way for us to move.

Unless we see just something that blow us away and consumers adopt it really aggressively and immediately. But it's just not our experience there historically, especially not at scale.

As far as Groupon Select, it's not something that we've discussed broadly, and I think that's just because it's very much in an experimental stage. It's very early.

There's still a lot of folks who haven't been exposed to it, haven't seen it on the site or in our apps. We're very excited about that program.

I think it has a lot of potential. But again, it's a membership-oriented program.

So it's a little too early to start discussing what the potential of it is ultimately, financially, for the company and for customer accounts because as you know, in membership, it's really about continuity. And that's where it counts, not just how many folks you can get in the program initially.

So I'd say, stay tuned on that one. We're watching it really closely.

We are excited about it as an early-stage initiative, but we have a lot of wood to chop on that one.

Rich, could you talk about the process of migrating through the SEO and e-mail headwinds? Did these remaine roughly constant from the fourth quarter?

And curious if they're, in any way, tied to customer losses in North America. And then, Mike, maybe just a quick one for you.

Nice gross profit and EBITDA outperformance. It sounded from your comments like International was maybe a little more challenging than expected at the start of the year.

Is this being more than offset by a better consumer in North America?

Great. Thanks, Tom.

So on migrating through e-mail and SEO, I -- when we discussed the full year last February or this past February, I said we -- generally, as we look forward, we really don't expect our challenges there to get significantly worse, but we expect them to be a real challenge as we move throughout the year. But we're not just accepting that.

We don't sit aside and say, "Oh well, e-mail is hard and SEO is hard." Our teams -- we have dedicated teams obviously for each.

And those teams are heads down trying to do their best to optimize the experience for our customers, particularly around e-mail where we control that channel really from front to back. So we have teams there.

And you can probably -- hopefully, you can see some of the changes that the team has made there. We're doing a lot to improve the relevance and even merchandising in our e-mail.

We're changing a ton in that space, and we're seeing some really nice early traction on some of our experiments there. And we won't stop.

So we're not just going to sit idly by and watch any of that happen. SEO is a little different.

Obviously, there's a bunch outside our control there with algo changes and UI changes, particularly in Google. But in the same way, our team is focused on really the core bits in SEO that ultimately have proven over time to be the most important when it comes to your position on the page.

And that's really the quality of our content and the amount of content we're generating, both from our customers and merchants as well as our internal teams. So a lot of our effort right now is around content development, and of course, all the other kinds of tweaks and tunes that you make to the core of how our sites are displayed to search engines.

And we have a lot happening there and we'll continue to tick away at that. Now historically the challenge at SEO is that it is -- it's not -- there's no quick fixes there and it is a shifting sands environment.

But we're focused on it. It is a big piece for us and our teams are cranking away.

With respect to is it tied to customer losses, the answer is yes. They are absolutely -- we feel an impact from those channels.

They are very big channels. Now it's important to remember that if -- they are not our biggest channels.

Our biggest single channel of traffic at Groupon is people who come direct to our app and site. So we have a built-in buffer and we have a lot of other successful channels and we have other initiatives like distribution to expand our reach and to broaden our traffic opportunities.

But we for certain feel that in the relative short term when you see channel compression. It's -- given the size of both of them, call it, e-mail around 20% and search being a double-digit channel for us, you're going to feel some of that as that has traffic headwinds in terms of customers and how they engage.

But the challenge is not on us to overcome those through new channel development, which we've done a great job on historically, and just making our existing channel that we control an awful lot better. So that work is ongoing.

But sure, we'll feel it. We'll continue to feel it, as we've said, probably throughout the course of this year.

But we expect it to start to temper as we move into 2020. And I think the opportunity is for us to accelerate that.

And the one thing I would just add on to that is when we talk about our commentary for $270 million of adjusted EBITDA this year and $300 million or more for 2020, that contemplates those headwinds. We're not assuming necessarily they go away.

We don't -- as Rich said, we don't accept them. But we do, in terms of our forward outlook, we are contemplating that those headwinds will continue as we go into next year.

So Tom, with regard to your question on International versus North America. First, I'd just highlight in International, on an FX-neutral basis, Local still grew by 6%.

So we -- while we certainly would have liked even faster growth than that, we definitely were pleased with that. What we saw on the International side is obviously a slower customer trajectory that we would have hoped, particularly in the U.K., as I highlighted earlier.

If you excluded the U.K., the rest of International actually grew. And what we saw was, in the face of a softer consumer, we chose to lean into that with pricing and promotion.

And the result was obviously lower gross profit, but we maintained the units that we would have had similar to last year. And with regard to North America, we have the benefit right now of having 30 million customers.

And what we saw in North America is that they responded well to the continued offerings we have and our continued improvements in supply and our ability to surface what's relevant to them in a way that maximizes gross profit for us. So kind of what I was touching on earlier is that we continue to see our customers shift to higher AOV, higher-margin categories, and that was the source of the benefit in the first quarter.

So how should we think about the relative purchase behavior of consumers who have downloaded your app versus those who haven't? And if you could update us on your mobile strategy as far as making adjustments to maximize mobile sales.

Sure. Thanks for that, Elliot.

The -- in general, if you downloaded our app and are buying in our app, you're a better customer. You buy more frequently and you engage more frequently in general.

And we discussed that a bit in the past as we've talked about our mobile transition. I think the important thing to remember here now is we're far beyond mobile most.

I mean we're, call it, 75% to 80% mobile any way you slice it, from traffic or transactions and everything in between. But that has come at some expense as really desktop to mobile web has been a challenging migration, not just for us but a lot of others as mobile web just generally doesn't convert as well as desktop.

But our mobile app users are now -- we're north of 200 million downloads in that mobile app. They really are our best users, our best customers.

And frankly, I think because it's our best experience. It's the best way to consume the content.

It's most contextually relevant. For most people, local is when they're out and about.

And that's when mobile works really, really well. So I think there's a lot of opportunity there for us to continue to do more.

And how to think about that for us in the future, again, as I mentioned in the past, like we're really aggressively testing new features, new ways to display our content and really build the app for what Groupon is becoming. And specifically with the size of our catalog, the current app is not built to handle the volume of offers we plan to bring onto the platform over the course of the next 4 to 6 quarters.

So I would expect some really significant changes. And again, I would point to play around a bit with mobile web and our mobile web experience because you're seeing some different things there and different ways to display content that is much more search/browse-oriented and broad catalog-oriented.

And that's a fast way to test new versions of app features as well. So I wouldn't expect, if you see something that's amazing in mobile web, that you wouldn't see it on app as well.

So it's just a faster testing ground. So look to that a little bit as an indicator of how we're thinking about the challenge and opportunity there and expect us to step that up, if anything, over the course of the next 1 to 2 quarters.

So first, just a couple of quick follow-ups. So on the overperformance in the quarter and leveraging that into further investments, 2Q and the rest of the year.

I know you gave a little bit of color around what to expect in Q2, but should we be thinking about the incremental investment and the maintaining of the full year EBITDA guidance as all that coming through in Q2? Or a good portion of it coming in Q2 but still seeing investment come through the rest of the year as well?

And then the second follow-up, on the distribution platform. It sounds really interesting.

Just trying to get a feel of what percentage of your total distribution do you think that could become over time. And then I'll leave one more follow-up for after.

Sure. So I'll start and then Rich could tag on.

So with regard to essentially the phasing of the year, as we talked about in the last earnings call, we'd expect the -- any of the net impact from investment to be biggest in the beginning of the year. And then as we get to the back part of the year, we're obviously investing but you obviously then start to reap some of the benefits of those investments.

And so as we've indicated, we would expect trends to generally improve as the back part of the year. And what we've talked about is while for the year we would expect gross profit to be down slightly, what we would expect is that GP per customer would continue to benefit as the year progresses, and that would largely offset a lower customer count.

And the GP per customer improvements are stemming from improvements we're making and our initiatives that we've talked about in terms of improving our product, our customer experience, what we surface to customers, customer segmentation. And therefore -- so that's kind of how we think about it playing out for the year.

And then, Ygal, on distribution, it's hard for me to say what percent of the total distribution it could become over time. But I suppose I think about it in maybe a too simple a way.

But we're roughly, call it -- in a place like the U.S., roughly 20% of the Internet comes to Groupon over the course of any given month. But that leaves a really big world out there that we're not tapping into.

And even if you look at our first partner in that space, of Augeo working with bank programs, there are hundreds of millions of cardholders in the U.S. And we have around 30 million active customers in the U.S.

So it's just -- it's so much bigger. So in terms of pure reach, it's not hard to imagine a world where distribution will -- for us, could reach far more customers or far more households than we do on a standalone basis.

The -- obviously, the economics of that are different, but it's not hard to see how over the course of a relatively short period of time, over the course of a couple of years, where distribution reach could be really significant off of this platform with just a handful of strategic partnerships that have real big populations. But we -- again, we still have of a lot of work ahead of us there.

The good news for us is there's a lot of demand for our content. And I think over the course of this year, specifically in the back half of the year, we'll be able to provide a lot more color on a go-forward basis of just how big we think it can be.

Great. Very helpful.

And then maybe just a question on bookable. So you noted in the shareholder letter that you're seeing in some countries the majority of your food and drink offers being bookable.

I was wondering if you could share what countries those are. And then you also note that you're moving into other categories there.

So what -- if you could share what some of those other categories that you think about moving into beyond food and drink as bookable. And then lastly, if you could share what your total inventory -- what percentage of your total inventory, whether it's North America or global, is bookable now?

Sure. Yes.

Thanks for that, Ygal. So we're really happy with how bookability had been trending on the platform and not just in terms of consumer response to it.

And response has been great and not -- and I think I mentioned it, it's not just because people are using it. It's that once they use it, they come back more frequently.

They're better customers for Groupon. And it's working better for merchants as well.

If anything, as we've rolled out booking with the merchant population that we're serving, particularly in International, we've actually seen our close rates increase because it's easier to actually work with Groupon. The onboarding load, the training load, all of it is easier when you're in a bookable environment.

So we feel really good about where we're trending on the booking side. We haven't shared a specific number of where we are with bookable inventory specifically.

But I'll tell you it's in the double digits and it's -- and we're making really fast progress there in total percent of our inventory. You're -- we mentioned that the majority of our food and drink in a number of countries, places like -- where we started really with bookings: France, U.K., Italy, is in that majority space now and growing.

And based on that feedback from merchants and consumers, you should first expect us to move bookability into the health and beauty space, where we're really strong and we're strong in every market that we operate in. And then you should see us move it into places like things to do and home services and auto services.

But we plan to -- this is a space where if it's not card-linked, it's going to be bookable. So you should be thinking about our entire inventory but facing more in that way.

Food and drink in Europe to start, then health and beauty, then things to do and home and auto and then really bringing that to North America and all those categories to the back half of the year. So that -- we are working aggressively toward our goal to make Groupon voucherless.

And that voucherless path is paid via booking and card-linked.

First one for Rich. You've been tweaking the marketing programs to focus on the higher-value customers for a few quarters now.

And obviously, the customer preferred that change. Can you provide some color on the churn rates and repeat purchase behavior you're seeing since some of these new cohorts of customers?

Is it any different at all compared to some of your traditional core customers? And then one for Mike.

It seems like the efficiency gains on SG&A, maybe in North America, is starting to tail off a little bit. How much of efficiency gains do we need to see from SG&A to kind of achieve the $300 million EBITDA for next year?

Thanks, Deepak. I'll start and then hand it over to Mike.

But yes, we have been tweaking for a while and tuning around quality. And it's -- for sure, it's easier to see that in terms of GP per customer.

And that's up by 5% so -- year-over-year in Q1 on a trailing 12-month basis. So you're seeing the results on that side and you're seeing our focus on GP per customer shine through there.

And what I'd say on our -- we don't give specifics on our cohorts other than to say they're trending in line with our ROI threshold. And in this case, I'd say they are trending to higher gross profit production over time.

So we feel good about the programs that we've put in place. As much as we don't like to lose customers, we also don't see it as a good use of our precious resources to invest in unprofitable customer relationships.

And our belief is that ultimately, a lot of these customers, when our product is at a place where it's more seamless and really a frictionless and rewarding experience, it's going to get a lot easier to bring a lot of these folks into the fold and a lot more profitable over time to bring folks into the fold. And based on what we see on our programs and voucherless in particular and its ability to increase purchase frequency over time, whether that's booking or card-linked, both of them impact purchase frequency in a positive way.

We'd expect that to continue to improve our gross profit profile per customer, which is going to give us more room to invest within our ROI threshold. So it's working as expected.

As much as we don't like to lose the customers, it's working as expected. And we believe, long term, it's the right call.

Yes. And on SG&A, so a couple of things I would highlight is, first, as you're looking at -- it sounds like you're looking at SG&A by region.

One of the things you have to keep into account, particularly around headcount, the way it's disclosed, the headcount, the way it's disclosed is based upon where the headcount is located, not necessarily the region it serves. So we have a substantial portion of our headcount that is located internationally that serves our North America business.

Second thing I would just highlight is if we look at this quarter, we ended the quarter with approximately 6,300 heads. And that's the lowest headcount we've had since Rich and I have been here.

And what I'd highlight is that was down 5% year-over-year and actually one of the biggest declines we've had in several quarters. The third thing I would just highlight is we continue to improve our process, improve our technology and continue to work on increasing the overall level of efficiency.

And what I would say is, over time, I think there's a lot of potential for that. So -- and what I would say is as we gain efficiency, that doesn't necessarily mean it manifests itself in lower SG&A.

We may choose to reinvest some of that efficiency -- some of those efficiency gains to support investment in our product, investment in our customers, investment in our merchants. With regard to 2020 and what's contemplated there, we do think about continued efficiency gains.

We also think about continued investment in our product. But what I would highlight is the bigger assumption there is and our thought is that we expect our customer trends to moderate in 2020, and we expect to continue to benefit in terms of GP per customer from the improved product, improved segmentation and the fulfillment of our strategy around card linking, bookability and improved mobile experience.

And so that higher GP per customer on a moderating customer decline -- on moderating declines, we think, ultimately aids and supports adjusted EBITDA of $300 million or more in 2020.