Tripadvisor Inc (TRIP) 2017 Q1 法說會逐字稿

Good morning, and welcome to TripAdvisor's First Quarter 2017 Earnings Conference Call.

As a reminder, today's conference is being recorded.

At this time, I would like to turn the conference over to TripAdvisor's Vice President of Investor Relations, Mr. Will Lyons.

Please go ahead.

Thanks, Brian.

Good morning, everyone and welcome to our first quarter earnings conference call.

Joining me today are Steve Kaufer, our CEO; and Ernst Teunissen, our CFO.

Last night, after market close, we distributed and filed our Q1 earnings release.

We filed our 10-Q, and we made available our prepared remarks on our Investor Relations website located at ir.tripadvisor.com.

In the release, you'll find reconciliations of non-GAAP financial measures to the most comparable GAAP financial measures discussed on this call.

Also our IR site contains a supplemental file information document, which includes certain non-GAAP financial measures discussed on this call as well as other performance metrics.

Instead of reading our prepared remarks on this call, Steve will jump into a couple of thoughts.

And then we'll jump right into Q&A.

Before we begin, I'd like to remind you that this call may contain estimates and other forward-looking statements that represent the company's view as of today, May 10, 2017.

TripAdvisor disclaims any obligation to update these statements to reflect future events or circumstances.

Please refer to our earnings release and our filings with the SEC for information concerning factors that could cause actual results to differ materially from those expressed or implied by such statements.

And with that, I'll pass the call over to Steve.

Thank you, Will, and good morning, everyone.

Thank you for joining the call.

Hopefully, you've had a chance to read our prepared remarks that we published last night after market close.

I'll summarize by saying 2017 is off to a productive start.

In Hotels, we're rapidly aligning products, supply and marketing as we drive towards long-term growth and profitability.

In a few weeks, we will roll out our streamlined hotel shopping experience, and soon thereafter, we'll launch our new brand advertising campaign, leveraging our strong global brand to build user awareness of TripAdvisor as a great place to find the lowest prices when you are ready to book.

In Non-Hotel, our continued strong demand growth and supply investments are really starting to payoff.

And while focus remains on top line growth, this segment is already beginning to exhibit attractive profit potential.

We are addressing large growth opportunities ahead and making great progress building end-to-end products that travelers love, a bigger and better platform for advertisers and a bigger and better business over the long term.

Ernst?

Thanks, Steve, and good morning, everyone.

We saw a strong growth recovery as the first quarter consolidated total revenue growth improved to 6% year-over-year or 7% in constant currency.

This revenue growth also improved from our fourth quarter results, driven by improvements in our TripAdvisor-branded click-based and transaction business.

We are encouraged by the significant recovery we saw in this click-based and transaction revenue growth over the past few quarters and the continued strong growth of our Non-Hotel segment.

Notably, we reiterate our revenue and adjusted EBITDA outlook, now inclusive of our brand advertising investment.

We are able to maintain this outlook by reallocating some less efficient online marketing dollars as well as expected profit favorability in our Non-Hotel segment towards our brand advertising.

With that, we'll now open the call for your questions.

(Operator Instructions) Our first question will come from the line of Ross Sandler with Barclays.

Steve, I guess the first question is just could we get a little more color on what's driving the improvement in revenue per shopper in the U.S. market?

Is it coverage?

Is it better competition in the auction?

Any color there would be helpful.

And then can you also talk about overall engagement?

As you've been making these changes to the mobile experience and as you see hotel shopper growth pick back up, what's happening with the engagement?

Meaning, is the session frequency or visits per hotel shopper also going up?

Any color there would be helpful.

Sure.

So with regard to our revenue per shopper growth in the U.S., the biggest factor is certainly going to be the lapping of Instant Book.

We had made a big bet in that category, built a really nice product.

But it did have some dilutive effects.

So now we're at a really good spot where we have turned the team towards the regular optimization efforts that we normally go through, and we're seeing results.

Simultaneously, as we indicated, we're working on our new shopping experience.

And any time that we focus on delivering a better overall experience, we would look to one of the measurements being an increased revenue per shopper.

When it comes to mobile, we've certainly seen meaningful growth aligned with the industry as users tip over to be using mobile apps and web.

And we've seen, on our own side, an improved revenue per session from users using our mobile app.

I don't have specific numbers I can share in terms of increased engagement to your specific question, but suffice to say, more and more users, more and more of the time, are using our apps for not only hotel shopping, but the full trip.

Great.

And if I can ask one more follow-up for Ernst on the Non-Hotel segment.

So you mentioned in the prepared remarks that you expect positive EBITDA for the 2017.

So can you talk about, where that margin might be exiting '17?

And remind us again what the longer-term margin profile for Non-Hotel looks like.

Yes.

We've always described this as a multi-year process in which we would initially invest in the Non-Hotel segment and then gradually turn it to profitability, as we saw more scale in the business.

We've been focused on building scale.

This is the year where we're seeing some of that building scale come to fruition in terms of leverage on the bottom line.

And as we say, we expect that this Non-Hotel segment, in aggregate, to be EBITDA positive, which, we think, is a great midpoint in our journey.

As we go forward, we expect our margins in the longer run to be robust and attractive.

We're not managing to a particular margin.

We continue to see this business as a growth and investment business.

But this year, we are going to see that profitability come through, and a little faster than we thought three months ago.

The first quarter has seen very robust results on the top line versus our expectations, but also on the bottom line versus our expectation.

We're seeing some efficiencies come through in this business.

So positive EBITDA for the year and a strong outlook for the future.

To add just a tiny bit of additional color to that, we're pleased to note that in our Attractions business, specifically, the TripAdvisor as a demand channel is growing faster than we had anticipated.

And that was, of course, one of the big pieces of synergy that we were looking for when picking up the Viator business and all that supply, how do we better monetize all of the demand that's already sitting on TripAdvisor.

And so on the Trip platform, demand continues to grow quite quickly.

And we're able to merchandise it better, having all the great Viator supply at our fingertips.

Our next question will come from the line of Eric Sheridan with UBS.

Maybe 2 also around the marketing theme.

In decommissioning inefficient spend and redeploying that money into brand advertising, I'm curious what you think that does for return on your marketing dollars long term for the business.

That will be number one.

Number two, as you do decommission inefficient spend, redeploying it into the brand, anything we should be aware of in the quarter-to-quarter cadence of marketing that might buck some of the usual seasonality trends on marketing?

Thank you, Eric.

In terms of the long-term efficiency profile of our paid marketing spend, we expect from TV to create much more lifetime value of users that we acquire and influence from TV than we are going to be able to get with online marketing.

And so with our online marketing, we have managed to, pretty short term, roughly breakeven as a channel.

The ongoing benefit is more limited and difficult to prove.

With TV, with a real shift of brand perception as TripAdvisor as a place just to use for planning, to use for planning plus price comparison and booking, we can get much more lasting impact on our user group.

And therefore, over time, that should actually benefit our marketing efficiency profile.

In terms of the seasonality of marketing spend cadence, not so much.

I mean, the marketing will still be aligned very much to the seasonality of the business, which doesn't change by doing advertising.

Our next question will come from the line of Mark May with Citi.

I have 2 questions, somewhat related.

They have to do with the near-term risks.

You're going to be reallocating a pretty significant amount of your online marketing spend.

How confident are you that you've had your head around the negative impact that that's going to have to the top line, at least in the near term?

And then you'll be changing your UI, it sounds like, as well.

How confident are you around modeling the short-term potential negative impacts?

And I guess, the question there would be also, what do you see as some of the near-term potential negative impacts from the UI change?

Okay.

This is Ernst.

I'll take the first question on the marketing.

And Steve will take the product question.

It is true that if we reallocate marketing dollars from online to TV is you get a positive impact in revenue on TV and potentially, a negative impact on online.

We have done quite a bit of work over the last 3 months in studying the different markets that we're going to be going with, with TV, studying our historical patterns of impact of TV that we have found.

And although there is some volatility around the results there, we feel good about the projections that we're making and good about reiterating this double-digit growth this year on our transaction-based business for TripAdvisor.

Steve?

Yes.

And with regard to the user interface, we're fortunate that this is a type of change that's mostly internal to TripAdvisor.

It's not requiring a lot of assistance or behavior change on the part of our clients.

So we're in control over the testing methodology, the rollout.

You can see some of the changes already on the site and some of the changes already on the app, depending if you get lucky enough to see it.

But all of that goes into our testing and how we're able to gain confidence that is, in fact, revenue neutral or revenue positive as we move to roll out.

I'm not saying we wouldn't roll out completely if we were up or down by a little bit.

But we do have a lot of capabilities and are experiencing this day to day in tuning and tweaking the design, so that it delivers on not only a better experience, but the revenue expectations that we have in our plan.

Our next question will come from the line of Lloyd Walmsley with Deutsche Bank.

In the script, you guys mentioned that absolute U.S. revenue per hotel shopper has fully recovered to levels to prior to Instant Book launch.

It looked like, by our math, that the desktop rev per shopper is still about 20% off the highs.

Is the delta just international?

And is that following a similar path to the U.S. in terms of improvement?

And then just following up on some of the TV questions.

It's always a challenge to change user behavior.

And you guys have been working at it on the IP side.

I guess, what is it about the new campaign and/or the new page that makes you feel more confident you can really get users to come back closer to time of purchase?

Is there anything just from the test you've seen in the new page design that gives you that incremental confidence?

Lloyd, I'll take the first one.

And Steve will take the second.

The comment about RPS being back to 2014 levels pertain to the U.S. And so we have seen that into -- in the U.S. Internationally, we have not yet seen that.

And so that was a clarification I wanted to make on that point.

And in terms of TV changing behavior, there are many examples out there where a brand in a category comes up with, you could call it, line extension, if you want.

So we all think back to Amazon that started at books and now did books and music -- and books, music, and of course, now they do everything.

And they were able to do that because they had positioned themselves as a shopping site.

Well, TripAdvisor is a travel site.

You come to plan your trip.

So we have the credibility to be able to play in a number of different aspects in that travel planning.

But I couldn't go -- we wouldn't be able to sell you a pair of shoes, that's kind of not what people associate TripAdvisor.

So we come from a strength, reviews, as Amazon had come from the strength books, but it's well within what users are using the site for to extend it to something else, and in our case, moving into that price comparison.

So we have a large portion of our current audience and have, for the past decade, come to TripAdvisor, read reviews and move into our price comparison engine and then click off to our partners or use Instant Book to book.

So that's an already established behavior pattern for a large portion of our users.

Our challenge isn't, "Hey, is that even possible?

Will users do that sort of thing?" They do.

It's a huge business for us.

The question is, can the TV campaign accelerate the move of the people who don't currently use us for that to join the crowd that does?

And a lot of people, when we ask, will say, "Well, yes.

I use it for reviews.

I didn't know that you had price comparison." And of course, I hear that and I think how could you miss that on our site.

Oh my goodness, it's plastered all over the place.

But we have to accept what users are telling us.

When we go on TV and we talk very clearly about price comparison: find the lowest price, help save you some money as a traveler.

And as we repeat that TV ad numerous times, we hope to educate that individual that the site that they already know and love can now be used for this new part of the travel -- the hotel shopping experience, which they're already on TripAdvisor for.

So we don't view it as a big leap or a big change.

And when we look at our last TV campaign, it was focused less around the benefit of comparing prices and more teaching users specifically that they can now book on TripAdvisor.

So we've tweaked that message.

We'll be back out to the marketplace at a heavier weight with a more streamlined user experience that's focused around that hotel shopping.

Think of it as less distractions on the page, more pointed on how can we help you save some money in your shopping experience, how can we give you the best value for the hotel that you're looking for, not just the highest rated, but the best rated in your price point, those sorts of things.

And we're giving ourselves time to make that message sink in through things like TV and in our other online campaigns.

So I view our past TV as very instructive, useful lessons to get us to where we are now, but we have a very compelling plan to achieve these objectives that we have in front of us now.

Our next question will come from the line of Justin Patterson with Raymond James.

Steve, I know, in the past, you've often referred to TripAdvisor as being indifferent between meta and Instant Booking.

With the new hotel shopping experience rolling out, could you just characterize what, if anything, has changed within there?

And then secondly, one for Ernst.

On click-based transaction revenue, you mentioned that international growth rates are still significantly lagging due to the constant currency.

You obviously had a strong quarter with the U.S. But how should we think about the cadence of improvement in that international growth rate?

And is there any reason that, that should improve at a different rate than the U.S.?

Great.

Justin, I'll take the first part.

So we look at meta and IB as two different methods to allow the user to find and book the hotel that they want.

They're on our site.

They're getting the best value that they can in the sort, they're looking at the photos, reading the reviews, building the confidence that they're getting the right hotel.

Now the next thing the consumer absolutely needs is to get the best deal on that hotel.

And so part of the new shopping experience helps pull people down that funnel of putting in your dates and actually getting to see our prices and noting that we've scoured the web and found you a great price.

The thing that's not new in this quarter or in this new design is where Instant Book fits into that mix.

We have been talking for several quarters about Instant Booking earning its position in our hotel shopping experience, so that we put it in front of users when we think they're most likely use it, either for price, because Instant Book might have the best price, or convenience, particularly on the phone, if it's something you've used before.

So we want to build that great shopping experience.

Instant Booking is a clear tool to help us do that and is a way that we can get better pricing in front of our consumers and make it more convenient, especially on the phone.

But as we look forward in 2017, '18, as some people commented, "Well, Instant Book is not as prominent as it once was." That's because our customers are telling us they're indifferent as to whether they are choosing to use it versus something else, and that's okay with us, hence, the comment on indifferent.

We want to do what's best for our travelers.

And we believe that Instant Book plays a key role in that, though not as a big a role as we had anticipated a couple of years ago.

And then just on your question about the click-based revenue and the international component of that.

So the U.S. is roughly half of our revenues and the other half is outside of the U.S. The U.K. is our largest market, but it's a long tail of many, many markets internationally.

Part of the differential in growth rate is attributable to timing in IB and lapping of IB, but that's not the full story.

Currency is one of the factors.

If you compare, for instance, the pound, the U.K. being our largest non-U.

S. market a year ago versus today, and then the mobile shift has tended to happen faster outside of the U.S. than in the U.S. And then there are many market-by-market differences.

But generally, the overall market, the robustness of our auction, the level of competition is more favorable in the U.S. than in many places outside of the U.S. And so our anticipation is that the U.S. will remain the pace car in terms of growth in the foreseeable future.

We believe, that in the long term, international markets can be robust growers for us, too.

But the pattern of growth recovery is likely to be a little slower than we've seen in the U.S., and therefore, lag the U.S. more than just the lagging of the IB rollout.

Our next question will come from the line of Mike Olson with Piper Jaffray.

I have 2 questions, please.

First, how much of the ability to maintain your EBITDA guidance for the year is a result of the reallocation of marketing funds versus the better profitability in the Non-Hotel segment?

I'd assume it's nearly all the reallocation of marketing dollars.

But first of all, is that correct?

And then second, I guess, following the previous question, assuming IB continues to be less of a focus, how do you envision navigating the challenge of the handoff problem for mobile that IB was meant to help?

Or is that handoff risk maybe just not as a significant of an issue as you previously perceived it to be as you see the mobile mix grow?

Thank you, Mike.

We're not breaking down in more detail where the EBITDA favorability comes from to offset the TV unfavorability.

But we called out these two factors for a reason.

They are both meaningful contributors to that.

And so it's not only one or the other.

We called these two out because they're both relevant.

And with respect to Instant Book on the phone, if you look back, I keep talking about building a great shopping experience for hotel shoppers on the phone.

And there is a click-off model that works quite well for us and others, but there's certainly a wonderful convenience of being able to essentially buy it with a single click, as sites like Amazon have shown.

We don't see a reason why we won't be able to get there in the hotel space.

And to do that, you need something that kind of looks like Instant Book or something pretty close.

So if we fast forward to the time when Instant Book really does have the best pricing available on the web, it's matches the lowest price anywhere, there's going to be, in our prediction, a set of travelers who love the TripAdvisor hotel shopping experience, have put in their credit cards and have now formed a habit where they're a frequent repeat user of our app because it's just simpler and you always get the lowest price.

We're not forcing that on people, to be clear, right now.

But we see that as a wonderful endgame as the world moves to mobile.

And if you're using us for price comparison, there isn't an obvious benefit of clicking out versus continuing your purchase on our site, unless someone else gives you the better price.

And that's the challenge for the Instant Book team, is always making sure that piece of the product has the best price.

When we do that, we'll earn more than our fair share for convenience.

Where we don't, our business funnel says we'll help the customer save the dollar by clicking them off to whoever does have the lowest price.

Our next question will come from the line of Perry Gold with MoffettNathanson.

Can you provide any more color on the recently announced GrubHub deal?

How broadly will it be rolled out in TripAdvisor?

How will it be promoted?

And how should we think about the revenue opportunity longer term?

Sure.

So as you know, we have a massive number of restaurants on the site globally.

And folks, when they are looking for that place to eat, sometimes want to make a reservation if it's a restaurant that takes reservations electronically, often want to read the reviews, look at the photos and sometimes want to do take-out, order the food delivery.

GrubHub is a great partner.

It covers many cities, certainly not all cities in the world.

Presuming that works well for us, you'd expect to see additional food delivery platforms coming online.

I'd caution against viewing that as a move-the-needle revenue opportunity for us, as our traffic certainly tends to be a reasonable mix of tourists and locals.

And tourists are going to be less likely to use a food delivery service.

But based upon the scale of our restaurant space, it's kind of yet another product in the mix.

Our next question will come from the line of Kevin Kopelman with Cowen and Company.

Could you give us any color on what you're seeing in April, especially with the Easter timing comp?

And then on that, you also advised against extrapolating Q1 RPS growth into Q2.

Can you give any more color on why you're more cautious on that?

Yes.

Thanks, Kevin.

In our prepared remarks, we didn't call out any specific comments about April, but we did comment more broadly about extrapolation from Q1 into Q2, as you point out.

And we caution against extrapolating some of the graphs that we put into our prepared remarks into Q2.

Q1, as we said in the prepared remarks as well, was a quarter in where the year-on-year comps for our paid marketing investments were relatively favorable.

So we saw an ability to significantly grow our paid revenue line in Q1, which may not translate directly into Q2.

And I'll make the point more generally, is some of these graphs that we put in, RPS and shoppers, if you look at the last few quarters, and you could be tempted to just draw a straight line through some of those curves, and we just want to make sure that although we have reiterated our expectation that we'll have double-digit growth in the click-based line this year, and we feel very good about where we are and we feel good about our outlook there, the quarter-to-quarter progression maybe not as smooth as some of the graphs have shown so far.

Our next question will come from the line of Jed Kelly with Oppenheimer.

On capital allocation, you continue to turn out a decent amount of free cash flow and had a second consecutive quarter of decent buyback activity.

Can you share how you're managing buyback versus acquisitions versus investing for growth?

And just given your comments around the long-term outlook for the business and where the stock is currently trading, would it make sense to be even the more aggressive around the buybacks?

Thanks for the question.

Indeed, we put $150 million to work in Q1 for share repurchases at an average price of $42.49.

We want to underline what you said in your opening statement is that we are indeed generating significant cash flows from operations.

As we think about allocation, the areas of where we could invest outside of the business / operations is M&A, as evidenced during in the last few years through acquisitions that we've made, like Viator, like TheFork, LaFourchette.

Last year, we made a number of smaller investments, but a number of investments there as well.

And we'll continue to look at opportunities that may present itself in the M&A space.

And then there's share repurchase as an opportunity as well.

And as you say, we clearly, in 2016 and in the first quarter of 2017, have seen an opportunity to buy back some of our shares.

The program is reviewed by the board from time to time.

In January, the board approved refreshing the $250 million allocation that we had for share repurchases.

And we have a 10B5 program in place with different triggers that we don't want to go into in detail, but they resulted in a purchase in Q1 and, from time to time, will allow us to benefit from periods where we believe our share price is attractive as an acquisition of ourselves, of our own stock.

Our next question will come from the line of Douglas Anmuth with JP Morgan.

Can you guys help us understand how much of the recovery and a return to the pre-IB monetization levels that you're seeing in the U.S. is coming from optimizing on Instant Book versus shifting back to meta?

And then also just wanted to follow up on mobile.

Can you just talk about what the mobile-to-desktop monetization gap is here currently?

And your comments on extrapolating, is it the same rationale, Ernst, that you mentioned a few minutes ago?

Yes.

So the U.S., I think it's a little bit difficult to untangle after so many years of how much of the RPS is this or the other.

I do want to comment that we make continuous improvements in our revenue per shopper on desktop and in mobile.

And in the U.S., we have been able to do that.

And we've been making continuous improvements on Instant Booking.

We said a few quarters ago already that in the U.S., Instant Booking was no longer dilutive to the overall offering.

And we've been improving both meta and IB since then.

So I think it's a combination of factors, but it's just a sign that after fully incorporating Instant Booking in the U.S., we're now better off than we were in '14 and continue to show impressive growth to our revenue per shopper.

In terms of the desktop monetization gap, we've said in the past, mobile monetizes roughly at 1/3 of desktop, and those trends changed a little quarter-over-quarter, but not dramatically.

We saw in the first quarter where there was revenue-per-shopper improvement on mobile.

We had 35% growth in our phone revenues and about 25% growth on the shopper line.

And so we're very pleased with that.

And we continue to plug away at making sure that we make that gap as small as we can.

And then in terms of your last question about the Q2.

Yes, again, with mobile as with the general comments I was making, quarter-to-quarter, there may be some volatility in revenue per shopper or in monetization of mobile, but the long-term trend is pretty stable.

Our next question will come from the line with Heath Terry with Goldman Sachs.

This is Daniel Powell on for Heath.

Just a couple of quick ones from us.

First, on the traffic growth and comments you've made around paid acquisition, just curious if you could give any comments on what you're actually seeing in organic search and how much of that is coloring your decision to lean more into paid acquisition.

And then on the Non-Hotel side, I just want to make sure we're interpreting our comments there correctly about it approaching 2016 growth levels, maybe suggesting acceleration in the back half of the year against some tougher comps.

Is that driven by seasonality mostly?

Or is there something else that changed in the business?

Thanks, Daniel, for the question.

This is Steve.

So I wouldn't say there's anything particularly different in our organic search trends.

Google continues to take more space on search result pages on all devices for their own products and for other paid ads.

There's nothing new in that statement.

I've been say something like it for several years now.

When we look at our overall marketing mix, we have more and more sort of brand traffic.

We have paid traffic.

We have organic traffic.

We have our CRM program.

And our focus, at this point, is on building, what we call, that branded direct traffic, focused with an offline TV campaign, in order to build more brand direct, build more loyalty to Trip when a user is in mind for shopping.

So I wouldn't say that shift is not a result of any specific organic search trend, but it is a change for us.

And we view a pretty meaningful one in terms of building the business for the long term on a large component of brand-direct channels.

In terms of the Non-Hotel, it is indeed a very seasonal business, and even the growth is somewhat seasonal.

Last year, in the first quarter, we posted, from memory, I believe, 14% growth for Non-Hotel, which was substantially lower than the rest of the year.

And indeed, we said in our prepared remarks that we expect for the full year, Non-Hotel growth will be approaching the levels that we had for the full year of last year.

So implied in that is that we expect growth to be better in the rest of the year than in the first quarter.

Other than some of the seasonality we've seen in Vacation Rentals for instance, where there is a continued shift to free-to-list model, which means we get paid on booking and tends to happen early on at the time of stay rather than at the time of booking.

And then there's some currency issues in there.

Our Vacation Rentals and our Attractions business are, for a large part, pound driven.

And of course, we had Brexit and the adjustment to the pound happening in the second quarter of last year.

So all in all, a number of factors why we believe Q1 should be a lower point compared to the full year in terms of Non-Hotel growth.

Our next question will come from the line of Nat Schindler with Bank of America.

Ernst, just wanted to go over a little more detail on the guidance and/or your profit outlook comments.

You said you could maintain profit outlook comments.

But at this point, your profit outlook comments are simply flat to down.

That's, quite frankly, an infinite range.

So if I look at ROI as negative on the brand campaign, I know you're switching out some inefficient marketing dollars to pay for that, but there would be some revenue lift from that brand campaign as well.

How much are you really thinking this hits on a negative line?

Or are you saying that this has no impact from where you were thinking last quarter?

Yes, Nat.

Thanks for that question.

We said last quarter, we said flat to down.

And you're right that we didn't specify what down really could imply.

What we really wanted to signal in this quarter is that, that outlook has not changed because of TV.

And that effectively, the negative impact on EBITDA we expect from TV, you said it's ROI negative, we believe it's ROI negative in the near term.

In the long term, we think it's very much ROI positive.

But it will have a near-term drag on EBITDA, just the TV campaign itself, which is offset by these two drivers that I talked about, partly by reallocating some of our online spend.

So we've looked at our total portfolio of advertising spend and said, "TV doesn't necessarily have to be just a bolt-on." We have looked at what does that mean for our online.

And we have been allocating some spend that was on the margin less efficient than the average and allocated that towards TV.

And then as we highlighted, we have upgraded our outlook on profitability of our Non-Hotel segment.

And those two factors, together for '17, happen to be roughly equivalent to the EBITDA loss we expect from the TV campaign.

So taken together do not have a meaningful impact on where we thought the business would be 3 months ago without this brand advertising campaign.

And therefore, we are reiterating the same guidance, flat to down, that we did a quarter ago.

Great.

And then just one secondary question on this.

Because this is happening fairly late in the year, I'm assuming that the $70 million to $80 million is smaller than you would do next year and the years after, and you mentioned a few years.

This campaign will continue for at least a few years.

Can you give us any color on what you think a normalized year on this campaign will be?

And then just another corollary to this brand campaign, you pulled back after the '14 and '15 campaigns, you pulled back on those campaigns because they didn't work as well as you hoped.

What makes you think that this time, TV will work better?

Yes, so your observation is correct, is that the $70 million to $80 million in 2017 is starting well into the year, of course.

And as such, as we look in following years, the investment will be over a full year rather than just over what is going to be roughly a little bit more than half a year in 2017.

So that's a good observation.

The additional point is that we're starting with the U.S. and a handful of other markets, not yet with a global campaign.

The U.S., plus a handful of markets, will be a very meaningful representation of our total revenues in these markets, of course, but not yet a global campaign.

And we may add other markets as we go forward.

It's a little early for us to be more specific about what our expectation is for spend in 2018.

We're obviously going to have more information under our belt in the next few months as we are actually rolling out advertising in these markets.

And in due time, we'll revisit that question of what the size would be for '18.

And with respect to the brand campaigns that we had in '14 and '15, it's not that they were failures, they just didn't quite achieve all of the objectives that we had wanted at the time.

So we're actually taking a number of the learnings from those campaigns and applying them.

One of the learnings was, "Hey, we don't want to just say that you can book on TripAdvisor.?

We want to drive home the point that you're going to find the best price on TripAdvisor.

And that's more of a value orientation in the message.

And then of course, the product needs to deliver that lowest price front and center.

And a couple of years ago, it didn't.

The hero price on a particular page was not necessarily the cheapest.

And so we've added more supply into our store, and we changed our algorithm, so that we're showing you better deals and we're showing you better prices all the time.

So the product is delivering more of what our current advertising will be talking about.

So you take those components, plus the learnings, and that gives us confidence that this ad campaign, just literally based upon what we did learn from last time, will work for us.

Our next question will come from the line of Paul Bieber with Credit Suisse.

I apologize if it's already been asked, but how should we generally think about the prioritization of meta versus Instant Book when you launch the new user interface in the coming months?

Yes, this is Steve.

You shouldn't think of those as necessarily connected.

We've been adjusting the placement of Instant Book over the past several quarters.

It's floating more , it's earning its right based upon what consumers want to do.

And that aspect doesn't change as we roll out this new interface.

And just a quick follow-up in terms of TV ad campaign.

Is the intent to drive more awareness the overall TripAdvisor brand?

Or is it to drive awareness of meta versus Instant Book?

Neither.

We have tremendous aided brand awareness in most countries.

We're looking to drive unaided awareness of TripAdvisor as a booking site.

And so many people know us just as reviews.

And when they think they want to read reviews, they come to TripAdvisor.

We want to make sure that when they're thinking about, "Hey, I want to get a good deal on this hotel that I want to stay at," that they come to TripAdvisor for that purpose.

So again, unaided booking awareness or unaided awareness as TripAdvisor as a booking site would be the key goal of that campaign.

And the campaign is completely agnostic to Instant Book versus meta.

It's about shopping.

It's about booking the hotel.

Our next question will come from the line of James Lee with Mizuho Securities.

Steve, can you maybe talk about some of the new targeting technology that you introduced for hotel search?

For example, from a couple of industry conferences we attended, a few agency talk about your firm offering like CRM targeting and retargeting using pixels, and just curious how those technology are performing versus your existing ones.

And are you also testing maybe other technology, like lookalike marketing as well on your platform?

So I'm going to answer the question as kind of what are the targeting technologies that we offer our CPC and CPM clients to best target traffic that they want that's on our site.

And so for CPM, we've kind of -- we have many years of dividing up audiences, luxury audiences.

There is a type of a lookalike product that our sales force has been using quite successfully.

When it comes to the CPC, we do offer things like retargeting pixels on our client sites, so that they can bid differentially on traffic that's been on their site that is now TripAdvisor.

Presuming that's a more attractive customer for them, they can bid up.

And that's capabilities that we've offered for several quarters now.

Are you seeing a very meaningful lift versus other targeting technology?

Talking to some of your peers, especially online marketing, when they start using CRM targeting, especially when it comes to search, they saw a pretty decent lift.

I'm just wondering from your process of working with the clients, how you're seeing that impact.

So yes, many of our clients have adopted many of the different things that we offer.

You have to put it a little bit in perspective because we're familiar with retargeting capabilities, as TripAdvisor advertises on other sites.

And we tend to want to buy traffic to our site that has travel intent versus buying a car or a house or something.

The folks that are looking to buy traffic on our site, all of our traffic is already prequalified as travel traffic, so it's already past the major hurdle.

Now we're looking to nuance it.

So our clients are looking to nuance it with their retargeting pixels or their lookalike audiences.

So yes, it's valuable, but not necessarily, which is why we offer it and why a lot of client use it.

And you can kind of infer some of that goodness in our CPC numbers from the number of bookings that we drive for our clients.

But it's not the same order of magnitude that a Facebook or a Google might be able to offer to their audiences.

Steve, can I also ask you a question regarding mobile monetization?

Some of your peers have been successful using cross-device attribution because they have [a lot of] log-in ID, so they can use mobile to provide [assist] credit for desktop attribution for conversion.

Are you doing the same thing right now internally?

Just curious where you are in that process.

Have you done any test at all?

And have you seen any success using cross-device attribution on your platform?

I'd say we, like many other companies, where a majority of our revenue is not from logged-in users, do have a really tough time tracking that mobile attribution.

It frankly drives us a bit crazy because we qualify millions of travelers every day on our phone, on our app and website, and then send them over to our clients.

And then those users come back the next day and book on desktop on our client sites.

And of course, we can't track it.

Our clients can't track it, so we don't get the credit that we deserve.

That's part of that big monetization leak to that we keep referring to.

And that, one day, Instant Book will help us address on the phone.

But no, I can't point to any tremendous data or successes in terms of cross-device attribution that we've been able to share with our clients, so still a work in progress for us.

Our next question will come from the line of Peter Stabler with Wells Fargo Securities.

I wanted to go back to the marketing, again, apologies.

You mentioned in the prepared remarks that -- a handful of markets overseas.

Can you just give us a little more color?

Are you going to go in with a similar stance in terms of weight levels in weeks?

Or is the international rollout more on a test basis?

And any color you can give, a sense of the coverage of your major international markets and whether the market list would be likely to expand appreciably next year or even later this year.

Sure.

So the handful of markets will sort of talk about the same message at a weight level that's appropriate for that market, i.e.

not a small test, but a meaningful spend, so that we can see results and optimize.

It would be perfectly reasonable to read into our comments that as we move on TV in 2018, since we're a global company, there's no logical argument that we would limit it to just a handful of markets that were in TV now.

You might instead read into, "Well, we couldn't tackle all markets in the very beginning." There's simply a bandwidth question that's involved.

So we think we're making a very meaningful move here, limited by our ability to execute and the desire to learn as much as we can, as quickly as we can in order to optimize for next year.

Our next question will come from the line of Naved Khan with Cantor Fitzgerald.

Steve, in the past, I think you've said that you think Instant Book is a better experience for mobile monetization.

Do you still hold that view?

Or do you think that it could be either Instant Book or meta?

And then I had a follow-up question on the Non-Hotel side.

So when the Instant Book product has a parity price, then my personal opinion is it's a better product because it's fewer clicks to get what you want done.

Where Instant Book has a worse price for whatever reason, we're not pushing it on you because we're not asking you to spend more money for the privilege of using this path.

And then rubber meets the road or the question's most relevant when you have a parity price with other providers, at which point, look, some users, we fully expect to be loyal to a different booking site.

They thank TripAdvisor for finding them the best price.

That best price might be on Booking.com, and they click over to Booking.com.

And they're happy with that experience.

And we're fine with that.

We got paid on the click and we helped drive the booking.

Other folks, because they're in the habit of booking on TripAdvisor, using our Instant Book with our stored credit card, will prefer the TripAdvisor option because it's literally fewer clicks to actually finish that booking.

No need to switch.

So our wealth of product capabilities here allows us to tailor the experience to what we think that individual traveler is looking for.

And we think that's a great option.

So Instant Book on the phone, when it has the better price, it's clearly a better experience than meta because it saved money.

When it's a worse price, you're probably not going to see it a whole lot.

And when it's a price parity, the user gets to decide, if you will, and we make our suggestion in terms of order based upon what we think the user wants to see.

Okay, that's great.

And then quickly on the Non-Hotel side.

I think -- and especially on Attractions, I think as previously said, you think of this as a 3- to 5-year sort of investment opportunity, where you actually want to focus on growing the business.

And I think we are in year 3 in 2017 and you're already looking at margins kind of turning positive.

Do you think that this business is going to -- is ready to see some margin expansion from here on?

How should we think about it?

Yes, observation correct, Naved, is we had described it as a 3- to 5-year journey.

And we are in year 3 and we're seeing profits coming through this year.

We have invested significant amounts of dollars in 2015 and 2016 in growing our supply, but importantly, as Steve noted earlier, in making all of this bookable supply also available on the TripAdvisor site, where we have most leverage in terms of exposing it to the largest number of users.

And we're now seeing that flywheel of supply, getting supply on, then making it available for all of these TripAdvisor users start to work.

And so this is the first year where we, in aggregate, are profitable for this Non-Hotel segment, which is a great trend, and we believe, in the long term, will make this segment not just a growing, but a very profitable segment.

Ladies and gentlemen, this concludes our question-and-answer session for today.

So now it's my pleasure to hand the conference back over to Mr. Steve Kaufer, Chief Executive Officer, for closing comments and remarks.

Sir?

Well, thank you, everyone, for joining the call.

The first quarter 2017 was a great starting point to the year.

We still have a lot of work ahead of us, but we are making progress on all of our initiatives as we build the best user experience in travel.

I want to thank our employees around the globe for the continued hard work.

And we look forward to updating you on our progress next quarter.

Thanks.

Ladies and gentlemen, thank you for your participation on today's conference.

This does conclude the program, and you may all disconnect.

Everybody, have a wonderful day.