Tripadvisor, Inc. (TRIP) Q2 2013 Earnings Report, Transcript and Summary

Good day, ladies and gentlemen, and welcome to the TripAdvisor Second Quarter 2013 Conference Call. [Operator Instructions] As a reminder, this conference call is being recorded. I would now like to introduce your host for today's conference, Will Lyons. Sir, you may begin.

Thank you, Sam. Good afternoon, everyone, and welcome to TripAdvisor's second quarter 2013 earnings conference call. I'm Will Lyons, Senior Director of Investor Relations for TripAdvisor, and joining me on the call today are our CEO, Steve Kaufer; and our CFO, Julie Bradley. Before we begin, I'd like to remind you that the estimates and other forward-looking statements included in this call represent the company's views as of today, July 24, 2013. TripAdvisor disclaims any obligation to update these statements to reflect future events or circumstances. Please refer to today's earnings release and TripAdvisor's filings with the SEC for information concerning factors that could cause actual results to differ materially from those expressed or implied by such statements. You'll also find reconciliations of non-GAAP financial measures to the most comparable GAAP financial measures discussed on this call in our Q2 earnings release, which is available on our IR site, ir.tripadvisor.com. Finally, unless otherwise stated, all references to selling and marketing expense, general and administrative expense, and technology and content expense exclude stock-based compensation, and all comparisons on this call will be against our results for the comparable period of 2012. With that, I will now turn the call over to Steve.

Thank you, Will, and welcome, everyone. I'm going to take a few minutes and highlight some key metrics and discuss some notable developments on our technology and growth initiatives. Julie will then give color on our financials and outlook before we take your questions. Our core hotel shopper growth metric remained strong and was up 41% during Q2 with broad strength across both our core and emerging markets. This drove 21% click-based revenue growth for the period. Our other revenue lines also fared very well, as 18% display revenue growth and 68% subscription, transaction and other revenue growth were ahead of our expectations. Collectively, this led to a strong Q2 total revenue and EBITDA growth of 25% and 16%, respectively. More users across the globe are engaging with TripAdvisor sites, whether it's on desktop, tablet or smartphone. Our average unique monthly visitors grew approximately 57% in Q2. At an average of more than 220 million monthly unique visitors, according to Google Analytics, TripAdvisor-branded sites had more than 1 billion unique visitors during the first half of the year. Roughly 79 million of these monthly uniques visit us via a tablet or a smartphone device, up 216% year-over-year, highlighting our continued strong mobile user uptick. We also accelerated user-generated content growth to an average of more than 70 contributions per minute on the strength of our social, email, badging and permanent branding campaigns. More recently, we introduced Review Express, which allows hoteliers to automate the post-stay review solicitation that will help drive even more fresh content to help travelers plan their trips. Review Express is just one of the many new features we've been working on to improve the TripAdvisor experience for our users and partners alike. The most notable and visible new future, which is now live to all of our global traffic, is hotel metasearch functionality. With the launch of meta, we now have the world's best travel content, combined with realtime pricing and availability all in one place. Because of a lot of outstanding work from the team here, we were able to roll out meta in 100% of our markets in early June, slightly ahead of our original plan. Everyone on the TripAdvisor team is proud that we've been able to delight hotel shoppers around the world with a new and improved interface. There hasn't been a bigger upgrade to our user experience in our history. From a meta monetization standpoint, as we predicted, the number of paid clicks is down multiples and conversion CPCs are up multiples relative to our classic bidding model. We continue to expect to see the biggest negative impact to revenue during Q3, as this will be the first full quarter at 100% meta rollout and as partners optimize and find equilibrium within our new meta bidding landscape. We've not yet reached revenue neutrality, but, based upon the progress we've made, we continue to expect that this transition to meta will be revenue-neutral by the end of the year. To get there, our core product and engineering teams continue testing and optimizing the site for commerce, making progress on generating more clicks through our on-site conversion initiatives. Simultaneously, our engineering and sales teams are helping the auction dynamics mature by working closely with partners to improve our client-facing tools, to help partners -- to help more partners spend and spend more effectively on our platform. Along those lines, we've heard a growing number of independent hoteliers, partners that traditionally -- that we haven't traditionally supported in our CPC business, tell us that they're interested in buying clicks. For years, small budgets and technology bottlenecks have stood in the way of addressing this vast underserved consumer set, but they are increasingly interested in working with TripAdvisor, given our growing importance in the hotel shopping funnel. Just earlier this month, we announced our new TripAdvisor Connect platform, which enables Internet booking engine providers to connect directly to our commerce engine. This allows them to provide their independent hotelier clients with the necessary technology connection that will allow them to bid in our metasearch auction. We have received very positive initial feedback from our announcement, and we expect that many of the largest companies that provide Internet booking engines to independent hotels will be working with us soon. Later this year, we intend to launch the second part of the TripAdvisor Connect platform, namely, the self-service bidding interface that will allow hoteliers to bid directly for our traffic. We believe this is an exciting development for a large number of independent hotels, especially those that are not currently represented by an online travel agency, as the hotel will be able to advertise their rates and availability for the first time to our massive travel audience. Another exciting aspect of TripAdvisor Connect is that it includes an adaptive Review Express product, which can automatically generate post-stay review solicitation emails. This will help hoteliers engage guests on a more frequent basis who will help them to add more fresh, valuable content about their properties on TripAdvisor. To be clear, we don't expect a material impact to our results from the TripAdvisor Connect platform in 2013. We do, however, think that this will improve our value proposition to travelers by providing them rates and availability for more properties, as well as help independent hoteliers advertise in our growing global marketplace. I'll now touch upon some interesting developments in a few of our growth initiatives. Earlier, I alluded to our rapidly growing mobile usage, reaching over 35% of total unique visitors. We've also reached more than 50 million cumulative app downloads, including those from our new Jetsetter and GateGuru apps, which are -- which is up over 125% year-over-year. This download number also includes a nice early traction in our new Samsung partnership, as we've seen a better than 10% app utilization -- app initialization rate on the reported 20 million Samsung galaxy S4s that have been sold to date. In terms of mobile monetization, as expected, tablet hotel shopper monetization has matched that of desktop, and I'm excited by how our teams continue to refine and improve user experience on that device. On smartphone, our hotel shopper growth continues at a triple-digit rate, but monetization remains at less than 20% of desktop. We attribute this to our ongoing priority in mobile on engaging and delighting as many users as we can on a more frequent basis, whether it's to read hotel reviews, send a postcard to a friend or reserve a table at a restaurant. Moreover, our smartphone team is dedicating significant resources to building native iPhone and android apps, which we will believe will allow a user to take better advantage of all that TripAdvisor has to offer on those devices. In our fast-growing subscription transaction and other lines, we've seen a nice uptick in business listing sales productivity, as we have more fully trained sales reps in more parts of the world. We are also excited by the prospect that our forthcoming TripAdvisor Connect platform may bring, which gives more property owners around the globe another reason to work with Trip. In our Vacation Rental business, early results from free-to-list remain positive, as overall listings are up, as well as mailable users, traffic and inquiries. On the consumer side, we've increased the percentage of our listings that are online bookable properties from 0 last year to over 20% this year, and we shortened the booking flow to improve conversion. We've integrated the majority of FlipKey and Holiday Lettings properties onto the main TripAdvisor site, and we're working on integrating new inventory to give vacation renters more choice in Spain. On the international front, we are working on 3 new TripAdvisor domains in localized Spanish language: Colombia, Chile and Venezuela, to better serve the local travelers and to better align our offerings with the local OTA partners in that region. Late in the quarter, we had the leading mobile resource for flight and airport information around the world, GateGuru, to the TripAdvisor family. Also at the end of June, we decided to sunset the SniqueAway brand, combining its assets with our recently acquired Jetsetter brand. We believe this change will be positive for the business, our partners and SniqueAway members. In summary, the second quarter closed out a strong first half for 2013 for TripAdvisor, as we made great progress to create the best experience for every user on every device in every geography. All TripAdvisor employees deserve thanks, especially the meta team who, in a true speed wins fashion, delivered a fantastic product upgrade in a short amount of time. I thank you, and I think travelers all around the world thank you as well. I'll now turn the call over to Julie, who will provide some color on the financial results, as well as our outlook.

Thanks, Steve, and good afternoon, everyone. As Steve just described, during Q2, we saw a continued strong traffic trend that provided a nice backdrop for our revenue growth. Second quarter revenue growth of 25% was driven by continued strong hotel shopper growth and strength across our product suite. Currency had a less than 1% impact this past quarter. Adjusted EBITDA growth slowed sequentially to 16%, versus 30% in Q1. Impact from the meta rollout, hiring and adding costs through our acquisitions were the key drivers. Our Q2 click-based revenue was up 21%, fueled by hotel shopper growth of 41%. As we rolled out meta to 100% throughout the quarter, it negatively impacted growth by approximately 6% to 9%. Hotel shoppers on smartphones and from international geos continued to monetize at lower rates, and each accounted for approximately 5% to 10% of the delta between hotel shopper growth and click-based revenue growth. Specific to geographic mix, revenue from international points of sale was 49% of total revenue during the quarter, which was slightly lower than Q1 due to the timing of our metasearch rollout to various geographies, with our core U.S. market being among the last markets to move to 100% deployment. Hotel shoppers continued to grow rapidly throughout the globe with notable continued strength in core U.S. and European markets. On the display baseline, revenue growth accelerated to 18% due to traffic growth and better sell-through rates. This past quarter, we saw some summer-specific ad spend, nice sales traction from our Delayed Ad Call innovation, improved customer targeting functionality and deeper industry relationships with DMOs, hoteliers and airlines. Subscription, transaction and other revenue, which include Business Listings, Vacation Rentals and our transaction businesses, Tingo and Jetsetter, accelerated to 68% growth, ahead of our full year expectations. Beyond having nearly a full quarter contribution from Jetsetter and nearly 2 months contribution from Niumba, we owe this acceleration to increased business listing sales productivity and higher vacation rental velocity from our new transaction-based offering. On the expense side, our hiring pace quickened in Q2, as we pursued our product, platform and global growth initiatives. Direct marketing costs increased slightly versus last quarter, as we started testing our new TV campaigns in 6 U.S. markets, in an effort to promote -- to provide a brand awareness and diversify traffic. We expect to increase TV spend materially in the second half of the year, as we broaden our testing domestically and internationally. Moving on to taxes. Our Q2 GAAP effective tax rate was 26%, which is consistent with our ongoing mid- to high-20s expectation for the year. CapEx was $14.3 million for the quarter or 6% of revenue. This increase is primarily related to leasehold improvements in several new and existing offices. From a liquidity standpoint, we ended the quarter with $616 million in cash, cash equivalents and short-term and long-term marketable securities. At the end of Q2, we had outstanding borrowings of $316 million, as well as an undrawn credit facility of $200 million. In terms of capital allocation, we paid approximately $30 million, net of cash acquired, to complete our 4 acquisitions in the quarter. Also, we repurchased 675,000 shares of our common stock for approximately $42 million. Subject to prevailing market conditions, we remain focused on offsetting expected employee equity dilution during 2013 with share repurchases. Finally, you'll notice a new deferred merchant payable line on our June 30 balance sheet. This represents cash net of commissions that we received from travelers at the time of the booking, which we ultimately use to pay hotels and vacation rental owners in some of our transaction-based and Vacation Rental businesses. As of June 30, our current balance was $31.3 million, of which $14.5 million was acquired through our M&A transactions during the quarter. As long as these businesses continue to grow, we expect the changes in working capital related to these balances will positively impact operating cash flow. With that, let me provide updated thoughts on our 2013 outlook. For click-based revenue, our thoughts from last year remain largely unchanged -- our thoughts from last quarter remain largely unchanged. We are very pleased with the positive user feedback on our new Meta Display and have made great progress driving more clicks and helping partners better determine their bidding efficiencies. However, the bidding landscape is still maturing. Given that Q3 is the first full quarter of 100% meta, we continue to expect the transition to have a larger negative impact on Q3 revenue than it did in Q2. Assuming continued on-site conversion improvement and stable partner pricing, we continue to expect click-based revenue growth re-acceleration in Q4 2013. For full year 2013, we are reiterating our click-based revenue growth guidance of high teens to low 20s. As it relates to display, based on our strong first half results, we now expect mid- to high-teens growth for the full year, with Q4 being our strongest quarter for this product. For subscription, transaction and other business lines, we believe continued sales productivity in Business Listings, Vacation Rentals traction and nice contributions from our recent acquisitions will now yield high 50s growth for the full year. Remember that traction of our Vacation Rentals free-to-list product and slashed sales will contribute similar seasonality to click-based revenue where Q3 is seasonally strong and Q4 is seasonally weak. In sum, we are reiterating our total revenue growth expectation of low 20s for the full year. I'd like to again remind everyone that our business is sensitive to fluctuations in hotel shopper growth and then [ph] partner CPC pricing, both of which are risks. On the flip side, remember that our forecast does not include any meaningful traffic or repeat visitor lift from our planned off-line ad campaign. So that represents some potential upside from a traffic and revenue standpoint. On the expense side, our investment and hiring plans remain unchanged, given our growth initiatives and large market opportunities ahead of us. From a forecasting standpoint, we have considered our strong result in the first half of the year, offset by our ongoing meta transition, off-line ad spend and our recent acquisitions. Net-net, we are reiterating our expectation of mid single-digit EBITDA growth for the full year 2013, which assumes negative Q3 and Q4 year-over-year EBITDA growth due to the timing of our off-line ad spend. In conclusion, we are very pleased with our continued improvements to the TripAdvisor experience for our users and partners. We will now open the call up to your questions.

[Operator Instructions] Our first question comes from Lloyd Walmsley of Deutsche Bank.

Wondering if you can just give us a little bit of color on how metasearch is changing the user experience and activities. So when you look at the activation of hotel shoppers into monetized hotel shoppers, are you seeing a higher percentage in metasearch actually click on ad units versus the traditional? And just as a follow-up, can you kind of parse out the puts and takes of metasearch in terms of how much increase you're seeing from end users and then what you're seeing in terms of volume and price on the paid click side?

Sure. Thanks, Lloyd, a good question. So the engagement metrics are basically all in the right direction when we look at how the new experiences rolled out. So page views per session, how much were people searching around, that's up by north of 10%. Bounce rate is something else that we measure, hey, how often when people come to the site do they leave without doing anything, and that's going down across the vast majority of our points of sale. People are staying longer on the site, so in general, that's a good thing, they're doing some more searching, there's some more researching. To the question of, "Well, are more -- if I'm understanding the question correctly, "Are more people shopping on TripAdvisor, entering their dates and then clicking off to our clients, the Expedias and Hiltons of the world?" The answer is, "No." It's a meaningfully smaller percentage of visitors that finish their shopping experience with a click off. That's entirely what we expected and what we expect to see, frankly, going forward because the traveler -- in our old classic model, upon getting interested in a hotel, would naturally want to know, "Hey, how much is it? Is it available?" And the only way to find out would be to click the show prices button, which would generate a paid event off to a client site, only to find out that the hotel is sold out or that it's $1,000 a night. So it's a potentially very low-converting click, but it was a paid event for us and it did count as a shopper who went downstream to click -- went downstream to a client site. Whereas now, they put in their dates and they see that, "Oh, the hotel is sold out or it's way too expensive for me." And a bunch of them kind of finish their shopping experience at that time. They don't click to a client so they don't generate a paid event for us. We're fine with that because it's actually keeping that shopper on TripAdvisor to help find the right hotel, not only from reviews but also price and availability. And maybe that shopper comes back the next day, maybe that shopper shops around a little bit more on the site, or maybe they were just looking anyway and weren't planning to book anything, in which case I had sent, in the old model, a very low-qualified click off to a partner site that wasn't going to buy anything, so at the end of the day I wasn't going to get paid. So we're still sending -- or in the new model, we're sending meaningfully fewer clicks to our clients, but they're converting at a meaningfully higher conversion rate and the clients, in turn, are paying us more. And all of that, net-net in Q2, was coming in at that, hey, we were estimating a 6% to 9% headwind when we did all that math and when we counted all the clicks.

And then just in terms of what you're seeing in terms of volume versus price, can you give us any color there on how metasearch is transitioning?

So what we have actually predicted in the beginning, in terms of, kind of, on average, and this again varies a lot by partner and by placement and by position on the site, but on average, we're seeing the click volume for these meta-hotel clicks being down, call it, 3x and the conversion rate and the pricing being up, call it, 3x, and those are approximations. And to be clear, they don't represent all of the clicks that we send to our clients, but for these priced hotel clicks, those are the kind of multiples that we're seeing. In the end, to reiterate, not all partners see the same thing, and it's different geographies, but I'm trying to give you a general average on what it looks like. What I'd also add, we did mention in the prepared remarks, is that we had rolled out meta over the course of Q2. So Q3 -- and we've rolled out -- we've kind of finished the rollout at the end of May, early June by finally moving our U.S. point-of-sale, our biggest point-of-sale, over to 100% traffic, after trying it at much lower percentages. So the biggest and one of the most profitable segments, one of the richest CPC segments for us, the U.S. market, rolled to a full rollout only for the last month. So when we talk about Q3 being a bigger meta headwind, that's because we've got that whole rollout in front of us, the whole site is on meta now, and we're making progress on improving the conversion, but we're dealing with 100% of our traffic. So there's just a lot of different things that are changing. The full meta rollout is a bigger headwind. The fact that we've made some conversion improvements along the way is a bit of a tailwind. We expect to make some more conversion improvements. And then we've got a whole bunch of other external events that have already kicked off Q3, which are, frankly, mostly headwinds. So we have our wonderful new citizen of the world, Prince George, who no doubt is causing various people to stay at home and watch TV awaiting the news before they travel. We have a big heat wave in the U.K. and in parts of the U.S. and other parts of the world that we believe have slowed down some of the traffic to our site. We got an airline crash, that is probably more shorter-term than longer-term, but we have seen, for the first couple of weeks in July, a noticeable headwind on the traffic, which naturally becomes a headwind on the revenue. We hope that, that's not a long-standing thing, but it hasn't been a great beginning of July for us.

Our next question comes from Michael Purcell of Stifel. Michael B. Purcell - Stifel, Nicolaus & Co., Inc., Research Division: My question for you is on the hotel shoppers. Steve, I was wondering, could you give us some insights into that 41% growth? If you could kind of give us some insights and geographically where that's coming from? And then secondly, if you'd maybe just offer some comments on what you're seeing in terms of your test markets ad campaign and how that is rolling forward?

Okay. So to the first part, on the traffic. So, in general, our core markets are doing fine in traffic over the quarter. As we had alluded to, some of our newer markets aren't growing at quite the rate that they were last year but, overall, our mature markets are hanging in there quite nicely. Again, the U.K., in particular, for the first couple of weeks in July, hasn't been as robust a growth as we had seen in the course of Q2. When we look at our TV test markets, we have been working on 6 different markets. We've been doing our best to understand the pieces that we get to measure, comparing city to city, as well as comparing what happens immediately post running a spot. So we've got some great data coming in. We knew it would be a learning experience for us. So, again, we've only spent a couple of million dollars so far, but we have no reason at this point to be pulling back on our expectations to spend kind of the full budgeted amount going forward. We're thrilled with the addition of Anne Bologna, who's our new VP of Brand, who will be helping us properly allocate and carefully track the spend for us. And again, we'll -- you'll be seeing us on air, not only nationally in the U.S. but also, at least, in several additional international markets over the course of the second half of the year.

Our next question comes from Nat Schindler of Bank of America Merrill Lynch.

Can you just talk a little bit about what you're seeing in the disparity in clicks from users, as well as what you're getting in differential pricing between the top spot, the second spot and the third spot on your normal Meta Display? And I know that tends to change. Additionally, could you talk a little bit about any impact you're seeing from Google Carousel? And finally, and I think this is all one related question because you asked for it that way, could you talk a little bit about why you saw such a striking acceleration in click-based revenue growth from your non-Expedia clients? And who particularly were they? Were they hoteliers or other OTAs?

Sure, Nat. I'll do my best to remember all 3 parts to the question here. So just like Google Search and most other placement-based offerings, being in the top position for a property yields meaningfully more clicks than being in position 2 and 3. A little different than Google Search in that the actual price that gets displayed, of course, does sway the click behavior. It's an almost guarantee for a client to get more traffic if they bid up to that first position. When you're in the 1, 2 or 3 position, you're getting the vast majority of the clicks. Positions 4 through whatever get a -- we'll just call it a tiny fraction. So to the fourth player, to the fifth player, we're always making suggestions on how to get into the top 3, because the listing traffic from having a big button placement versus a text link is quite meaningful. Even if you have the lower price in the text link area, sort of, beneath the buttons, it's always best to get into that top 3. Again, the analogy is, in a Google Search result, you've got 3 placements right at the top of the page and then a bunch of others in the right-hand rail. And everyone knows the right-hand rail gets a small fraction of the clicks that go to the one from the top 3. Google Carousel, yes, we saw it come out. If you noticed when you click on it, it actually just redoes the search on the name of the hotel that the picture is for, and what we found or what I personally found when playing around is that TripAdvisor usually ranks pretty darn well for that individual hotel search. I have no insight whatsoever as to how many people are actually clicking on the Carousel. It's nice to see pictures, but it's really not much of a hotel selection tool, so maybe it's helpful in engaging the searcher, but I don't view it as a particularly strong shot across our bow in terms of hotel selection. We continue to have, obviously, strong hotel shopper growth. And we have, frankly, no way of knowing whether it would have been stronger but for that, or it made no difference at all. So we're, let's say, jaded on the number of things that Google does in the search result page. And most of them, at the end of the day, tend not to affect us. They rolled out several new search engine algorithm updates over the course of the quarter, and we've got a pretty powerful brand with pretty strong placement so we tend not to see any interesting changes. Again, if history is something to judge on, we tend not see too many interesting changes when they've rolled out their algorithms. And then the last part of the question, the CPC growth acceleration from non-Expedia. It's an auction, and so this was the case where we rolled out meta, and some partners were a little faster on the uptake than others, some partners measured things a little bit differently than others. And as you could guess, maybe Expedia wasn't in at the top position as they had been across all of our markets, not to say that they won't reclaim that position where they were or that people are just taking a different approach to getting started in our auction. I don't really have any comment as to who would have taken their place or who would have taken their click share.

Our next question comes from Douglas Anmuth of JPMorgan. Bo Nam - JP Morgan Chase & Co, Research Division: This is Bo, on behalf of Doug. Just one kind of related to Nathaniel's end question was just on the reduction in Expedia, if you can give any more details. And related to Expedia and the other partners, we know that your bidding platform has been getting a lot better. And how the bidding -- how dynamic the platform is at this point? And then, just secondly, a quick question on just traffic sources, how has that changed over time and have you maybe shifted some traffic -- incoming traffic away from Google and towards some other sources?

Sure. So we don't really have any more comment on the Expedia piece. The bidding platform, we have made sort of great strides in automating components, certainly on our end. Our larger partners have the ability to bid every single day if they want. I don't believe any of the big partners are leveraging that fact. I'm not surprised because there's just a lot of data to analyze each day. And at the end of the day, they want to make sure they are getting conversions out of it. So again, I can't speak for them, I don't want to speak for them, but if I were in their shoes, getting to daily bidding wouldn't be top on my list. I'd want to make sure I'm getting all the information I can from TripAdvisor, and that's exactly what we're providing. We're providing the ranking reports. We're providing the bid recommendation reports. We're providing them on a super frequent basis, by markets, for all the individual properties. And I would venture a guess that many of the bigger partners are now saying that while they love some additional API automation, which we're working on, but while they'd love some of that automation, they're getting plenty of data right now, and the onus is on them to find new ways to consume it and analyze it. And then traffic sources. Again, we don't really go into quarterly commentary on traffic sources. It's always been known that I'm always interested in growing additional traffic sources in all cases. The TV is certainly a meaningful strategic opportunity for us to drive more domain direct, more awareness of TripAdvisor, thereby reducing our dependency on any and all paid traffic or search traffic in general.

Our next question comes from Tom White of Macquarie.

I was hoping you guys could maybe talk a bit about the ways you still have opportunities to maximize on-site conversion. I imagine you're focused on stuff like improving the speed of the site and the search results, but are there other sort of specific examples of things you guys are focusing on to drive that? And then just secondly, on the shift to meta, I realize it's still very early but if I think about sort of the 2 main monetization parts of the site, the hotel-specific pages and then kind of the more general travel metasearch, can you give us any sense of sort of what is the breakdown of the percentage of paid clicks that's coming from both of those 2 areas and maybe where you envision that breakdown being in the next 12, 24 months or so?

So the -- certainly, the vast majority of our clicks are in the hotel category. We have air clicks and we have some restaurant clicks, and we have some other clicks. We use Google AdSense as a backfill in many cases and we don't even know what types of clicks those are. When I think about the business and when I think about, over the next year or 2, I am still thinking mostly hotel clicks because there's a lot of money there, there's a tremendous amount of demand, we're still tapping into a relatively small portion of the online travelers and that's our core, that's absolutely what we are best at. When you combine the price comparison, the rates and availability, the whole metasearch piece, with our content, we feel we stand meaningfully above all the competitors in providing a complete solution for that hotel search. To back up to the first on-site conversion, yes, yes. In fact, we do a lot of different aspects to improve the conversion. I can -- an example might be I'm going to Prince Edward Island, and I'm looking for some interesting places to stay, I land you at a page that might be just hotels. But we may not have a lot of hotels, so I want to make sure that you also see that we have some bed-and-breakfast, we have some other specialty lodging. And integrating those cross-sells on to the right page helps the user and, likely, generated another commerce click for us. When we look at personalization, we have a tremendous amount of information that's a bit hard to bring to bear at realtime. I've told you before that when I can use the information from your Facebook friends and you are friend-connected, you become a more valuable consumer, a more valuable traveler to me. You stay a bit longer and you do a bit more commerce. As that continues to grow, we get some additional benefit. But even if you're not instantly personalized, even if you're not Facebook connected, we have efforts underway to help understand a better property to cross-sell you to if, in fact, the one that you're looking at is sold out or more expensive than you want. And since we already have you on our site, getting an extra 10% of activity from you, wow, that's a huge opportunity for us. And it tends to happen, by experience, in increments of a percent here or a percent there, but adding up in our volume, turns it into real money.

Our next question comes from Mike Olson of Piper Jaffray.

For Vacation Rentals, could you just talk a little bit more about the traction of the transaction-based offering and remind us when that kind of fully got rolled out and how widespread it is, et cetera? And do you believe that it's being used by new customers or are existing subscription customers switching to a transaction-based listing to any significant degree?

Good questions. I don't have all those figures with me. Certainly, we're pulling in a lot of new customers that might have said, "Well, I'm not really sure it's worth the couple of hundred dollars for the subscription. Oh, you have a free-to-list option now. Great, I've got nothing to lose." So you get to add that new inventory. It just makes it a much easier sale. We do have some owners that upon renewal time especially, because that's when we frequently deal with them, say, "All right, well let me try that free-to-list transaction model." And to the extent that we get more and more of those folks converting over, that's great for us. We're fully in the transaction angle because at the end the day, we think that, that is what's right for the consumer. We want our incentives in Vacation Rental to be wholly aligned with providing the best experience. And the challenge, as we articulated before with the subscription model, is that the internal dynamics or that "I'm going to want to try to keep a subscription member live on the site." But if they're not getting enough leads because the property isn't good enough or it's overpriced, I end up having to feature that property high in the sort order or high on the page. And therefore asking the consumer, "Hey, you should try this property, by virtue of where I placed it on the page." Even though it's the kind of the wrong answer for the consumer. It's the right answer for my business because I want the renewal, it's the wrong answer for the consumer because there's probably a reason why that property didn't get a lot of clicks. As we move to a transaction model, poof, all that goes away. I'm hugely incented [ph] to put the property first that's going to get the best conversion rate because, at the end of the day, it's a transaction, that's how I get paid. If you're on a subscription model and you try to add a transaction model, but then you allow subscription people to pay to get ahead of the listing, ahead of the transaction listing, then you've kind of got a mixed model that says, "Well, we want to do what's best for the consumer on the transaction side, but we're still going to let other people kind of buy their way to the top of the recommendation order." And again, it's, I guess, better than pure subscription but not entirely in the camp of what's best for the consumer. So we do have a subscription model. We're not planning to get rid of the subscription model. But when you look at the pages on our sites, we always want to have firmly in our mind that we're going to be showing you the best possible property, based upon what we know of you and how well that property ends up becoming a successful transaction for our shoppers. Business is growing nicely, and we continue to invest in it, such as our recent acquisition. And we're putting a lot of effort against it and feel it's a great aspect to the overall TripAdvisor story.

Our next question comes from Heath Terry of Goldman Sachs.

Can you give us a sense of how meta is impacting revenue concentration, particularly to the extent that it's impacting supplier-direct relationship? And then on the mobile side of the business, any sense of sort of the different consumer behavior, whether it's direct or through search or conversion, that you're seeing app versus mobile web, and how that may be impacting the way you're thinking about growing your app installed base through marketing?

Okay. So the 2 pieces, has revenue concentration changed much, does [ph] supplier-direct matter? I'd say, at this point, you shouldn't think of the revenue concentration between OTAs and suppliers as having changed in any meaningful way. I wouldn't think of it -- I wouldn't expect it to change in any meaningful way going forward either. We're always trying to get more suppliers to recognize the savings they can achieve if they bid into the first position. But that was just as true in classic as it is in meta. So suppliers with good rate parity enforcement probably do pretty well on TripAdvisor because of the occasional bias toward the supplier-direct. To the question of consumer behavior, mobile app versus web, we're still early into the Samsung arrangement so we think that there's a nice potential to learn something new about how an app, now pre-installed, gets a different style or perhaps a different level of engagement. We've gotten a lot of trial. We'll see, over time, how we get used when it comes time for that individual to go take a trip. I'm not sure there's a lot I can add in terms of marketing or customer acquisition channels between the 2. We obviously get a whole lot of downloads of our app because when you're searching on your mobile web, on your phone, using your web browser and you land on TripAdvisor, we're sure to give you a message saying, "Hey, come on down -- come on, let's download the app. It's a faster, better experience." That -- and just by the sheer nature of our traffic, that drives a fair number of downloads, which we're very pleased to have. I think the bigger shift will happen after we release our Truly Native [ph] app, because most of our mobile app is technically on a web view. So it's not as fast as it should be. So we're quite pleased with the redesign that we did in the past 6 months for the mobile app. But we know, going entirely native on the iPhone and Android, we expect will give us a meaningful boost in usage and engagement, and that's going to be great for the consumer and great for us.

Our next question comes from Scott Devitt of Morgan Stanley.

This is Nishant for Scott Devitt. For some of the markets where you rolled out metasearch early, such as the U.K., are you close to or at revenue neutrality that makes you confident for -- to achieve revenue neutrality for all sites by the end of the year?

No, it doesn't really work like that. We get to revenue neutrality by rolling out improved placement of the meta widget, improved cross sell, improved engagement, improved -- the different ways that folks interact with commerce on a page. So we rolled it out in U.K. and one of the reasons, hey, we can get our clients used to the bidding. And some of the clients did get used to bidding there. And the initial rollout, initial implementation was really quite painful on the revenue side, which was good that we only rolled it out to a small percentage of our traffic. And so we saved the biggest for the end, which is when we rolled the U.S. to 100% in June, because we were able to apply the learnings in April and May so that when we rolled it out in June, we already have the benefit of some meaningful conversion uplift. So we don't expect to be kind of revenue-neutral country by country. It's more the optimizations we make on the site improved conversion pretty much everywhere at the same time. And then the pricing, on a per country basis, is what'll, at the end of the day, I think, make some points of sale might be up a little bit from classic, and other points of sale may be down a little bit and we'll claim success when the overall net-net is rev neutral for all of our points of sale.

Our next question comes from Ken Sena of Evercore Partners.

This is Andrew, in for Ken. We were hoping you could provide a little more clarity on the definitions of marketable members versus hotel shoppers, the difference between them and how you measure them.

Sure. So a marketable member is an email address of a traveler whom we have permission to email, essentially of any frequency but we typically email them weekly. And we gain new members everyday. As new people write new reviews, they become marketable members. And then people unsubscribe from time to time. And so we'll lose marketable members. They don't drop their TripAdvisor account, they just go to the Subscription page and unsubscribe. So that's our marketable member. Our hotel shoppers are the subset of all traffic hitting all TripAdvisor points of sale, all the devices that touch a hotel-related page. So either a specific property page, a page about the new Marriott hotel, specifically, or a page that lists all of the properties in Hilton [ph] or all of the properties in any city. So that's our rough definition of an intent for a -- someone on our site to visit -- to do a hotel booking. To be clear, we still get, and it's fast-growing, a lot of traffic around the globe, in great categories like restaurants and attractions and overview categories like a city page for us. And we love the traffic, we tend not to buy any traffic to those pages because the folks are looking for attractions or restaurants, or things to do in a city, which is great from a brand experience, great for travel planning, part of the complete solution that we offer and none of our online travel agency clients offer, that Google doesn't offer and all the rest. But since that number fluctuates a lot and can go up and can go down or -- recently just going up, up, up, since we don't monetize those folks as well, we stick to the metric of hotel shoppers as the more meaningful number to report on.

Our next question comes from Mark Mahaney of RBC Capital Markets.

I just wanted to ask Steve a broad question about your latest thinking on how far down the transactions funnel you would like to see TripAdvisor go? I know it's something you've thought about over time and you've moving a little bit, but what's your latest thinking on the risks and rewards of doing that?

Sure. It's an excellent question. We always come at it from the perspective of, "how can we help the consumer have a better experience?" And so moving from classic to meta, pretty clear. Moving from meta to an assisted book or to some additional help in the booking, well there are some pros and cons. If it's assisted book, are we helping you to book a hotel that's a more expensive room than -- or a more expensive price than you can book at somewhere else. Is that helping, really? Well it might be a little bit more convenient, but we don't want to imply that the room is going to be any different because you're paying us potentially a higher price than you're paying someone else who's advertising on our site. Where I think there is good room for improvement that is pretty clear to us, at least, is on the phone side of things, where it's just not a very pleasant experience. Still, to go from our meta, we'll show you the price, to the supplier side or an online travel agency side where you're still in the, "All right, well I'm still selecting my room." And then, heaven forbid, I'm having to enter in my credit card at that point. And so a lot of friction. And we think that some of that friction is why our phone monetization is so much slower than our desktop, and we feel that there's just meaningful room for improvement. We haven't articulated how we plan to improve it, but there's -- it fits in with our goal of helping the consumer. And so that is an issue or challenge that we look to help our travelers with. Beyond that, we feel it's great to be able to offer choice to the consumer. So a price comparison that really does find you the best rates, that really does search all available partners. And I can pull back to the earlier announcement of Trip Connect as yet another way where we're going to help show you some availability that OTAs may not have or that the individual hotelier may not even -- beyond any OTA, let alone be sold out on that OTA. So again, be able to offer that direct connection for the consumers and the hoteliers that are interested, it helps the user experience.

Thank you. And at this time, I'd like to turn the call back to our CEO, Steve Kaufer, for any closing comments.

Well, very good. I certainly appreciate it, and thank everyone for joining us today. We look forward to updating everyone in just a few months. Thanks much.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude today's program, you may all disconnect. Everyone, have a great day.