trivago N.V. (TRVG) Earnings Call Transcript & Summary

The next session of the Deutsche Bank Technology Conference. I'm Lloyd Walmsley, the Internet Analyst here at DB. We're excited to start the next session with Matthias Tillmann, the CFO of trivago. Matthias -- sad, we can't do this in-person as we have previous years, but glad to have you joining our virtual conference. Thanks for being here.

Yes, absolutely. Thank you, Lloyd.

So maybe we can start with an update on travel trends you've been seeing. Any signs the recovery continues to improve in August? Or has this kind of spike in cases in certain regions slowed down the recovery? You guys had talked about Europe and rest of world recovering a bit faster than the Americas in July. Kind of -- any changes regionally? What are we seeing right now?

Yes, absolutely. I mean, Europe recovered nicely in July as the seasonal summer peak kicked in. Seasonality wasn't much different compared to what we usually see during this time of the year. But there were some shifts in demand that we have talked about before, like people looking more for local destinations and a shift towards alternative accommodations. Beginning of August, volumes peaked, which is also in line with the historical seasonality. Having said that, overall volumes are still down significantly year-over-year, as we called out during our 2Q earnings call. On the other hand, in the U.S., absolute traffic volumes plateaued for us in July. While in Lat Am, we did not see any meaningful recovery. In August, week-over-week volumes were holding up against the normal seasonal pattern in the U.S., and we also saw some signs of recovery in Lat Am in the second half of the month. So clearly, positive what we saw there. In general, I mean, as you said, new cases continue to have an impact on travel behavior. For example, when cases started to go up in Spain in August, and Germany issued a travel warning for this country, including the Balearic Islands. We did not only observe a decrease in volumes in our point of sale in Spain, but also in Germany. So in essence, demand continues to be heavily affected by the overall situation, for sure.

Yes. Okay. And have cancellation rates kind of compared to new booking levels started to -- have you seen any stabilization there as we come out of this?

Yes. The first thing I need to mention is that we do not have full visibility on that. So obviously, we saw advertisers bidding down across the board, which was also related to higher cancellation rates. And since the end of April, we have seen higher engagement from our advertisers again, which could also be an indication that cancellation rates were coming down and are closer to historical levels. And looking at the part of our traffic, where we do have visibility, there's a clear indication that cancellation rates peaked in April. This is also consistent with the feedback we are receiving from our partners. What I would also mention is that the overall travel behavior has changed now. So the time to travel window has decreased. And usually, the cancellation rate for those trips is lower than, let's say, when someone is looking for a trip couple of months out. So if you take the U.S. as an example, in July, the average time to travel was less than 30 days, whereas normally it is higher than that, at least for the traffic we see on our platform.

Okay. And on RPQR, you all had called out last earnings period, it was down more than 50% in July. How is this recovering? Should we think about the recovery rate varying by region?

Yes. We called that out in our last earnings call, so the more than 50%, that's correct, and there's not much I can add here. I mean, since end of July year-on-year rates rather stabilized, and this holds true for all regions. But I would also be careful when looking at this metric, let me remind you that RPQR, our revenue per qualified referral is the product of the booking conversion, the average booking value and our monetization. And given the shift in demand, for example, higher share of domestic destinations and alternative accommodations, there are some mix effects in there. But the biggest driver, for sure, is our monetization, which is still down significantly, year-over-year. The uncertainty on future cancellation rates plays a role here, in particular, for searches with longer time to travel windows. And secondly, some advertisers are only now returning to the auction, activating their campaigns again. So it will take a bit more time until the auction is back to pre-COVID levels.

So how do you kind of describe your visibility into the holiday travel season at this point? You mentioned that the booking windows are still quite narrow. Is there anything in the search queries you're seeing that would give you any increase in visibility towards holiday travel?

Yes. I think there are 3 clear trends we saw in July and August this year. First of all, clearly, higher demand for local travel destinations. And I mean, many, many companies in our space talked about that. And intuitively, that makes sense. People feel more comfortable traveling domestically or locally. That's clearly one thing we saw as well. Then secondly, as you mentioned, shorter time to travel windows, which also makes sense. I mean, if your trip got canceled because the flight was not taking off or whatever it was, and you are looking for something short-term to get away. That is clearly something that we saw. And then thirdly, and we have also talked about it before, as have many others, a shift towards alternative accommodation. So those are for me the clear -- 3 clear trends we have seen this summer. In addition, there was a lot of demand for city trips and, obviously, international travel as well. And those are trends we expect to continue to see for the remainder of the year as well. And clearly, there's lower demand for business trips, which usually pick up after the summer vacation period. So this would be now. And there, let's see -- but from talking to different companies, many companies are still not back in a fully office setup. And I would expect that to happen first before companies seriously think about all kind of city trips. And then, I mean, the fact that we are now talking virtually and having this conference virtually also shows that in many parts of the world, trade fairs are not happening, conferences are not happening. And obviously, a lot of the business travel shares is related to that as well. So let's see, it's early in September, but I would expect that to be much lower, at least for the remaining of this year.

Yes. Okay. If I had a crystal ball and said that travel would recover to 2019 levels by the end of '21, say, when do you think digital advertising in the sector would recover? Do you think it's concurrently? Is it on a big lag? Are there any structural changes that you see in travel advertising? What would your perspective be on that?

Yes. I think for digital advertising, I would say, it's instant. So the dollars usually follow the pocket of demand. So when volumes pick up in paid channels, then surely advertisers are happy to pay for that traffic, in particular, if it's high-quality traffic, I mean, that's the 1 precondition. I think it's different for traditional offline brand marketing channels. Because there you need to require an upfront investment that doesn't pay off immediately. And for that, I would expect many companies want to wait and get more comfortable that the situation has stabilized. As if you start investing too early and then things worsen again, you do not benefit from the branding effect at all. So there's a clear difference. But yes, for digital advertising, I don't think there's a big lag or there will be a big lag.

Okay. And I guess if you guys said you started to dip back in the marketing channels, starting in June, mostly performance but also some bits of TV in markets like Germany. Are you guys continuing to scale that spend up?

Yes, that's correct. We tested in most European countries. And indeed, we're happy with the response we got. Obviously, overall, on a much lower level than what we historically did. But the relative response was good. Now that we are going into off-season again and knowing that business and city trips will lag behind. We will be a lot more mindful with our TV spend. We might opportunistically invest or run certain tests, but you should not expect us being constantly on air everywhere. And also keep in mind that historically, we have always decreased our brand marketing spend in the fourth quarter significantly regardless, yes? So that is not a quarter where we usually spend a lot of money on TV advertisement. So that's where we are right now. Obviously, it will be interesting to see how things shake out towards the end of the fourth quarter because that will also determine how we start the new year. And there, usually, after Christmas, we start going in and increase our spend again. We will certainly use the learnings from more creative that we got right now, that will be helpful, and the team is working right now on fine-tuning that and taking the learnings in. With regard to the media spend, I think, there, we will take our time and assess the situation in a couple of weeks from now.

Okay. And you called out, I guess, alternative accommodations, exceeding 20% of referrals. How does the RPQR in that side compare to the core lodging business? And I guess structurally, is there more or less competition for that traffic?

Yes, that's a good question, especially given that we see a shift towards alternative accommodations. But it really depends, and it is difficult to make a general statement here. By definition, part of that inventory is unique. And obviously, on that part of the inventory, there is less competition, especially if it happens to be located in a destination with very little alternatives. So like the -- a small skiing village in the alps, there, you might not even have a big hotel. You might have a few apartments only, and then your options are very limited. So obviously, there, the competition is more. Do I want that apartment in that village? Or is there another nice place to stay in the village next to it? So it's different levels and dimension of competition there. But also, I mean even if there might be less competition, the overall basket value could be higher for this type of inventory, which then, again, has an offsetting effect on RPQR. So as I said, I mean, competition or the monetization is one part that influences RPQR, but another is the average booking value, which tend to be higher for that type of inventory. So again, it's very difficult to make a general comment here, but those are the puts and takes I would mention.

And I guess sticking on this product alternative accommodations, you've seen a big pickup in the percent of referrals here, largely a function of the time period we're in. Do you feel like the product is where it needs to be? Is it still something you're investing a lot in, in terms of getting the product right or getting supply on board? Or do you feel like it's now fully baked, and it's just a question of getting your customers aware that you offer it and using it more? How do we think about that?

Yes. No, I don't think we are at the end of our product development. I mean, we mentioned it before. So far, our focus has been on onboarding the inventory, getting advertisers onboard and filling it up. If you look at our product right now, just anecdotal evidence, go to our website, and you see it's still mostly about hotels, yes? So you can see it in our tech line, you can see it in our search bar. You can see it in the filters. So it's not that you go to our website and you think, wow, this is the best place to search for vacation rentals. And that is also not our intention, yes? So when we went into that opportunity, we said, first of all, we want to have it in one search. So we integrated it. There's not a separate product for that. But you see that in one search along hotels -- hotel items and it's more offering you an alternative when it could make sense, yes? And again, as I said, I think on the inventory side, we have come a long way. I believe right now, we have over 3.8 million units on our platform. And we said it before. Now it's more the time to build additional features and help users to navigate through that, but it won't be an aggressive push, but it's more gradual and incremental, and we also learn how the users react to it, yes? And this could be different by travel type and by use case, yes? Let's say, someone comes to our website through the search term hotel in New York, then for sure, our list would show more hotels. But then that user engages with our platform and clicks maybe on an apartment that we show in the mix, then we would dynamically update our search results and show you more of the apartments. This is kind of how we think about it. And then we -- going forward, we will build additional features. Another example is the carousel that we are testing right now. So there, we are trying to cluster similar apartments in one. Again, for the user to make it more convenient and easier to determine whether that's something he or she is interested in and should explore further or not. And then you don't need to see 20 different apartments for that. But you see 1. And if you engage with it, then that's good. We show you more, if not, then we show you something different.

So we're getting a relevant question from the audience right now, which is what share of referrals did alternative accommodations make up pre-COVID?

Yes, that is indeed a relevant question. So we never disclosed it. So in Q2, it's the first time we disclosed it, and we said for the first time it exceeded 20%. But let me give you some history. So before 2017, we didn't have any of the inventory on our platform. So there the share was close to 0, except for a few items here and there. And then end of 2017, we only started to consciously invest into that category. And then it took some time to build it up and now 3 years later, the share, obviously, it also benefited from the COVID situation as we saw a shift in demand towards that category, but it was above 20% for the first time. So between then and today, I mean, obviously, it hasn't been a straight line. But yes, it's -- that's roughly how to think about it. So we got literally from 0 or close to 0, end of '17 to now over 20% in the second quarter.

So has the local product moved out of beta? Can you tell us a bit about the product and how receptions by users and how it may be impacting monetization?

Sure. I mean, it's still early stages. And for sure, too early to talk about how it's impacting any of our key metrics. And at this point, we're also not putting any marketing dollars behind it. But it is available in the app and on the website as a separate tab. So the key idea is to present you with options and great deals for local travel destinations. So it is more about inspiring the user rather than comparing your options when you already know where you want to go. And the way we want to do this -- to do it is by using technology, not having u read tons of reviews. So there are different use cases for this. It could be about destinations itself, for example, like nature with beautiful hiking path, beach locations, mountains, et cetera. It could also be about an event that is happening. Obviously, which is not as relevant right now, but selectively, it will be more relevant again going forward. And all it could also be just about the accommodation, like a nice bar hotel, an accommodation that is perfect for families, et cetera. So as a next step. So it's live and globally, and we do see some traffic on it. And as the next step, we will work on improving the search results and also then adding destination content, which is quite relevant for that product. And make it also relevant, even if you don't want to travel or can't travel right now. And the focus here is not on conversion but interaction so that we increase the frequency of our product.

So if we think back to kind of the time of the IPO, you all were definitely not interested in kind of reviews generally or having people spend time on the website beyond your core price comparison and conversion functionality. Has that attitude changed a little bit since the IPO in terms of just having people spend more time on the site? Moving perhaps a little bit more top of funnel, how do you philosophically think about that?

Yes, I think a very, very fair and relevant question. And I think the one thing that hasn't changed is that -- is our philosophy, and that is to provide tech solutions for our users, yes? And that's why I also said before, we don't try to build now an inspirational product where you get a lot of content, read a lot of content, et cetera. We believe that technology can do that for you. So obviously, we need the data. We need the information, and selectively also have to give it to you so that you can decide for yourself and you get an idea. But then we want to use technology to help you find out what's good and relevant for you. And there, the use cases we are working on right now, like the local travel product, for sure, they are a bit high up in the funnel and more for users that don't know exactly where to go and what to do and are not ready to convert, that is something that at the time of the IPO was too far away for us, and we are moving in that direction. But again, doing that very mindful and using technology to build products that have you achieve that.

I've got another question coming in from the audience. Just to try to clarify some of the comments from earlier when you said things kind of peaked in July in the U.S. and then followed typical August seasonality. Can you elaborate a little bit on what that means, perhaps relative to comments on the EU?

Yes. Absolutely. Well, first of all, so the comment on peak in July in the summer peak, there was -- there I referred to Europe, not to the U.S. So usually, when we look at travel demand on our platform, historically, we see that the summer peak or the summer seasonality starts kicking in, in around June, and then it builds up. And then usually, end of July, beginning of August is, when we see the peak in travel demand on our website. And then it slowly starts to decrease again until end of August, and then kind of the peak season is over for us. That's typically how we see it in Europe and similar in Americas and the U.S. and I said, that this year, we saw the same pattern in Europe. So demand increased towards the end of July and then peaked around end of July, beginning of August and then we started to see a decrease. But yes, that's also what you would expect. So there was, in a way, no shift in seasonality. Obviously, if you look at the year-over-year, that all happened on a lower base, yes? So in general, traffic was down quite significantly year-over-year. But the overall pattern is quite similar. And if you look back at my comments from the second quarter earnings call, I gave some numbers on how July went. So this gives you a bit of an idea how then August went in comparison to that. So now specifically on the U.S. or on Americas segment, there my comment was that we didn't see the same peak in July in the U.S., but in July, the U.S. in terms of travel demand plateaued. And on the other hand, we didn't see the seasonal decrease in August then, but it stayed roughly on that level. So it was more stable. And that just against the normal seasonality, it was obviously then more positive than what we anticipated. And then I also made a comment on Lat Am. So there, we didn't see any recovery in June, July. And that now in the second half of August has started to show some signs of recovery, which then also added to a bit more positive picture, if you just look at the seasonal pattern in that segment.