All right. Good afternoon, everyone. Welcome to our next Fireside Chat keynote, here at the Morgan Stanley 2026 TMT conference. We're thrilled to have Ewout Steenbergen with us from Booking. Good to see you as always. Thanks, Brian. We're on a roll, Singapore, London, so. We've been touching them all. Yeah. We've been spanning the globe, which is very fitting with Booking Holdings. Let me do the disclosures first. Always the best part. Please know that all important disclosures, including personal holdings disclosures and Morgan Stanley disclosures, appear on the Morgan Stanley public website at www.morganstanley.com/researchdisclosures. Some of the statements made today by Booking Holdings may be considered forward-looking. These statements involve a number of risks and uncertainties that could cause actual results to differ materially. Any forward-looking statements made today by the company are based on assumptions as of today, Booking Holdings undertakes no obligation to update them. Please refer to Booking Holdings Form 10-K for a discussion of the risk factors that may impact actual results. Good to see you. Yeah, we've been doing a lot of these discussions across a lot of continents, and I feel like the agentic discussion is sort of remains on front burner. Maybe before we get to all of that and how, what the world looks like in 10 years, just remind everybody of sort of how you're thinking about the multi-year growth framework for top line margins and overall bottom line. How do you think about the next few years and the key drivers of each of those factors? Yeah. We refer to this as the eight/eight/fifteen framework. We expect to grow our top line in terms of gross bookings and revenue by 8% from a medium-term perspective, and then earnings per share by 15% from a medium-term perspective. It's a very repeatable model, and I will later on explain why it's so repeatable, but those are the medium-term targets that we're aiming for. Where were we last year? Of course, we had some benefit from FX, so let's take that, take that out. On a constant currency basis, our results were 10%/10%/18%. Clearly we exceeded in 2025 our overall medium-term targets by quite a strong margin. 2025 was a really good year for the company. Of course, we're already the largest travel company in the world, and to grow from that base, at such a high level, we were very, very pleased about that. We also gave guidance for 2026 that was actually above that framework because the guidance we gave was 9/9/15. We believe that this year, our top-line metrics on a constant currency basis actually can grow 100 basis points faster, and then still we'll be able to end up with about 15% EPS growth. Why can we grow the company in such a way and deliver those results? I think if you look at the algorithm, it's the following. In general, travel markets grow faster than GDP across the world. Why? People at some point get more income, and what they do, they do with it after they, of course, taking care of their living necessities, they start to spend on other things, and they start to spend on travel, and they want to explore the world. In general, travel grows a few points faster than GDP in general. We believe, and we are having a conviction, that we can grow faster than the market in general. Although we are already the largest player, we have a lot of idiosyncratic growth opportunities. Think about we are, at this moment, the already largest online travel platform in Asia, outside of mainland China. That's a huge growth market. We are very focused on investing in Asia and capture that growth opportunity. We see ourselves still as a challenger in the U.S. Our market position is a little bit lower. We're very focused on growing U.S., get to a fair market share in the U.S. Actually, in the second half of last year, we grew at a double-digit level in the U.S., much faster than the market in general. That's a growth driver. We're expanding in many different verticals, where we're relatively smaller, like in flights, we have grown in the 30% level last year, in attractions around the 80% level, and so on. We have so many different growth opportunities there. We're also reinvesting a big part of the savings from our transformation program in strategic initiatives, and that will help further drive top-line growth in the future. Therefore, these levels of 9% of top-line growth this year, we have some leverage from a fixed OpEx perspective, marketing perspective, gets our EBITDA at a low double-digit level. We run very large buyback programs, and you get to the mid-teens from the EPS. Again, very repeatable model, and we like that, and that's where we're very much focused on from the execution perspective. That's good color. I'm gonna dig into a few of those areas. Maybe we'll span the globe a little bit. Maybe let's start with Asia. You know, you mentioned sort of Asia being one of the investment areas. I think, you know, Glenn has talked about it as being structurally faster growing. What has changed most about your Asia go-to-market strategy in the last year or so that sort of gives you confidence you're gonna be able to drive durably faster growth throughout 2026? Yeah. In Asia, we have a number of our brands active. Our main brand in Asia is Agoda. Agoda is originally founded in Asia, is headquartered in Singapore, has a big part of the operations in Bangkok, local management. Agoda is very focused on localization. Agoda makes sure that, for example, in Korea, it presents itself as a Korean company to Korean travel customers. Mm-hmm. It has Naver Maps, it has KakaoPay, it has a UX that is very much focused on Korean users. We use brand advertising campaigns and celebrities that are local. That's just an example. The same thing is what Agoda is doing in all the other markets, and therefore it's very deeply ingrained in Asia. Booking.com is also active in Asia. It's much more the global optimized model that we're applying across the board, but it's very attractive for many categories of travelers, particularly travelers that are thinking about outbound travel. Booking also is investing in Asia, sees also clear opportunities. Let me give you an example of that. Until recently, Booking.com didn't have a domestic payment option in India. Mm. Booking was effectively only on the agency model. The agency model means you pay as traveler when you check in at the hotel. Now we have the merchant booking, where we facilitate the payment in India. That is opening up really some growth opportunities for Booking in India. Booking is also very focused on Asia. We are already, as I said, a very large player. We are very focused on maintaining our market position in Asia. Just by the growth of the market itself, that is expected certainly to grow at a very high pace. It's by far going to be the fastest-growing region in the world over the next decade, and we should be very well positioned. Okay. Let's go around to the U.S. You know, throughout our discussions, typically the way they go is I say the U.S. has been growing slower than it should. Why is it growing 1%, 2%, 3%? Now the tables have turned. You guys have been beating my numbers in the U.S., growing much faster. What has changed in the U.S., and how should we think about, you know, the drivers of this durable outsized U.S. growth? Yeah. The market in the U.S. is still growing low single digits, so that's the market. Yep. For the last half year, we have been growing low double digits in the U.S. There's a few reasons for that. One reason is we have seen some opportunities with respect to performance marketing. We're always looking at optimizing our bidding algorithms. We saw some opportunities to lean in where we saw attractive ROIs, and by really pushing there harder and spending more, we have been able to grow at attractive ROIs in the U.S. I would expect some of that to continue still for some part of this year. This year. The second reason is we have been growing well in B2B. B2B, Just to explain it quickly, this is not business travel. This is sometimes being confused. This is where you distribute through another partner. This could be, for example, a credit card company where people want to use their points to book travel, and then you are the entity that is powering behind that to make that booking of points to ultimately booking of trips through points to make that possible. It could be airlines that after you book a flight on the airline saying, "Hey, do you also need a rental car? Do you also need a hotel accommodation?" That is also B2B. We have done very well in the U.S. with B2B, so that has also been a growth driver. Third reason is, this is actually the one that makes me most excited. Mm-hmm is we have seen growth in our direct channel. Right. Direct channel means that customers, traveler customers coming directly to our web or to our app and book directly on our app, their travel. This is not something you just change overnight. This is something that comes through years of investments. Investments in brand, in familiarity. People come to us one time through a paid channel. They have a good experience. They come a second time to us through a paid channel. They have a second time a great experience. They become a member of the Genius loyalty program. Mm-hmm. They book a 3rd time, and then maybe the 4th time, they say, "You know what? Let me go directly to the Booking app and check it out." They become a direct customer. That's, of course, really the ultimate point. To see that going up in absolute terms in the U.S., the direct channel, I think is really encouraging. That's definitely something that is then, I think, positioning us very well for the foreseeable future. When did that direct kicker start coming through? Was that a fourth quarter dynamic, or is that sort of over the? A little bit mid of 2025. Mid of... It really started to change, the trajectory. Got it. Okay. Noticeably, within the U.S. driver, you don't really talk much about the alternative accommodation business. Just maybe talk to us about sort of the U.S. alternative accommodation business at this point. How do you think about key areas of growth there to sort of continue to grow that business? Yeah ... in the U.S.? Thank you for mentioning that because alternative accommodations, I think for us globally, is a very important business. We're about 75% of the size of the largest player globally. Yep. We're definitely smaller still in the U.S.. We have, if you think about alternative accommodations, about 8.6 million properties, listings on our platform. It is year-over-year up by about 8%. Actually the growth in the U.S. is faster. In the U.S., we are relatively small in alternative accommodations. We're seeing this really as an opportunity to further expand, and we're clearly investing in this. It's going well, I think we are definitely would like to close the gap more. How do you do that? Investing in product. For example, for the host, it's very important that they have direct communication options with travelers. It is in the way how we present the offerings. I think supply, adding more supply that people know that they can also come to our platform for alternative accommodation. We're investing in all of that. I think that's definitely going to continue to be helpful in the future. On that point on supply investment, how do you think about the take rates and sort of the relative take rates that you offer to hosts who want to list on Booking as opposed to the take rates that are lower on the two other alternatives in the U.S.? I think actually that is all converging now. Mm-hmm. People need to look through all the different elements. The good thing is with the fee transparency rules in the U.S., the optics of we might ask more commissions or our price is higher is very different if you look below the line. In our case, we don't add a administration fee and a cleaning fee, and a service fee, and all of that. We always had an all-in price. Mm-hmm. That has basically in the U.S. now become the standard. I think in terms of presentation of fees and commissions, it's much more aligned now between all market participants. I think that was optically a negative for us, but I think that optic negative has now disappeared. Okay. Maybe just to complete our little trip around the world, let's talk about Europe. This is your biggest, most mature business. It's still, I mean, it's still growing very healthy, especially given the size of the business. Can you give us some examples of where you still see opportunity to improve execution in Europe to keep that business growing even faster than the overall travel industry in Europe? Well, we are growing faster than the travel industry in Europe. Mm-hmm. We were last year consistently every quarter at a high single-digit growth level. I think that's really a great performance, particularly given the market position we're already in. I think what we are very focused on is. That's always important to explain that, particularly to investors in the U.S. We have a market position in Europe where people really the go-to is going to the Booking app, planning their trips on the Booking app, having it all in one place on the Booking app. It's easy to find what you need. It's easy to adjust. It's easy to cancel. Your credit card is there. It's, if you need to make an update or a change, you know that's taken care of for you. We have more and more options on our platform. Think about it in terms of alternative accommodations with us. You can get everything side by side on the platform. Maybe you're looking for a hotel in a city, but we also show you an apartment option, and people search for one and sometimes book the other. That's very natural. We have added a lot of flight options. For example, Ryanair was added as supply last year on our platform. The options get broader and broader. Because we are so much the go-to and the standard, I think, and people like it. We give them great service. We give them great value. They are part of the Genius program. They get even better value from us. I think that is ultimately the effect we're seeing in Europe. In Europe. Okay, great. What's your How do we think about the direct mix of the business in Europe versus the rest of the world? How, how big is that gap, wherever you'll say? Yeah, we're not really breaking that out. Yeah. obviously, the markets where we're stronger, that is also, a bit better. Yep. That sometimes it's underestimated what is the direct mix in other parts of the world. Even Agoda, for example, I think in general has a good direct mix that is better than people think from the outside. Interesting. Okay. Let's talk about marketing a little bit. You've been a very successful spender on Google Search for decades. Now I know over the last year or so, you've been talking more about spending on social, which, you know, is code name for spending on Facebook. Maybe talk to us about what has worked spending on social, and how do you think about the ROI on social and sort of continuing to spend on social in 2026? Yeah. It's actually much more than Facebook/Meta. Mm-hmm. We have expanded, I think, a lot with other social media platforms. Yes, it started very much with Meta originally. We set up a special environment where for us, it was important that we were able to measure incrementality. We are a very scientific-driven organization. Mm-hmm. If we spend on performance marketing, we need to be convinced that it adds to the results. Adds means it needs to have proper ROI, but it also needs to be incremental. If this would have been a booker, anyhow would have booked with us, then spending on advertising there is not a great way to use our scarce resources. That's why we ultimately have set up that environment where we can measure this. Once we're sure that it was incremental and the ROIs were really attractive, we're leaning into that. In the meantime, we have expanded structures with other social media platforms. Mm-hmm. I'm not going to tell you the names because I don't want to make some others in the industry smarter than they are. Right. We are seeing also very good results with others, so we're continuing to spend. I've said after the third quarter that the spend on social media is a couple of $100 million a quarter. Right. We said in the fourth quarter that it grew 13% year-over-year. It's a meaningful number. Realistically, if you compare to the total marketing spend for us as a company, in total marketing last year was a little bit over $8 billion. We also said that we have merchandising, which is discounting or the incentives. That shows up as a contra revenue. For us, we see it in the same category. In the end, the dollars are fungible. Where do we get the highest return? Where should we spend this? That's another $2 billion or so. We're speaking about ballpark $10 billion of marketing and merchandising spend. We're always looking for where do we find the most optimal places where to spend. Social is an important part. It's great that we are diversifying. LLMs will be diversifiers in the future, but it is still a relatively smaller part compared to the rest, what we spent in marketing. Okay, you brought up the LLMs. Let's talk about agentic a little bit. I know you get asked this question a lot. I know I get asked this question a lot about the future of agentic. I've changed my view even. I was wrong about sort of the way things have played out so far. What do you think is the most misunderstood aspect when investors or people ask you about Booking being disrupted by a horizontal agent? I'm taking out my phone. Okay. because I would like to read something. This is of course, we are in a situation with a huge conflict in the Middle East, in the Gulf region right now. For us, the most important at this moment, we're looking after our travelers. We're taking care of them, the ones that are stranded there, the people that have trips planned to the Gulf region that we need to take care as well in terms of cancellations, refunds, rebookings. This is a moment for us towards travelers and supply partners to really help them and really step in. Of course, we're doing that in the right way with force majeure and treating our travelers and supplier customers in the right way. I was thinking about this because there is this perception of, oh, some of those LLMs, if it is Gemini or ChatGPT or others, they are going to do all of that in the future. Yesterday, I asked one of those platforms, let me not mention which one. Mm-hmm. that I said, "I have a flight planned to Dubai, and obviously I can't fly there anymore. But hey, you platform, you recommended this flight to me. How are you going to help me to get a cancellation and a refund?" It gave me actually an answer that was quite unsatisfactory. Mm-hmm. I asked, I said, "Hey, but aren't you the agent that helped me?" I got the following answer. I said, "Yeah, that's a very fair question. Short answer, no. I'm not an airline agent. I can't directly cancel or refund flights to you. I don't have the ability to." Then it gave me a few elements. Then it said at the end, "Think of me as your travel strategy advisor, not the one pushing the button." I thought, this is so interesting. I am the travel strategy advisor, not the one pushing the button. Yeah. When the moment is there, that it really matters. The moment is there where a traveler needs support, they still need a party they know, they trust that is fixing things for them, and that's the moment that they realize where they want to go back in the future. To your question, there's 2 things. 1 is disintermediation. These LLMs that directly connect with the supply and the OTAs are not necessary in between growth underestimation of the complexity that goes in between in terms of fulfillment, in terms of customer service, in terms of payments, in terms of dealing with tax matters there, with regulatory matters, with data privacy matters, to really help the supply partners being able in control and understand what they're doing, how they're showing up, the visibility, the global marketing we're getting. The most important thing, as I just highlighted, I mean, we are there when people need their help and they can trust, and they know who to call. If things go wrong, they know who to call and who is there for them. I think that is why I don't think disintermediation is really a risk at all. The other part, I presume you want to go there, is direct and the impact on direct mix. All right. I think that's probably a more realistic risk. Yep. obviously we are strategically very focused on that to mitigate that risk in the future. How do you do that? You know, what safeguards do you have in place with these early partnerships with any of the horizontal agents, we don't need to name any of them. We can if you want to. so that you have the data capture that you maintain merchant of record. How are you structuring these so that you can continue to grow your direct and repeat business, for second, third, fourth transactions? Yeah. Well, first of all, if we look at the numbers now, the traffic that comes through the LLMs to us is very small. Yeah. It's not actually, it's not growing. I think that's for me, a bit of the surprise. Over the last few months, it's more or less stable. Some months are slightly higher, other months slightly lower. Okay. There's not really an underlying trend. I think most people are using the LLMs to do research, to get inspired, to maybe build an itinerary, but are still doing research from many other places, and ultimately they go to the place they know and trust for ultimately the booking. That could change over time. I'm not saying that that is always going to be like that. What should we do? Because about two-thirds of B2C travelers are coming direct to us. Two-thirds of traveler customers are coming direct to us. For us, it's really important that they continue to come direct to us in the future and not start with an LLM. Ultimately, yeah, it's going to be maybe a lead that is handed over to us, but the LLM becomes a paid channel and therefore our direct mix is coming down. We do that by developing agentic tools on our platforms that are as good as the agentic tools that people will find on horizontal agents. Mm-hmm. We're working on many different variants of that in different parts of the organization. All our brands are working on it. We have two startups within the organization that are focused on that. To create an agentic experience top of funnel. That is as good as you can find as a consumer on a model of a hyperscaler. In an environment you know you trust, but also an environment where it doesn't stop with the booking because those horizontal agents then ultimately have to give it to someone else. Mm-hmm. You continue in our environment. We give you an agentic experience through the whole funnel, pre-trip, in-trip, post-trip. We can help you with preparation for your trip, your passport, your visa, your check-ins. We can help you with things that are happening, changes in your program, fill in some gaps in your... We can make it very convenient, peace of mind. If something needs to be adjusted, everything is being updated at the same time. We can make it personalized because we know what you like and what you usually like to do. We can put enhanced loyalty on top of that, and we're working on that as well. We can create an agentic experience where everything is logically fitting together, everything is aligned to each other, and there's so many ways that that can be, can be done. For example, you travel with a child, and automatically the child seat is being added to the rental car, and the child bed is automatically already being added to the hotel. We know your child has an allergy, so the restaurant is already informed about it. I mean. Mm-hmm ... there's so many ways that you can build that peace of mind intelligence layer on top of it. That is ultimately the experience we're working on. Mm-hmm ... over time, and I think that will be something that will make it very attractive for travelers to continue to come to our environment. I would even go the other way around. I see this as a real opportunity for us as a company to grow even faster. This is not a threat. This is an opportunity. Mm-hmm. This would mean that it is so much easier to put everything together in that environment with us. If you do it in different places, now you have the hassle that everything needs to be updated. You need to contact so many different providers. If that's all in one environment and it's peace of mind, I think that's a really strong proposition. You guys have a lot of leading travel data, leading travel data sets of, you know, millions or billions of daily searches. Really, the value of the outputs from these models and the agents is often the data. You should be positioned to do that. My question is, what is sort of the gating factor, the hurdle that's gonna determine if that product comes out this year or next year? Yeah. We are not really a company that is doing any flashy demos and videos of things that are coming. We are more a down-to-earth company. We say, "You see it in our numbers, you see it in our results, and you see some of our agentic tools showing up. Once you see it and you see it on our platforms, then that will tell the story to you." That's how we operate. That's philosophically who we are and what we do. You should expect some progress in these tools this year. I think, definitely if we're here together next year, I think we have a different kind of conversation. There are a couple of good examples we already have. I think if you look at a tool that is called Penny on the Priceline platform, that's a really very advanced already travel tool, AI travel tool that we have. Agoda is working on a variant of that. Booking.com is working on it. As I said, for example, even OpenTable has already an AI concierge on top of it. There are many of those agentic developments we are doing. More to come, but I think definitely when we're here in a year from now, I think you will be able to ask me more questions, and I can point to more examples. We'll use it on stage. I'll pull my phone out and use it on stage. Okay. On gen AI and internal sources of efficiency, this is the other area where you've been pretty vocal about. Maybe, you know, there's a lot of focus on ROI of all these investments. Where are you seeing quantifiable savings from GPU-enabled products or gen AI in the P&L that we can look for in 2026? Yeah. I'm actually really proud already that in 2025, I can point to real concrete savings that have fallen through to the bottom line, and that is mostly from customer service. If you think about customer service, last year in absolute spend, it was slightly below 2024, and volumes were up about 10% year-over-year. We have introduced a lot of agentic tools in customer service. What we're seeing is that the average handling time is coming down. We see the contact rate is also lower, contact rate in the sense of where a human agent needs to be involved. The average cost per booking is down about 10%, as I just explained. By the way, customer service is holding up very well. Customer satisfaction scores, I mean, is holding up very well, which is of course also really crucial with those numbers. I think that's not done in customer service. Mm-hmm. I think we have more potential there in 2026. Obviously, we're also looking at Gen AI savings in other areas. A little bit too early to go in detail of that, but I think we're very focused. We're very disciplined. I think you have seen this in the overall transformation program. We're already Yep ... at the end of last year. Well, in the third quarter, we increased the targets by $100 million to a range of $500 million-$550 million, and at the end of the year, we already achieved that. We will definitely be focused on what is next to come in that area. Maybe remind everyone about the transformation math because I think there's a lot of expectation that you will reinvest in Gen AI. Getting the agent out that I'm gonna use on stage next year, there's obviously investment behind there. Just remind us of how we think about the gross savings and the net savings after the reinvestment. Yeah. In 2025, we for the first time introduced our investment program, or we call it reinvestment program, because it is self-funded in the end from the transformation program savings. $170 million, and some of those initiatives were the first year, so they were in startup phase. We have scaled this up in 2026 to a reinvestment program of $700 million. Mm-hmm. Definitely at a higher scale, higher level. We think that's the right thing to do with a business that has EBITDA margins at a level of approximately 37%, fully loaded, including stock-based compensation and so on, around 37%. Growing the company as fast as we can is the most value-enhancing way we can manage the company. That's what we are focused on, and those reinvestments are going to help us. That's why we're confident that those growth objectives for this year are at the 9% level instead of at the 8% level that we normally would say. The $700 million investments will lead to approximately $400 million additional revenues this year, so that helps with that 1% growth. Got it. The net impact on the bottom line is approximately $300 million. We have about $250 million additional in-year savings from the transformation program. Yep. Basically, we self-fund this and can become a faster-growing company. I think it's a very positive, very motivating. Mm-hmm ... overall way how we run the company at this moment. Yeah. Before I open it up to the audience for Q&A, if anyone has a question, raise your hand. I want to ask you about AI hype. You know, I'm sure you've been investing in a lot of different types of generative AI tools internally. Can you give us some examples of things that have actually proven to be more challenging than you thought, or areas that are just overhyped externally? We have a lot of discussion with our CTOs. Mm-hmm. We probably see all these companies announcing, oh, 30%-50% more of code is written by Gen AI. Right. Which is probably true, but there's also very much a nuance to it. Because the nuance is that doesn't mean that every developer is 50% more productive, because there's many other parts of the whole development stage that goes into it as well. It isn't that that is all more productive. The other thing is we, in that area, deliberately want to use that additional capacity and velocity to do more development in the future. That is one area, for example, where I think some of the numbers are overhyped. Mm-hmm. The reality is actually slightly different. It's still an opportunity, and of course, we're going after that, but I think the optics of those numbers are sometimes, I think, a little bit too much. Question for you, Bob. This is an inventory, an inventory question, I ask this question as a Genius level three user of Booking, by the way. Thank you so much. You seem to have locked up an enormous amount of inventory, but here's the Google question. Google feels like it's this terrifying threat. You know, they put this advert out just before Thanksgiving last year. Turkey wants to go away from the United States, and it just... I think it got a lot of attention. How does Google take that inventory directly? Can they take that inventory directly that would either undercut you because they're offering something else to the person with the inventory, the hotel with the inventory? Can you just help me understand why you can continue to lock somebody like Google out? Yeah. A couple of important data points around this. We have 4.4 million properties that we connect to today, and we need to have real-time data with those properties. We need to know real-time availability. You don't want to say, "Oh, I booked this hotel room in this property," and once you press book, you say, "Oh, it's not there. It's not available." This is something that we have built over the years. 90%, close to 90% of the room nights that are booked on our platforms are coming from independent hotels, alternative accommodations, and smaller chains. We said the top 10 chains in the world is about low double-digit level of room nights. The vast majority is independent hotels, alternative accommodations, and the smaller chains. Often to investors in the U.S., I need to explain that because they think that the U.S. hotel market, that that is the standard in the world because it's all about the big chains. Actually, the U.S. is more the exception than the rule around the world. I think what we do for those properties is a lot of that supplier value add that I explained before. There is a lot that goes in the mix. It is not just communications between us and that property in terms of data, intelligence, the whole fulfillment, the customer service. We process payments in over 100 different payment forms across the world and 100 different, 50 different, more than 50 different currencies. We help them to be in control of the data, how they show up, when they show up, what kind of promotions they do. There's a lot that goes in the mix there with it. Google, I don't think it's so much different than in the past. They are a very good partner for us. We are a very good partner for them. We are, I think, one of their largest customers in the sense of how much we spend with them. It's great for them. They have really been able, their advertising model, to really optimize it. It's good for us, it's good for the consumer. I think all three have really done very well over the last decade in that environment. We very much believe that that will continue in the future. They have, by the way, said publicly that they are not planning to be an OTA. They do not want to be a merchant of record. They will continue to be a lead generator for us, and we like that. We have fantastic algorithms to do the right bidding on their platforms and get the traffic to us. Therefore, we can continue to be, I think, very good partners together. No, I am not concerned about this. I mean, you could say if Google wanted to do it, they could have done this already many years ago. In the end, we found a very good way that was beneficial for everyone. Great. All right. Time for one more before we get the hook. Thank you. How do you respond to the risk that over time, the horizontal agents, while they may source the inventory that you've aggregated over time, the consumer stays on the agent, and it's harder and harder to get the direct traffic and maybe that direct mix, as Brian highlighted in his recent note, starts to plateau or maybe come down? Yeah. I think that's very similar to what Brian already asked. I actually don't see this as a threat. I see this as an opportunity. We have been talking for a decade about what we call the connected trip. Gen AI, agentic experience of connected trip is one and the same. This is the same thing. We can now make that possible to create that connected trip experience for a traveler. We can naturally put all those pieces together. We can help them to get inspired, to make it very personalized. We can make sure that there's a logical connection to it. We can say, "Hey, next week you have your trip, but it's bad weather on Thursday. We should probably change the itinerary and make sure that you don't have the outside tour on Thursday, and we do that on another day. Press here. We can do the integration with OpenTable and say, "Hey, you have no restaurant reservation. We see that evening. Press here." We can make it proactive. We can make it personalized. I think this will be an experience that is something that doesn't exist today, that is further enhanced with loyalty benefits that we're going to provide to travelers. If you add that up, I think that is such a great experience. Yes, we are working very closely with the LLM providers, with the hyperscalers. We're a launch partner with OpenAI. We have a very close relationship with Google. We are also working together with Microsoft and Amazon and many others. We see them as ultimately, enhancements and diversification of pay channels. We believe we can develop an experience for travelers that come direct to us, that they will continue to come direct to us, even more often come direct to us, and even do more of their trip with us in the future. All right, great. Ewout, well, we are against the hook. Thank you as always. We're anxious to see the other business continue to change and the direct business keeps growing the next 12 months. Thank you. Thank you so much. Thanks.