Good. Right. Okay. Good morning, everyone. My name is Adrien de Saint-Hilaire. I work at Bank of America. I've got the great pleasure of leading our European Media Research Department. It is my great pleasure to be welcoming Arthur Sadoun. The pleasure is mine. Thank you. Great to see you guys. I think it's the first time you come to this event. We're very, very happy. That's the first, yeah. To have you there. We've got Arthur for about 40 minutes. If you don't mind, maybe, Arthur, we can kick off with some questions of mine. Maybe what I suggest. As long as we don't talk about current trading. I'm kidding. I'm kidding. I'm kidding. I'm kidding. Let's kick off the current trading questions to start off with. Then we can broaden out maybe to the longer-term discussion if that's okay with you, Arthur. To whatever you want. Fantastic. Talking about trading and current business trends, at your last results, you slightly raised your full-year guidance. It's been implied that H2 would be a bit slower than H1. I think you said that your guidance would assume that there would be some outcuts maybe sometimes later in the year. Where do we stand on that? I think we can fairly say that we had a good summer. Summer has been good. What we have experienced is, first, the marketing cuts that we have discussed in Q2 did not happen, did not materialize so far, which is a good sign. Second, it's very interesting to see that we have more AI projects at Sapiens every day. Don't get me wrong. It's still the strategic phase, which is not where there is material impact on the revenue, but we can see a cadence. By the way, although it has been a very busy first part of the year in new business, we have continued to win over the summer. When you add all of this, actually, you look at the underlying business between H1 and H2, and there is no deceleration. There is a comparable that is tougher that you need to take into consideration, but the business is still on the same pace. We have to be a bit cautious because we just have the first months of the second half, but we feel very confident that the five now is very solid, to make a long story short. Yeah, that's very clear. That's good, no? I think that's great information. Can you perhaps talk about the pitching activity of late? Because you've been very, very successful in the first part of the year. I think you also touched on the fact that maybe pitches could be a bit lower in terms of intensity in the second half. We see, I mean, again, H1 has been historically high for us. We won, I guess, 70% of the pitches and 100% of the big one. By the way, I just have to say that once for all, which is the big one we won in H1, won't have impact before next year. We won't see any impact for the moment. It has been an amazing time. I think it has been at the crossroads of two things, which is best model, best people, best tech, basically, on one side. Second, a single focus on our client while some of our competitors are pretty busy doing other things at the moment. That was awesome. This historically high track record, I don't know if it's going to replicate in H2. What I do know is that we have a good momentum at the moment. I would say the pitch rate is closer from what we have experienced last year than what we have experienced in H1 this year, which was, again, a very particular situation. No real slowdown on new business. Again, some wins, some that are public, some that are not, but feeling good about that. Just so that we tackle any potential downside risk, is there any major account that you're currently defending and that there could be a risk of maybe a loss at some point? At the moment we're talking, no. We are in a business where I could have a call on my phone that has been charging at the moment that will change what I'm telling you. What is sure is that if you look at 12 months, there could not be any impact of big losses that will come onto our numbers. I would have to receive a call when I leave that would tell me that maybe there will be a pitch. I will have to pitch. I will have to lose. I will have to transition. This will take 12 months before we see that. I thank you for asking the question because I think that what the market is not understanding enough, at least, is that where I think we are doing a great job is not in new business. We are number one by far, as you know. It's in client retention. It's funny because if you look at H2 last year, some of our competitors have major wins. Some of you, by the way, started to think there is a revival here. You look at the number this year, and you don't see in the number the wins. Basically, for two reasons. Maybe they have given away too much, which we don't do. We don't buy market share. This is why, by the way, there are some pitches. Talking about big pitches, there are some pitches last year that we didn't take for this reason. That's number one. Second, because they are losing. If you ask me, what keeps me awake at night is not to win a new client, even though we are winning more than anyone else, it is to make sure that we bring to our existing client the best model. I guess if I was an investor, this is what I would look first, which is, are we doing what is right for our current clients, so much so that they don't even want to go and see outside if it means something for them? Yeah, understood. I think we build on that later. No doubt. Maybe if I can touch on one part of the business which hasn't exactly perhaps lived up to expectations, or maybe that's been a little disappointing, let's say, is Sapiens. Q2 was actually better from your perspective. Is that the start of a proper recovery or are you still seeing things touch and go here? I mean, first of all, this kind of wait-and-see attitude that has been described by all of our competitors on Sapiens, which are basically the Accenture and Capgemini of the world, is still here. Clients, and I guess you see that over everywhere, are kind of reluctant to spend a big CapEx. Everyone is excited about the promise of AI. Before you move again from the strategic phase to the implementation phase, there is a lot of money that needs to be spent. We still see that. I want to be, again, careful. What we are seeing at the moment is pretty encouraging. We are seeing, as I said, more AI projects going through. We believe that Q1 was definitely the bottom of the swimming pool, which is a good thing. Sorry, that's a French expression. I don't know if you say that in English, but you understand. That's the lowest we could go for sure. If you ask me what we see at the moment, because of the reason of the momentum coming back slowly, we see a Q3 that will improve versus H1. It gives you an indication. We are one month into the quarter. It's project-based. You need to remember that Sapiens is roughly 15% of our revenue. Whether you have new projects that start earlier, it has an impact. I won't tell you that I'm ecstatic about what I see. I think that with the capabilities we have, it should already be going double digit as it did in the past. We are seeing slowly, and hopefully my competitors will see the same and stay the same, we are seeing that project. I was telling you, you know we're having our COMEX at the moment in New York, so I left it to come with you. I should thank Maurice Lévy every day for having bought 12 years ago a tech consulting that is going to be able to do true AI integration for all of our clients. What's absolutely certain is that we can talk hours about the potential of AI, and I can show you a super UX of what we can do with production, blah, blah, blah, blah, blah, blah, blah. If our clients do not build all of this AI machine on the right tech foundation, it will never work. It will be built on sand. We are in a unique position with Sapiens to help every of our clients truly transform their marketing. If you add to that the fact that Sapiens is a business that doesn't do outsourcing, what is going to be disrupted is, of course, this kind of heavy headcount business in transformation. We don't have this part. We just have AI and engineers that can bring our clients through new products into the AI transformation. We are fit in terms of structure for what AI can do for our clients. We are at the heart of all of our clients speaking about transformation with our Sapiens people. I have never felt so confident about the power of Sapiens as long as clients realize that they now have to move from architecting what they need to do to putting the money and doing it, which, again, I don't want to be too optimistic, and things have changed again the last days. I think clients start to stabilize. If you look at where they are today, they just have two main concerns. Concern number one is how I'm going to deal with the tariff, how I'm going to deal with the macro situation, and how I'm going to deal with my own industry that is being disturbed. That's one thing. On the other side, you have, is AI an empowerment or a challenge for me? We are in a unique place because of the trust, because of our capabilities, because of our model, because of our people to actually help them do that. I want to stay very cautious on how fast Sapiens can take over, not take over, take off. What is certain is we will see an improvement in Q3 versus H1, for sure. We feel better about the number of projects. We feel super confident when we look at what we can bring and the structure we've got. We have to work. Interesting. Very oftentimes, I think people refer to Epsilon as being a key success factor in pitches, but maybe not so Sapiens. Is Sapiens something that is being used in day-to-day pitching? It's funny you said that because it's exactly I shouldn't share that. I hope there is no competition on the line. I think that one of the big reasons why we've been winning in media for so many years is we've been saying, look, it's great to have scale in media. We have it as competitors. It's great to have access to a wealth of data through the platform. We have it as the competitors. What we have on the top of that is Epsilon with identity that allows you to anchor all the source of data you have into a single persona and connect this persona not only with the media, with the content, and measure it if you want to make a long story short. That was the pitch. That was prior to AI. Yeah. Because now that you have AI, you can do two things. First, you can superpower what we do with Epsilon. A good example of that is what we do with influencer. Second, it's not enough to have the data, and we are leading on that. It's not enough to have the capabilities, and we are leading on that. You need the tech expert that will build the tech foundation. That's another area where we are the only one to have it end-to-end, I would say, with Accenture. Accenture doesn't do the media. Yeah, yeah, yeah. That's it. For the moment, maybe the challenge. Maybe that's a question that I will be asking potentially a bit. That will be fun. Okay, can we maybe switch gears a bit and move to another activity of Publicis, which is doing actually very well, but generally speaking, has been perceived as at risk from AI disruption, which is the creative part of your business? I think you mentioned that it grew high single digit organic in the second quarter. What's the driver behind this? The common view in general is that AI is leading to price deflation, fee pressure, etc How are you able to drive that high single digit? If you don't mind, I'm going to take a step back. Because the question that is behind your question, if again, I read what I've read in terms of reports, and by the way, how I've seen the stock react in the last quarters, is there is a common thought that there will be industry that will be AI winner and other that will be AI loser. I think this is completely wrong. Yeah. I think you will be AI winner and AI loser in every industry. It's not that I think. You just have to look around. I mean, financial services, you already start to see who's going to be the AI winner. Tech, I mean, tech companies that you saw would be thriving forever are starting to have problems. Others that were maybe more in the midst are doing well. The question is, are we winning or are we losing with AI? That's the big question. Yeah. It is going to be true for every industry. I think that when you look at the marketing services, what you need to take out, and I'm going to come back to your point, what you need to take out of our result in H1 is that we are clearly winning with AI. I'm not saying we're going to be an AI winner. It's too early to say. Today, we are winning with AI. How do you think we get 500 basis points of gap with our competition on growth and 300 basis points on margin if it's not by our ability to take what we were doing well and beat the competition and accelerate it through AI? Why do you think we are winning all of our media pitches? Maybe you remember two years ago, we launched CoreAI that was based on the $12 billion we invested in data and technology. 100% of our win, and 100% of the reason why we win is because we put CoreAI at the center of our media platform. Why are we growing almost double digits on creative when the rest of the industry is down? Our AI production platform growing double digit, representing a third of our creative. What is very interesting for people to know here is that creative for us is only 25%. Yeah. Out of this 25%, 8% is production, which means that we are exposed to storytelling, which is definitely the place where there would be the most disruption, only by 17%. Yeah. It's a third of our competition. Yeah, yeah. Which means that, to come back to your question, the reason why we are winning is twofold: AI production platform and our ability to win market share. To be clear, no one so far has seen the impact of AI on their storytelling business. If there is a decline somewhere else, it's because they are losing market share. It doesn't mean that it's not coming. Yeah. This is why we feel good about the 18% that we will have to reboot like everyone else. That is, and I can go on and on on the reason why we are winning with AI. Another one, for example, is when you look at M&A. Yeah. I mean, we spent $2 billion. They are growing 20% on average at the moment. Why are they growing way more than they were growing before acquisition? Simply because thanks to AI, we've been able to connect them to our data. It is particularly true for creative because this is where you see the biggest gap. AI is helping us to win. It is what is in our number. It is particularly true for our creative business. Great. Maybe a bit of a funny, sensitive question, which is. I love sensitive questions. Tough questions and sensitive questions are what I like. Since Publicis and all the agencies are people-based business, that's why it's maybe a bit sensitive. Do you see that there is an opportunity around further margin expansion thanks to AI implementation in the future? Of course. The question is, how much are we going to have to invest before it delivers the savings? That is true for everyone. If you ask me, there is a reason why we have actually increased our margin despite all the investment we have made this year. We are basically the only one who still pay bonuses. Yeah. Raise our people. Yeah. Which is why we're winning also. We have spent $300 million of OpEx into our AI. I'm not even talking about what we spend in terms of M&A. We have a huge cost of onboarding and winning new business because this is a cost. Despite all of that, we're growing 18% plus, a bit more than 18%, which is again 300% more than. Why? We have already automated many things in the last two years since we launched CoreAI, and we're going to continue to do that. What we experience is for $1 of saving, there is in year one basically $0.80 of investment. Yeah. On year one, let's come back to the point I was making before about our clients. We have too many clients at the moment that are not ready to spend this $0.80 on year one, although they can amortize it to get the $1 of savings a year after. That's what we're expecting now to move forward. Interesting. Again, switching gears to the last part of the business, the biggest, in fact, the connected media part, which is going high single digit. Loads of questions that I can ask on that. Oftentimes, we hear from investors that they don't know or they're unsure about the sustainability of the model. Yeah. The fact that you integrate scale and data, for example, what would you say on that? I mean, we have been outperforming for six years the market pretty significantly. I've got news for you, which is in 2026, we're going to still outperform. That's done already. I don't know what I should do or say to convince. By the way, we are winning all the pitches, and we have a retention rate of 100%. I don't know what I should do or say to convince them. Now, if I want to give you a bit of granularity on that, I actually think that the connected media we have created is going to even accelerate it with the rise of AI. I'm going to tell you why. It's a very simple model. It's to say, which is a big difference with WPP. I guess we are aligned now with Omnicom. Omnicom, for a while, I mean, John was saying, we don't need the data. Email, you remember that time. Everyone changes mind one day, which is great. I think both of us are aligned now. By the way, that might be a reason why we're winning, both of us, is that at the end of the day, the future of marketing lies in identity. I'm sorry to be is that. I guess you asked a very good question the other day on the call. It's the only way for our client to be successful is to make sure that they can engage directly with their customer at an individual level. It was true four years ago for a car, and all car manufacturers realize that, that it was a big money item. It is true also for pet care today. If you have a cat, you want to be able to talk to the cat owner, not only to make sure that you know what kind of product, but also to make sure that you can go to the vet. This is how it works. It's true if you eat chocolate. I know everyone has a mouse, but now you need to go directly to a customer. Why do you need that? This comes back to your point, which is you need identity to basically do three things. First, to be able to recognize not only your customer, but your prospect. For many of our clients, they don't have direct access to prospects, so they need to start building their audience on those two. The second thing is to make sure, to come back to your point, that you can connect the full media ecosystem. The reason why there is still growth for the year to come is that we have actually increased considerably our addressable markets. Yesterday, we were media buyer. Today, we help our client connect their full media ecosystem from publishers to CRM to commerce to, again, influencer, which again increased our base of what we can do for our client and increased what we do for our client. This is why we are growing so fast. Where AI makes a big difference is that as AI is only data connecting to data, three years ago, I was not able to connect 5 million influencers into a single platform and truly understand what are the followers of those 5 million influencers that could be useful for my pet food company, okay? Now I can. I can do it at an individual level. To come back to your question, and I'll finish with my last point, what I guess the market is not understanding is the reason why we are still growing at that pace and we will continue to grow is that every day with connected media, we are increasing our addressable markets. Every day, we are making it more efficient thanks to AI. You know, it's very interesting. The last point that you do with identity is measurement. I think the thing that we are not doing a very good job at at the moment where we really need to progress is that the $12 billion of investment we have made are actually over-leveraged now thanks to how we can connect it thanks to AI. This allows us to empower our clients with what they really need to be AI ready. More importantly, to think in terms of business outcome. You will have a lot of tech companies that will come and explain to you that they have the best product. They can't talk to their client about the end benefit, which is, what is this going to bring to my business in terms of saving and in terms of growth? I'm sorry I'm talking around about that. I can go on and on on this question. This is, at the end of the day, ultimately what clients are expecting from what we bring. By connecting the entire media ecosystem, we're able to deliver those business outcomes. Sorry, it was a long answer. No, sure. Super interesting. I want to stick to that topic for a minute because in June, there were some announcements just before the Cannes Lions from Meta that they would launch during 2026 some tools that would allow advertisers to create their own ads, even buy their own ads, and do the things that people perceive to be done by agencies. I know that Meta clarified later on that this was more aimed at SMEs than large corporations. What would be your take on that? Do you see that as a threat to your business? I'm going to make a short answer and a long answer, okay? The short answer is Meta is a bit less than 5% of our client spend. Thinking that our clients, not talking about SMO and SMEs, thinking that our client can go end-to-end with one platform makes no sense, okay? I think there is something that is bigger beyond that. Maybe I'll spend a minute on that. Maybe it's interesting for the audience. If it's not, you said like this, and I would stop immediately. How can I say that? I'm being careful with the words now because I like to be provocative. I'm an ad guy. Sometimes my provocation doesn't help my stock price. I'm being careful. I think we should never forget that at the core of who we are, we are a service business. We are today a service business that is by far the most advanced when it comes to AI. We are still a service business. The reason why we've been successful in the last eight years is because we always said to our clients, don't choose us if you're looking for a media agency or a creative partner. Go where you get a better price going somewhere else. By the way, they will kill the margin for you. Just do it. If you're looking for true transformation, we are the right partner for you. This is the reason why we've been winning is because we've been telling our clients, the world is changing. If you want a transformation partner, we're here for you. If you want a media agency, go to competition, okay? This has been increasing through time. Why? Because clients want more transformation, and AI is helping. This is why we're winning. Where we feel very confident, and I would say incredibly excited, actually, is that the world is getting more complex every day. The number of LLM, agentic network, systems, the palette of possibilities you have to be AI empowered as a client is just massive. Our role here, which I think is a fantastic white space, is to be the one that truly connects all of this great technology. We will never compete with Meta or Google or Nootropic on spending hundreds of billions to get the best product. We're going to let them race. We are going to make sure that whatever they create, thanks to our tech infrastructure that we can create for our client, thanks to our data and our identity, thanks to agentic AI that we can put on the top of that, and by the way, thanks to our people and expertise, we can actually connect our data, connect our technology, connect their agents to deliver on business outcome. The reason why we're winning today is this. It is because we are the only one who's having the tech experts, the data, the expert in media and technology, and the AI through CoreAI to bring everything together. To come back to your question, this was the long answer. Meta is a very important partner for us. Everything they are creating at the moment in terms of AI is very useful for us. It is only a tiny bit of the ecosystem our client needs to build. By the way, and I'll finish with this, people are starting to get fired because they come to their board or their COMEX with a single interpretation of what AI can do in silo. That's over. Everyone understands now that it's about connecting end-to-end all of your AI strategy in order to truly deliver business outcome and not just the nice things that you can show on the webcast. This is exactly what we do. We connect all of that. OK. That was long, I know. Yeah. No, as always, I think very, very interesting and detailed. Sometimes the perception that investors have about agencies is that they're chiefly, I would say, cost-plus businesses. Maybe one other way to split your revenue would be to look at what is actually driven by business outcomes, as you said. Yeah. What is still driven by, I would say, the traditional legacy part of agencies? Is that a split that you ever look at? There are two questions in your question. First of all, we have a revenue mix that is very different from competition. Again, we have spent $12 billion in data and technology on one side. We created the power of one to make sure that they are truly implicated. Today, 60% of our revenue is in connected media growing at a single digit. 15% is in Sapiens, not growing as it should, but coming soon. Only 25% is in creative, with again 8%. If you look at Sapiens, it's 100% AI, 100%. Media is 80% AI already. Creative is 30%, which is again the production. We still have work to do there. To come back to your question, when it comes to Publicis, when you look at our revenue mix, the legacy business, I mean, if I want to be negative, I will tell you, yes, we still have 18% of legacy business. That's it. That's our reality. Again, this revenue mix, we are the only one who have it. Now, there was another question in your question, which is remuneration. Today, we have two legs. We are paid for our people, and we are paid for our technology. Technology is taking over people. We don't want this to be too much the case. The reason why we are profitable, again, is because we are a service business. People, clients value what we bring. People understand that it's not only about giving you a product and do whatever you want with it. It's about telling to the client, for this, you should take this product. For this, you should take this product. For this, you should take our product. We need to put that all together in a safe place, infused by data, delivered by great people, focused on business outcome. We don't want to lose it. By the way, this is far from disappearing. Yeah, yeah. If you ask me, I've got a big question about the SaaS business. Yeah. Which is, is SaaS going to be disrupted by AI so much that it will disappear? Maybe. At least I will be retired for a long time. All right. Maybe another theme in the industry, which is resurfacing, is M&A, and large-scale M&A, obviously. Yeah, I've seen that. Obviously, we've seen the deal between IPG and Omnicom. There have been some press articles recently about dentsu putting their international business for sale, and some speculation about what may happen at WPP or Esport Capital. Yeah. What sort of role do you want to play in there? Do you want to participate in that sort of large-scale consolidation between agency holders, or? First of all, the Omnicom IPG merger, acquisition, whatever you want to call it, is going to happen. I know you're seeing Philippe tonight. This is one of the best things that could happen to our market. The problem we're having at the moment is a perception from my investors that we are in a declining market, which, by the way, is not the case. It's just that there is, first of all, way more players than those four old co. They are, most of them, growing. It's true that at the moment, we have two wooden players. Yeah. We have WPP on one side that I'm confident that now Cindy is going to come with a plan and make something good. You have IPG that is moving to Omnicom, and we know that John is doing a great job. We're going to find ourselves in a position where hopefully, touching wood, we're going to have three strong players taking basically 100% of the big pitches. We're going to reduce drastically the competitive landscape by 25% with players that know that we are here for the long run, and we should not do anything stupid. I feel very good about that. Coming back to your question, I think I've been very clear so far. If I need to be, again, I will. By the way, look, reading at the price this morning, maybe I should. We are not interested in consolidating more of the same. If it was true a year ago, it is even truer today. I mean, it's actually not even truer. It's the gap. I saw myself seeing that exactly a year ago for good reason, because we are not interested in more of the same for the sake of efficiencies. One year later, with the speed of what AI is bringing us and the question we should ask ourselves and how we can automate many, many, many things, it's definitely not what we're interested in. We are interested in, and I know you're going to come with your share buyback question soon, we are interested in buying capabilities that will continue to make us outperform the market and deliver what we believe is the best shareholder value of this industry by far. Okay, I'll reserve the share buyback. You can. If you're buying your own capabilities. By the way, we're always raising cash. We don't buy with our shares ever. Yeah. Okay, I think your message on large-scale M&A was. I'm not fully at it. Yeah, yeah. Maybe if we can discuss a bit about the acquisitions that you have completed in the last two years. Yes. You have acquired more data sets and more assets into influencer marketing. You touched on the growth that these new, I would say, units deliver. What has been the rationale for spending, I would say, that much money on those deals? From the outside, it seems like the initial multiples seem a bit elevated, maybe. I do think so. They seem so, yeah. No, you know. Maybe for the first. It's funny you said that because 2023 was a year where multiples were very high. We didn't buy anything. We buy like $200 million. I believe that every acquisition we have made in 2024 and in 2025, multiples were pretty reasonable. Apart from the multiple, our strategy here is very simple. We have made a very painful transformation. You have seen it on the other side. Sorry to say it, but we spent $10 billion in data and technology when you guys were putting me a gun in the head and saying you need to do share buyback. My competitors did that. Look where they are, okay? It has been extremely painful. The second thing is we said no silos, no solos, no bozos. We fired all the bozos. We broke down the BML barrier and truly integrated this data and technology into our media and creative, which means that they are AI-proof today. This has been incredibly difficult. Third, and I don't know if you were in Cannes at that time, we said AI will matter. It was in 2017. We started playing with AI almost 10 years ago now. Thanks to that, we have built a unique model. Now our obsession is to make sure that we keep growing and keep innovating to keep leading, thanks to smaller acquisitions that come and complement what we do. What we are looking for is acquisitions that bring us IPs, technology, people, and that can accelerate our existing business. If they don't fit, and by the way, the right price, if they don't fit with that, we are not interested. If you take the last big thing we did, which is all-around influencer, I don't even know where to start there. The excitement and the interest we got from our clients when we told them, if I want to cut a long story short, thanks to the AI platform we have bought and the 15 million influencers we've got that we can connect to our identity, we are able basically to get to the reach of the Super Bowl for a tenth of the price. By the way, know exactly to who you talk and can measure that. This is invaluable. By the way, I'm not even that interested about the fact that we are going to double our acquisition in terms of revenue this year. The impact it's having on the rest of the business is way more important. We would not win all the picks we're winning if we didn't have that. Yeah. I'm obsessed by that. To come back to your question, if it's to add more people that are doing media, more people that are doing creative at a time where, by the way, it's going to be disruptive, I'm not interested. I'm interested in buying those capabilities in identity resolution, in intelligent content, in this new kind of media that, by the way, can include very traditional things. We have made some acquisitions in sports where we're able to take those sports events and link it to influencer and then to commerce. This is amazing. That's the only thing we're interested in because, again, it delivers the growth. Yes, we have been spending money for sure, roughly $2 billion in two and a half years, I would say. It's growing 25%, and it makes the rest grow 5%. You're going to ask me, are we going to do 4.8% or 5.2%? At the end of the day, it's the same number. It's 500 basis points more than competition, which is a demonstration that thanks to AI, we actually extract ourselves from the pack. Understood. I don't want to end the conversation on share buyback. Let me ask it. You sound incredibly confident. You've shown the numbers. No, I'm not. I'm sorry, but I'm cutting you. The good news is we have a track record. Yeah. I was confident seven years ago, and I didn't have a track record. I'm very serious. People are moving away from our results. Now that you have a view of the industry, and I'm not only talking about those guys, I'm talking about the tech industry overall. I'm talking about AI winner and AI loser. If you take 30 minutes to look at the results in the different industries that broadly do marketing services, you will see that we are winning thanks to AI in an incredible way. It's not that I'm confident. We have a track record now. Yes, I'll say both. You sound incredibly confident about the future. Yes. You have a strong track record. However, the share price performance this year has been a bit mixed, let's say. The multiple of Publicis has de-rated. You generate a lot of cash. You have a very, very strong balance sheet. Would you be open to the idea of actually buying into your own capabilities by doing a share repurchase program? You know, I think that the reason why the stock went down is because we did a poor job in communicating what we do. I really do. I'm going to spend more time on that now in the coming weeks because I think there are three reasons. Reason number one, we wanted to demonstrate to the market that our guidance was bulletproof despite all the difficulties of this world, and by the way, the difficulties of our peers. I think we spent too much time showing all the risk that was baked into what can happen. I think I would have said H1 is going to be like H2 despite the comparable. This is something that we need to reassure on, and the summer helped a lot because again, we feel more confident. I'm coming to your question. The second thing is this market thing. We came just after the WPP warning. By the way, as you remember, Omnicom in Q1 also did the same. I think John and I and Cindy tomorrow, but let's give her a bit of time, need to do a better job to explain that we are in an industry that is growing and creating value. I think as an industry, we have a better job to do, okay? Even though I think we're extracting ourselves a bit. I could put also the CEO of Accenture or the CEO of, I think we have a better job to do. The last thing is we definitely did a poor job explaining how much we were winning thanks to AI because we didn't talk about it basically. For me, it was assumed we had people, and so I think people put us in the wrong bucket. Yeah. To come back to your question, the reason why we are winning thanks to AI and delivering what is, again, the best TSR of the industry for the last six years is because we invest. We invest in capabilities. When we have the opportunity to invest, we do at a reasonable amount of time because we have a cash allocation strategy that I think is very shareholder friendly. 50% of our dividend goes in cash. Our profit goes in cash and dividends, so that's not a small thing. We buy back share to make sure that the share count doesn't move. As you've been following us for a while, these are two massive progress. I'm not saying that if we don't find what we're looking for and what stays forever, we won't do some share buyback. I don't want to tell you that this is our priority. It is not. Our priority is to continue to maximize shareholder value. We have made a clear demonstration in the last year that we have the right strategy to do that. Believe me, we're going to continue to do so. When I look at the perspective of this year and next year, at least for the next 12 months, let's say, I feel very confident that the kind of bumps we're having at the moment because of this uncertainty on the macro, because of our industry, and because of this AI winner-loser thing will actually go away and make us come back where we deserve to be. Since we have about 90 seconds left, I think I want to finish off on talking about maybe 2026. You've already won a lot of business. Did you mention the biggest one, Mars, isn't impacting 2025, so it will impact 2026. How much visibility do you have now on 2026? Visibility is increasing, which is good news. It's increasing for two reasons. First, again, because new business starts to materialize. We see how this is going to trend. Second, this works for the first H1, we don't have anything very dangerous coming, at least today, maybe tomorrow, but give us a bit of time. Third, we are seeing more AI projects and AI opportunities coming every day. For all of these reasons, we are very confident. We know that for the seven consecutive years, we're going to outperform GDP and our peers. For the rest, we'll have to wait a bit, I guess. Maybe I'll close with that. It has been extremely painful to be where we are. I think we have been working very hard. We deserve a bit of it. Also, we've been very lucky because we made some investments, particularly in data, which again, you don't have AI if you don't have the data, and in technology, that put us today in a position where we can feel very confident about what is coming in terms of technology transformation, raise of AI, and these needs for every of our clients to be AI empowered. We are a solution that can connect all the solutions around to truly deliver business outcome. I think that's a great way to conclude. Merci beaucoup. Thank you. Merci. Merci beaucoup. See you again. See.