LiveRamp Holdings, Inc. ($PUB)

Good morning. This is the conference operator. Welcome, and thank you for joining the Publicis Group's acquisition of LiveRamp Presentation. [Operator Instructions] At this time, I would like to turn the conference over to Mr. Arthur Sadoun, Chairman and CEO of Publicis Group. Please go ahead, sir.

Thank you, Sherry. Hello, everyone. I am Arthur Sadoun, and I'm here in New York in the middle of the night with Carla Serrano, Chief Strategy Officer of Publicis Group; and Scott Howe, CEO of LiveRamp. Loris Nold, our CFO, is also on the line, but from Paris. Thank you for joining us at such a short notice. We wanted to make sure we give you more color following yesterday announcement that we have entered into an agreement to acquire LiveRamp to become a leader in data co-creation and help clients build more intelligent agents. With these strategic investments, we are demonstrating once again our commitment to accompanying our clients in their transformation and continuing to power our growth and financial KPIs by leading the industry into the AI era. But before we get into the presentation, please read the disclaimer, which is an important legal matter. Okay. Let's start with a quick overview. LiveRamp is a global data collaboration platform and a strategic AI enabler. The transaction represents an all-cash deal for an enterprise value of $2.2 billion. Among its many benefits, it will allow Publicis to become a leader in a new high-growth segment, data co-creation to fuel more intelligent agents for our clients. LiveRamp will also strengthen our ability to deliver Agentic business transformation. We expect this acquisition to be accretive to our headline EPS from the first year. It will also allow us to raise our '27 and '28 constant currency growth objectives to plus 7% to 8% for net revenue and plus 8% to plus 10% for headline EPS versus previous objectives of 6% to 7% and 7% to 9%, respectively. In terms of next steps, the acquisition is subject to regulatory approval and approval from LiveRamp shareholders. It is expected to close by the end of the year. Now let's get into the detail of this announcement, including what LiveRamp is, the financial details of the transaction, the rationale for this strategic investment and how it fits into Publicis model. We will then answer what we believe are the 3 key questions you could have before opening up to Q&A. For those who are not familiar with LiveRamp, we thought Scott should start by giving you a quick snapshot of the business. But before, first, Scott, as we say in French, bienvenue to Publicis. We are very happy that you are here with your team and that you will be soon joining Publicis family.

I'm thrilled, thrilled to be here with you and have even moved on the French time. LiveRamp. Well, it's a data collaboration platform that allows companies to connect, unify and activate data throughout the digital ecosystem. Its interoperable technology connects data across all major cloud environments with robust governance tools and a commitment to shared standards that help customers collaborate with trust and transparency at scale. It enables data activation at scale through plug-and-play connections across our extensive collaboration network made up of 25,000 publisher domains and more than 500 data and technology partners in 14 markets. We have over 800 clients, including more than 25% of the Fortune 500 and cover nearly all segments and verticals, including advertisers, retailers, publishers, platforms and holding companies. Customers. Customers are our North Star. So what I'm most proud of over this period is our ability to attract world-class clients and profitably grow their success over time. Our revenue has grown by a 13% compound annual growth rate over the trailing 5 years. Recurring SaaS subscriptions represent 76% of our business and our customer revenue retention has averaged 107%. In fiscal 2026, LiveRamp's revenue reached $813 million and our non-GAAP EBIT margin of 22% improved significantly and regularly compared to the 4% margin we delivered in 2021.

Scott, I'm sorry for the 2:00 a.m. meeting. Yes, you're right. We're on French time. I'm now going to hand over to Loris, who will tell you more about the financial details of this transaction before we come back with Carla on the strategic rationale. Loris, over to you in Paris.

Thank you, Arthur, and good morning, everyone. Let me walk you through the main financial aspects of the transaction. The total enterprise value amounts to USD 2.167 billion. This corresponds to an equity value of $2.546 billion based on a share price of $38.50 and fully diluted shares outstanding of circa 66.1 million and an acquired net cash of $379 million. This transaction implies a forward adjusted EBITDA multiple of 12.3x. This multiple is based on a calendarized 2026 non-GAAP EBITDA of circa $126 million for LiveRamp based on consensus that includes a cost of $80 million of share-based compensation to align with Publicis Group's accounting policies and an incremental $50 million of savings on a run rate basis, which I will detail later in the presentation. The all-cash transaction is accretive to headline earnings per share on a fully diluted basis from the first year of consolidation. Closing is expected by year-end 2026, subject to customary approvals. Moving to the next slide and how we are planning to fund this transaction. As said, this is a 100% cash transaction fully financed through cash on hand and financial debt. We anticipate issuing new bonds in the second half of 2026. Assuming closing by the end of 2026, this transaction would result in a maximum net financial leverage of around 1.2x in 2027. We expect to confirm our current BBB+ and Baa1 credit ratings post financing of this transaction. Moving to the next slide and the illustration of the impact of the transaction on our 2026 fully diluted headline EPS. We expect a 2.9% positive impact on headline EPS, assuming calendarized 2026 figures based on consensus for LiveRamp and a post-tax run rate cost improvement of EUR 0.13 per share based on an assumed pretax $50 million savings. The incremental cost of financial debt is expected to be circa EUR 70 million on a full year basis or EUR 0.21 per share after tax. Moving on to the next slide. We have a strong track record when it comes to accelerating growth of our acquisition post closing. Since 2019, we have closed more than 30 acquisitions and with an accretive impact on our top line. To give you just a couple of illustrations. First, Epsilon delivered double-digit growth in 2021 through 2024, well ahead of our initial expectations. Second, our bolt-on acquisitions completed over the last 2 years have delivered circa 20% organic growth per year, again, above their stand-alone business plan. We are expecting to see an accretive impact of LiveRamp, driven by its strong growth on a stand-alone basis and as it starts benefiting from the power of One model. Second, we are absolutely committed to fully preserving the neutrality and inter-operability of LiveRamp's operating ecosystem. In doing so, we intend to secure revenue with all partners and clients, including other holdcos, which represents circa 5% of LiveRamp revenue as we did with Influential and Captiv8 since we acquired them in 2024. In fact, we proactively wrote to all the other holdcos to make this commitment clear to them. Moving now to cost improvement opportunities. At this stage, we are targeting a minimum of $50 million of savings on a run rate basis. These savings will come from 4 main sources. First, the continuation of the stand-alone margin improvement plan initiated by LiveRamp with its Rule of 40 objective that aims for a non-GAAP operating margin of 25% to 30% by 2028. Second, the post-transaction elimination of all public company costs at LiveRamp. Third, the integration of LiveRamp's back office into Publicis Group shared service centers; and last, some procurement synergies, including IT, hosting and real estate. With this, we anticipate LiveRamp to deliver an operating margin in line with Publicis Group's margin as of year 1. Moving to the next slide on outlook and capital allocation post acquisition. First, we are confirming our 2026 guidance on all KPIs. This excludes one-off transaction-related costs. Second, with this acquisition, we are raising our 2027-2028 objectives at constant currency. We now expect net revenue yearly growth of 7% to 8% versus 6% to 7% previously and headline EPS growth is expected at 8% to 10% per year versus 7% to 9% previously. In terms of capital allocation, we confirm our dividend payout of 45% to 50% of free cash flow with a floor of EUR 3.75 per share. We will continue our share buyback policy to offset any potential dilution effect, and our focus will be on balance sheet deleveraging until early 2028. Moving to my last slide on the next steps. The transaction has been signed and will now proceed through customary regulatory approvals and LiveRamp shareholder approval. Scott Howe will remain CEO of LiveRamp and continue to oversee all of its operations, reporting directly to Publicis Group Chairman and CEO, Arthur Sadoun. LiveRamp will continue to operate as an independent business and for external reporting purposes, its numbers will be reported within the group technology pillar. We expect closing to take place by the end of 2026. This concludes my financial presentation, and I now give the floor back to you, Arthur.

Thank you, Loris. Back to New York. Now let me hand over to Carla, who will explain why LiveRamp with Publicis will position the group as a leader in data co-creation to fuel more intelligent agents for our clients.

Hi, everyone. First, let me take a beat to explain what data co-creation is. Data co-creation is the process by which companies connect multiple high-value data sources across partners in a secure environment. This generates new data assets that companies could not build alone. This sets the foundation for building more intelligent agents for clients. It's a valuable capability for clients in today's world. As AI adoption accelerates, the AI paradox is becoming impossible to ignore. Companies are investing millions and only getting thousands in return. When you roll those investments up across industries, that disconnect becomes even more staggering. $1 trillion in expected AI investment with only 5% delivering meaningful value. That's $1 trillion of spend without material returns. But let's be clear, the problem with Agentic development isn't the AI itself. It's the data foundation. The fundamental issue is that the agents companies are building today do not have the data required to compete and grow their business. In fact, 93% of companies don't have the right data to support effective Agentic builds. This is because of 3 things. First, most companies are running agents on legacy enterprise data built to report on the past, not make decisions for the future. Second, everyone has access to the same data for the same agents, killing their competitive advantage. Third, all of that incomplete and disconnected data increases AI hallucinations, moves agents further away from their set objectives and means companies can't identify where an agent went wrong or how to fix it. That is why we have decided to invest in LiveRamp. Thanks to their data connectivity, marketplace, collaborative clean rooms and partner and agent network, combined with Epsilon's identity, we are going to accelerate on data co-creation to help clients build more intelligent agents for real business outcomes. In doing so, we can deliver 3 important advantages that will help clients close that gap between AI investment and return. First, greater speed, security and scale. They can now unify fragmented internal and partner data to enable secure collaboration across organizations without exposing sensitive underlying data. For example, a bank could build a powerful wealth management life cycle agent. The agent could use unified customer data from its retail banking, credit card and wealth management and securely connect it with partner data from merchants, payment networks and travel providers without exposing sensitive customer records. This agent can now cross-sell faster, coordinate efforts across multiple lines of business and more accurately detect fraud. The business impact of the agent is transformed from narrow task completion into a tangible competitive advantage in customer lifetime value, customer experience and retention and risk mitigation. Second, generate proprietary intelligence. By creating proprietary data assets from new combinations of signals and data sets, they can unlock hidden insights that drive smarter strategies and sustainable competitive advantage. For example, a retailer could build a comprehensive retail journey agent. The agent could connect data from CRM, loyalty to in-store to retail media network inventory to partners in order to measure the incrementality of each touch point and to build new proprietary journeys for shoppers. The business result of this agent now becomes faster, more efficient shopper conversion and more value for retail media partners. Third, continuously train and fuel enterprise grade A agents with co-created data to accelerate responsiveness and decision-making. For example, a global pharmaceutical company can build a therapeutic area optimization agent. This agent can compliantly use clinical, commercial and operational signals with patient, prescriber, payer and supply chain data across their brands and at a therapeutic area level. This agent can now use new dynamic signals to balance distribution by brand, optimize field force deployment in the context of marketing and uncover and navigate any barriers in the payer system. The business impact is incremental growth for each brand, more efficient and higher ROI field force activities, therapeutic area product lifestyle management and total enterprise growth. These are just a few select examples of how, as Scott would say, together, we can democratize innovation and data for the entire ecosystem.

Scott highlight the value of democratized innovation and data for the entire ecosystem. It's very inspiring. Look, put simply, agents build on co-created data, learn and improve with every signal, separating them from competitors that train their agent on stagnant generic data. Building smarter agents by leading in data co-creation opens up a new addressable market that allow us to raise our '27 and '28 financial objectives. It is important to note that it also complements our proven growth model and boost our ability to accelerate client Agentic business transformation. Thanks to Publicis Sapient, we can build and modernize technology and system foundations to make our client infrastructure AI rich. Epsilon market-leading identity connects clients and their agents to real people, behavior and deterministic transactions as a fundamental source of truth and growth potential. With the addition of LiveRamp, we will enable clients to collaborate safely and securely across partners and platform to co-create new data that fuels smarter agents. Last but not least, with Marcel, our Agent platform, we can activate this co-created data across all of our clients' enterprise function. Just to wrap up, as you can see on this chart, LiveRamp perfectly fits into the architecture of our entire model, delivering more intelligent agents to accelerate on our clients' agentic business transformation. Now before we open the floor to the Q&A, let me first address 3 key questions that I think will be top of mind for many of you. First, for those that are not that familiar with our industry, let me explain why LiveRamp is so different from Epsilon. Actually, Epsilon and LiveRamp serve 2 fundamentally different purposes. Epsilon is a marketing activation engine, focused on using identity and data to drive direct consumer engagement and business outcomes. On the other hand, LiveRamp is a B2B data collaboration platform that connects partners to enable multiparty data collaboration. When you look at their capabilities in detail, the differences are even clearer. On data, Epsilon is focused on organizing and unifying deterministic transactional, behavioral and proprietary data to power personalized marketing and media activation. Meanwhile, LiveRamp specializes in connecting and unifying fragmented enterprise and partner data across the entire ecosystem. On technology, Epsilon is focused on identity, building a unified identity graph to support audiences creation, media planning activation and measurement across paid and all channels. LiveRamp expertise is in inter-operability and collaboration, enabling multiparty data collaboration through clean rooms, data onboarding and ecosystem connectivity. On client access, Epsilon is delivered as a managed service designed to help clients reach audiences by activating data through deep insights and by providing performance measurements. In contrast, LiveRamp operates as a SaaS platform used by publishers, retailers, platforms, brands and partners. To cut a long story short, they serve different purposes, have different data approaches, complementary tech stack and distinct go-to-market. LiveRamp is not duplicative. It will be an additional building block of our growth model. Second question you might have is how LiveRamp neutrality and interoperability will be preserved. Let me be very clear on this one. As with all Publicis operations that work directly with partners and competitors, LiveRamp will maintain total neutrality. It will continue to operate as an independent business and ensure open access. It will not prohibit or restrict access to its service for any current or potential customer and will remain fully interoperable. When it comes to privacy and control, LiveRamp will not use or share client publisher or partner data in any way that is not explicitly identified in agreement with them. Last but not least, LiveRamp will not engage in pricing changes beyond standard business practices. As Loris this told you, we have sent a letter to the holding companies that represent roughly 5% of LiveRamp revenue to make those commitments very clear. It's very important to note that independence has been a key growth driver for many of Publicis acquisition, including Influential, Captiv8, Lotame and several sports acquisitions. To give you one concrete example, when it comes to our connected influencer platform, 51% of its growth come from non-publicis clients and competitors who have chosen to use it. Last important question, why acquiring LiveRamp when we are already partnering with them? The #1 reason for this acquisition is that the addition of LiveRamp will allow us to expand into a new addressable market. In 2019, we acquired Epsilon in the name of leading personalization at scale to enable our clients to take back control of their data from the walled garden by shifting from cookies to identity. Since then, we have been outperforming the industry. Now with LiveRamp, we are looking ahead to what's next by building the future of data co-creation. It is how we will enable our clients to generate new exclusive and proprietary data to build the smartest and most differentiated AI agent on the top of the leading LLMs. Second, we are confident that the power of One means that we will be able to quickly unite and deploy LiveRamp capabilities for all of our clients globally. With LiveRamp added to our ecosystem of Publicis Sapient, Epsilon and Marcel, as you have seen, we will go even further and faster in delivering agentic transformation for our clients safely and transparently and even more importantly, in their own environment. Last but not least, we have talked a lot about data and technology today. But at Publicis, we continue to believe that people are a key differentiator. We have absolutely no doubt that LiveRamp highly talented teams will have a great impact on our organization. In fact, we worked closely together over the past 6 months through our commercial partnership, which has allowed us to test the cultural fit, and I can tell you something it has been excellent. Of course, the deal still has to go through the full regulatory process, but we are really looking forward to welcoming Scott and his team to Publicis. Well, thank you for listening. Thank you for joining. And now we are ready to take all of your questions.

[Operator Instructions] The first question comes from Nicolas Langlet of BNP Paribas Exane...

Congratulations on the announcement. I've got 3 questions, please. First of all, on the network neutrality. So do you plan to implement any new governance structure to ensure that all the competing data providers continue to view LiveRamp as a neutral platform. So is there any change on that front? Secondly, on the synergies, so you have mentioned the $50 million cost synergies, but do you expect any revenue synergies over the midterm? And when do you think they might start materializing and what magnitude we could expect? And finally, on the ID system you will get. So post the acquisition, Publicis will operate 3 distinct identification systems with Epsilon, and LiveRamp -- is there a plan to consolidate those data systems at some point? Or you think they all have their own specificities and they can remain like that?

Thank you, Nicolas. I'm going to take 1 and 3, and I'll pass on to you, Loris, for 2. On the identity, no, we have no plan to consolidate. As you have seen in the presentation, each of our operation actually fit a different purpose, but also can be connected to really bring agent transformation to our clients. And I think this is the power of what we are doing. And by the way, the power of One, which is to bring very different expertise together, but each of them with their culture, with their way to go to market, okay? This leads me to the neutrality. So I want to be very clear on that. There will be absolutely no change in how today and tomorrow LiveRamp will be led. I mean I've got Scott next to me here that will be able to tell you a couple of words about that. But he will remain, of course, the CEO. The team will still be in place. Again, we have this great partnership for 10 years. We have accelerated for the last 6 months. We have seen how well we work together. And again, one of the big reasons why we are doing this deal is a cultural and people fit. So to come back to your point, absolutely no change. Now before I pass on to Scott, let me take a moment again to come back to this question because we have seen since yesterday that was a big topic for you guys on the analyst side. So first, we absolutely want to preserve the inter-operability of LiveRamp because it is part of the business. And that's how they grow and how they will grow in the future. And that's why we are very confident that this is going to only increase the Publicis performance. Second, we are making very strong statement about that here on the call. But also, by the way, with our competitors that represent today 5% of the revenue. I have addressed a letter to all of them, and we feel confident that we can still work in good condition. And this is only work. The things that matter are the facts. And the reason why we are so confident in our ability to keep this interoperability is actually that the neutrality that you see that you will see with Live is the one we are having today with other of ours. And we mentioned a lot of them, but I think the most interesting one is actually Influential and Captiv8. You would remember at the time we did those acquisitions, you could have hear the market, yes, but the competitors are going, blah, blah, blah. The result is half of our growth for those platforms come from client, non-publicis clients and competitors. So we know how to do that. But maybe, Scott, do you want to say a word on those 2 points, way, you're more than I am. So feel free to talk about that and definitely on the independents.

Sure. I think you covered it well. And Arthur, you know that this is a topic of conversation that started a long time ago between us. I mean this was very important. We have a 10-year track record of being neutral in the industry, and that's helped fuel our growth. And the commitment that you made already in terms of being committed to remain interoperable and the letters that you sent, I think, went a long way. But what gave our team even more confidence is just seeing how you've managed your businesses over time. And as you've made past acquisitions, you've allowed them to be neutral and interoperable with the ecosystem, and that's accelerated their success. I think the strongest thing I would say, though, is whatever concerns I may have had melted away yesterday. Yesterday, after we announced the deal, we reached out to hundreds of clients and virtually every major publisher partner and we talked about this, and there was 0 concern across all of those conversations that this commitment would be upheld.

Yes. I think this is a critical point. Again, we were not worried at all, and we have good experience, but nothing replaced the client feedback. And on both sides of the equation, we actually sent, of course, a lot of e-mails. I personally sent like 500 e-mails yesterday. And the answer we had from the clients was, of course, very strong, but you touched on the point you made, which is, first, and very importantly, and hopefully, you saw that in the presentation, every client understand that building the right data to build the right agent is mission-critical. We were actually -- I was a bit surprised by how much we were clear on that and that we were really at the right place. But the point that Scott made about neutrality is very important. Honestly, it's easier on our side because, of course, we are bringing a new service, but we have a lot of Publicis clients that are LiveRamp clients. And of course, they trust us to keep this neutrality because they have seen it, of course, with others. And again, coming back on the fact that we are stronger together, the notion between data collaboration and data co-creation is very important. What they can do today is definitely starting to collaborate. But the ability that we're going to have together to create new sets of data, sets of data that will make our client agent kind of super competitive, unique, proprietary is something that they found very interesting and, of course, very appealing. And I would say, last but not least, and I know it's going to be a topic maybe when we talk about our cash allocation they so much feel the need for us to invest in new talent and new capabilities to make sure that we are still relevant in helping them in this AI world. This is a point that comes from everyone. It is a challenging macroeconomic context at the moment, but they have never ever needed us to invest more in order to make sure that we can continue to be this most valuable player for them. Loris, hopefully, we gave you enough time to prepare the question on the synergy. it's a bit difficult because we are in New York and in Paris. So over to you, Loris.

Thank you, Arthur. Nicolas. So just a couple of points on the top line. The first one is very important is on a stand-alone basis, LiveRamp is an asset that is performing really well. I mean it delivered 13% on a 5-year CAGR. If you translate its Rule of 40 objective into the next few years, you can assume that it will sustain double-digit growth. Now when it comes to revenue synergy on top of it, it's a bit too early to say. But what I can tell you is that we are expecting the integration of LiveRamp into our Power One model to generate some significant opportunities for LiveRamp clients, for our clients, for new clients. And obviously, this would have an impact. And if you're looking for, obviously, evidence, look at the acquisition that we have closed in the last 5 years. As I said earlier in my presentation, they have surpassed our initial expectation, both at Epsilon and all the bolt-on acquisition we've closed. So we feel pretty confident around unlocking growth on top of the strong performance that LiveRamp has already been delivering and will continue to deliver.

The next question is from Tim Nollen of SSR.

Congratulations to all parties. And also, thanks for giving us a lot of preemptary answers to the questions that we've got. I wonder if you could expand a little bit maybe, Arthur, on what the addressable markets are that you referred to. There's a lot of AI discussion here. I think you're talking about Agentic AI opportunities. But really, what are these addressable markets that you referred to? And Scott, if you wouldn't mind just explaining why have you decided to sell LiveRamp now and why to Publicis?

Thank you, Tim. Look, I'm not going to come back on all the story about why data co-creation is so important to build the right agent. But happy to take that offline because this is absolutely critical because as I said, and you have seen that in Carla's presentation, we are talking about roughly $1 trillion by 2029 that is going to be spent in agent transformation by our clients, $1 trillion. It's going to be more than advertising. And as you can imagine, one will grow faster than the other. The question there is that there is no way you can capture a part of this investment if client doesn't have the right data. And this is where we see a big opportunity for us is to start capturing this new addressable market. And to be clear, and that's a very important point. We absolutely do not need LiveRamp to win in the marketing space. I think we made a clear demonstration over the last year that we can massively outperform our peers and winning new business with our existing structure. And by the way, with the partnership we're having with LiveRamp. But coming back to this addressable market of more than $1 trillion, our ability to co-create within the same environment with LiveRamp, new products and services that we can sell end-to-end is, we believe, a big for our growth in the future.

And Tim, maybe I can address the why now from the LiveRamp perspective. As you know from covering LiveRamp yourself, you know that clients have always been our North Star. And when Carla walked through the 3 examples around how different sectors are starting to think about AI, she could have just as easily given 30 examples. Literally, there is not a client that we work with that is not struggling with some of these same issues, how to harness the power of AI, how to move quickly. And you know what, it's hard. It is so hard for them to do that. And there's a couple of reasons for it. One is the data that they need isn't necessarily stuff that sits within their own walls. We can solve that problem for them. But in addition, the technology that they need is so disparate and it's often hard to bring together. Well, together, this combination solves that problem for them. And so it came back time and time again as it always has in LiveRamp's history is how do we solve the problems that our clients have and do it at scale and generate better performance for them. And the answer kept coming back to, hey, together, we can do this more effectively than LiveRamp could on its own.

And if I may add on that, I think Scott made the right point, which is it was the right time because it is the right time for our clients. And again, everything we have heard since yesterday is a good example of that. And of course, we tested the water in the past to make sure that this was the case. But it's also, I think, the right time for both Publicis and LiveRamp. I mean this long-term partnership that we have been accelerating since the beginning of the year help us to realize how much we knew each other, we were sharing things and we were ready to go. I mean, again, you know us. We have a very good track record in terms of integration. This integration starts with the people. And the fact that from day 1, we'll be working with people that we are already working with makes a big difference. I think the second thing that is very important is that we are talking since 3 years now about the ones that are winning in this new AI world and the one that are losing. And the truth is when you look at the performance of both companies, Publicis and LiveRamp, today, the number shows and again, today with the partner of LiveRamp that we are winning. And that makes a big difference because we have nothing to fix. We just have to make sure that we grow more together and that, by the way, we bring to clients what they need. And finally, honestly, and that's what gets me very excited is that the complementary products and services we are adding are just a perfect match. And hopefully, you saw that in my chart. They are very different. They are serving very different purpose. They will have their own way to go to market. But when we come to clients that today are looking for end-to-end solution in this new Agentic world, we come with a perfect fit.

Thank you for the explanation and please keep the ramp-up conference going.

The next question is from Ciaran Donnelly of Citi.

A couple of questions from myself. Firstly, revenue. I think looking through the accounts, it's 95% U.S.-based. Can you talk about the effectiveness of the platform outside of the U.S.? And is there any reason for not expanding historically outside of the U.S.? And then two, can you just talk about, I guess, in terms of Sapient, is this going to be a positive tailwind for Sapient? And if so, do you expect this to come through in 2027? And maybe just finally, can you provide a split of revenue for LiveRamp from the other holdcos?

Maybe I take the first here, which is expanding outside the U.S. So you're correct. Our business is very concentrated in the U.S. And that's disappointing in some respects because if you look at our client base, our clients are global. If you look at our publisher partners, the Metas, the Googles, the Disney, the Netflix of the world, they are global companies. And if you look at our technology, it is globalized and can be deployed any place in the world. And so we have clients and partners who have actually been pushing us to expand with them internationally. And we just didn't have the footprint to do that. So I think there's a really nice opportunity here given Public's global footprint and client connections to tap into that over time. And I will tell you, we had a we had an announcement yesterday with our senior leadership, the biggest smiles in the room were the folks that were joining from overseas because they look at this as the opportunity that they've been waiting for, for a decade.

Yes, that's part of our plan, of course, the international expansion. And I think it's interesting to see that both in this case, Sapient and Epsilon, when we acquired them, were roughly 5% international. And you have seen the growth we have been able to deliver on the international side. So this is very promising. If I understood well your first question about the split on holdcos, we don't give those numbers because we don't disclose any client number to be clear. But what I can tell you is that it's roughly 5% for all holdcos together, except for Publicis, of course. Your question with Sapient, allow me to tell you a bit more about how we see things. Again, each of those operations fit a very different purpose, okay? And so you have to think about Sapiens as how we're going to be able to modernize the mainframe of our clients. And that's a very big topic because the other big reason why AI is not working with most of our clients today is because they don't have the modernized framework. And this is where we can do a lot of work, and this is where today Sapiens is really starting to do an inroad. The second thing which comes on that is once you have the mainframe, you need identity. I mean I think that now the market has understood, and I'm talking about the financial market because our clients understood that very early. This is why we have been striving with Epsilon and other group is that identity is the qualifier for AI. If you don't have the identity, you just don't win with AI. Just look at all the platforms, not talking about Publicis, but outside, you don't win if you don't have identity and everyone has understood that. Where LiveRamp adds something great is that data co-creation, meaning collaboration to get new set of data is going to be the multiplier. It's going to be what makes client win. You get qualified with identity, you win by creating new sets of assets and new set of data. And that's why those 3 things with Marcel on the top, make a difference. So to come back on your question, they will stay independent. But yes, they will collaborate. And we are, of course, planning in the future for clients to look for an end-to-end solution. But this is not also what we're going to push for. First of all, because most of our clients already work with different suppliers, and we want to make sure we can adapt -- and second, because we have a huge belief that -- sorry, is a bit technical, but absolutely key is that all of this has to be built into our client environment. Where we're going to make a big difference is that we're not here to sell something that will be apart from their business. We want to make sure that it's core to their business in order for them not only to grow, but to prepare the future. We still have like 10 minutes, and I'm sure there is other questions.

The next question sir is from -- the next question is from Adrien de Saint Hilaire of Bank of America.

I've got a couple of questions, please. Arthur, perhaps can you talk on how much of -- or how many, sorry, of your top 100 accounts currently are using LiveRamp? And maybe a couple of questions for Loris. Why would you raise bonds when you actually have access to EUR 4 billion of gross cash today on your balance sheet? And then I apologize if it's early in the morning, but you've raised your 2028 constant currency growth by about 100 basis points. But I think LiveRamp is going to account for, call it, just about 5% of revenue. And as you said, it's growing like low double digits. So how are we getting to like 100 basis points of like revenue growth accretion?

I won't give you a precise number on how many clients out of the top 100 use LiveRamp, but I can tell you a big part of them. The reason why I won't give you a number is that it can go from a very small service to a very big relationship. So it won't give you a real idea of that. What I can tell you is, first, 100% of our top 100 clients know LiveRamp and have a great image of LiveRamp. This is something that we check. Second, 100% of those top 100 actually answer to our e-mail yesterday and feel very confident either to reinforce the relationship or to know more. And now, of course, there is nothing we can do until the period that is going now is over. And I see my general counsel saying yes, yes, yes, we will. But we feel very, very confident that in a way or another, it will serve most, if not the totality of our clients. Maybe Loris, I give you 2 and 3, I guess.

Yes, sure. Adrien, -- so I'll start with the question on the top line assumption. So as I said earlier, when it comes to net revenue growth at constant currency, we are expecting 7% to 8% versus the 6% to 7% previously. I mean the real assumption here will be on the timing of the acquisition. As I said, we are assuming that it will close at the end of '26, which essentially means that you should see a full impact from the acquisition on reported growth in 2027. In '28, we will include the impact of LiveRamps on organic growth. combined with what will be the normal effect of incremental bolt-on acquisition, which we would resume and that would close in that year. Of course, timing as well as what we should assume for LiveRamp soly growth are 2 important variables. So probably the easiest way for you to look at the objective for '27, '28 is an average range for the period, mindful of the fact that if our assumption is correct on timing for the closing, you're right, we are definitely on a conservative end for '27, given that we have to be very clear on one fact that we are not changing our assumption for Publicis stand-alone. On your first question, which was the financing, I mean we will evaluate in H2 what are the financing requirements and decide how we tap the bond market. I think you have to look at it for us, which is financing is about maturity. And so we're looking at our overall debt structure and also any other needs that we might have, working capital funding and all the other requirements of the business. So this is not only specific to the LiveRamp acquisition.

Thank you. We're going to move fast. We have 10 minutes a bit less. So Sherry, back to you.

The next question is from Conor O'Shea of Kepler Cheuvreux.

A couple of quick questions from my side as well. Just firstly, maybe for Scott, just who would you consider your main direct competitors? And in particular, would you consider InfoSum, which I think one of the other agencies acquired about a year ago as a competitor? Do they do the same things as you do or to what extent? Then second question, I think, Loris, you said you expect LiveRamp's margins to be equivalent to publicis at a group level within the first full year. Can you just indicate what the starting point for LiveRamp's margins are in terms of equivalent accounting policies for operating margins for 2026 or 2027? And then the final question, just in terms of how you plan to integrate LiveRamp. Would you -- are you going to take an approach similar to what you've taken to Sapient in the sense that they are very autonomous within the Publicis Group or more like the Epsilon model where they're closer integrated to Connected Media? What would be the approach there?

Yes, Scott.

Maybe I'll start in terms of competition. I mean, listen, we play in a competitive market. I will tell you that you're direct question was on InfoSum. They're not a company that we run across very often. However, like we do with other clean rooms, we can certainly make them part of our integrations or our networks in the rare instance that a client would want us to do that. I would tell you that I often think about our main competition as our clients just doing it themselves because they're not very sophisticated. And therein is the challenge, but it's also the opportunity for us because in a connected world where clients are going to need ever-increasing amounts of relevant data to power their models, they're going to need to connect. And so if someone doesn't choose to work with us today, maybe it's because they're going to integrate directly with Google or directly with Meta. It's just such a small number of integrations that they can manage that degree of complexity on their own. I think over time, as companies have to be more sophisticated about this, the number of integrations actually increases. And we've certainly seen that in our portfolio where the number of connections, the number of collaborations increases over time. And that just fuels additional revenue growth and makes even more use cases available for our clients.

Thanks, Scott. Loris?

Sure. So on the margin, if you look at the consensus for 2026 and you look at it on a calendarized basis because, as you know, LiveRamp's fiscal year ends in March, and you include the cost of share-based compensation of $80 million for LiveRamp, you get to a margin of roughly 14.2% for 2026. On top of that, when you look at '27, you build the acceleration, the stand-alone acceleration that LiveRamp is forecasting as part of this Rule of 40. And the savings that I mentioned that will be partly realized, actually mostly realized in 2027 and you very quickly get to a margin for LiveRamp, which is at or slightly above Publicis Group's margin in 2027.

Okay.

Thank you, Loris. We're going to -- now I've got to answer your question on integration, which is very important. I mean the truth here, and you can see that in all the acquisitions we have done is that each operation is special. And for each operation, we have an integration that is really tailor-made and definitely in the case of LiveRamp for everything we discussed. So clearly, LiveRamp will remain as a stand-alone business -- it will be led by Scott and his team. And Scott, as you might have seen, will report to me. For reporting purpose reporting purpose, sorry, as we said, I mean, we will map LiveRamp into our technology Pillar simply because when you look at this model and his capabilities, this is where it belongs. It is a tech business. Maybe a last question very quickly, if we can, Sherry.

Yes, sir, the final question is from Julien Roch of Barclays.

So you already had a commercial agreement with LiveRamp and LiveRamp will stay neutral. So what does ownership brings you? What will you be able to do that you could not do before? I know there's a slide in the presentation, you say addressable market, data co-creation and complementary model. But why would you not be able to do that with a commercial agreement? That's my first question. The second one is if LiveRamp stays neutral, why can't other holding company do data co-creation as well? And then the last question is you said Epsilon was 5% international when you bought them. How much international was Sapient in 2015? And how much Sapient today and Epsilon today? So 3 numbers, please.

Okay. I'm going to go fast. I'm going to leave the last question to you. So again, hopefully, we explained pretty clearly why this deal makes a lot of sense in getting together. I won't come back on everything we just said and you wrapped up. But if you want to look it in a very simple way, we want to capture the growth of this huge market that is coming. We want to make sure that we can grow, thanks to that. And second, we want to start building product and services together that no one else could bring, okay? That's basically the 2 reasons why we want to make this deal versus just a partnership. And again, when you look at the potential of the data co-creation market and more importantly, the potential of the agency transformation, you understand why there is growth. And when you look at our product and our model, you understand why it makes sense. Why other can do it? Let me be very clear on that if I haven't before. We are not talking about the marketing space here. Again, as we just said, we did not need LiveRamp to be part of Publicis to win in the marketing space where we make a difference and where we think we're going to bring something very unique. It is in the combination of Sapient plus Epsilon plus LiveRamp plus Marcel to go for this agentic transformation market that is roughly $1 trillion today and that is only going to be growing versus other industries. Loris, maybe?

Yes, sure. Julien. So if you look at Sapient and Epsilon, so I mean, as Arthur said, when we acquired those businesses, they were primarily U.S. domestic base. And if you look at the trajectory, today, Sapient is roughly 40% international. Epsilon is probably closer to 15% to 20%. The difference also is that the starting point on Sapient was slightly higher outside of the U.S. and also the fact that when it comes to Epsilon, they are servicing a number of international clients or the market from the U.S. But take those numbers, 40% for Epsilon and close to 20% -- sorry, 40% for Sapient and close to 20% for Epsilon.

All right. We are almost out of time. We have just a couple of minutes. So if you don't mind, I would like to close with a couple of takeaways. First, as you have seen, and that's a very important point for us, we continue to invest in new talent and innovation. This is what our clients are expecting from us. I guess you hear it from Scott, you heard it from me. The reaction has been overly positive because they need more than ever partner that can help them in this AI journey. Second, hopefully, we made the demonstration that we know how we are going to be able to continue to thrive and to make sure that LiveRamp thrives as a neutral and interoperable platform. We did it for others. We will do it for LiveRamp today. We feel very confident. Third, and that's come back to your last question is that with this acquisition, we are opening up a new addressable market, which is data co-creation. We're going to be able to fuel smarter agent for our clients and win over a part of this $1 trillion market that Carla described. And last but not least, we went fast on the number, but there will be a lot of news coming next. I mean LiveRamp will allow us to accelerate across all of our financial KPIs. It will be accretive to our EPS, and it will drive faster top and bottom line growth. Well, I hope we've been clear in exactly 1 hour. I'm sure there will be a lot of discussion in the coming days. Of course, Jean-Michel and his team are here for you. Thank you so much for joining us so early. Thank you so much for taking the news yesterday on a Sunday. The reason why we did that is that we need to communicate at the moment where both financial markets are closed for LiveRamp and Publicis. So it has to be early in the morning. And this is why, again, we put the press release yesterday. Have a great day and talk soon.

Ladies and gentlemen, thank you for joining. The conference is now over, and you may disconnect your telephones.