Ipsos SA (IPS) Earnings Call Transcript & Summary

[Foreign Language] I am Ben Page, the CEO. Thank you to so many people who have made the journey to our Paris HQ today. It's great to see you all. We will keep proceedings to about 2 hours, we think, which is good in case the room becomes warm. And we have some really good new news for you. So you're about to meet, obviously, me; Dan Levy, the CFO, will take us through some of the latest numbers. Jean-Michel Mabon will explain what we are doing on acquisitions where we have made a lot of progress over the last 18 months. You'll then hear from Eleni Nicholas who would love to be here in person but has been in hospital for 2 weeks of appendicitis. And so as a result, will be joining us on the screen because she can't travel. And then Michel Guidi, who some of you have met before, will take us through what we're doing on our tech and in particular, on Gen AI, where we, again, have made some real progress over the last year that we'll update you on. So in terms of where we are, the world is as complex as ever, if anything, looking at politics in France, even more complex. And in our latest global trends report that we will launch at Cannes Lions next week, if you look at the macro forces affecting the world, whether it's the changes in society; aging populations; the acceleration of technology; the challenges that's reflected in French, but also British, European generally, American, Latin American, Asian politics in terms of inequalities and opportunities; the climate, we've just had the hottest may ever and there is no sign of anything changing there; political tension nearly everywhere; health care systems under pressure; all of those things concern all of us and Ipsos, of course, is built to help us make sense of all of those things, all that complexity. So just a reminder, really, that we are here with one key purpose: to deliver reliable information, to give people a true understanding, not just of people as consumers but also of societies as a whole, of people as voters, citizens, parents, of markets and ultimately, to use that information and help our clients use that information to help the world make better decisions. So we are proudly in 90 countries, 220 cities, 20,000 of us working for over 5,000 clients, over 6 million authenticated and Michel will talk about the need for quality and dealing with fraud when it comes to online interviewing, but 6 million-plus proprietary authenticated panelists. And of course, a unique company. One of our big competitors said a few years ago, Ipsos is a little bit different to everybody else. I think we would say Ipsos is a lot different to many of our global competitors. We remain an independent company led by people who know what they're doing, market researchers. And that's important because it allows us to take a long-term view in terms of the investments that we want to make, the things we want to do. We are not here on a sort of 4- or 5-year time horizon and then to sell the company and pass it on to somebody else. We have specialist expertise in all areas of markets and opinion research. And that means that we're able to deal with some of the most complex challenges our clients face with our combination of broad geographic coverage, real local understanding, but combined with increasing global efficiency through our global services. That combination of things is increasingly distinctive in the market in which we operate. So whether you want to just understand society, in general; whether you want to develop new public policy, if you're a government, if you're a brand and you need to define your market strategy; if you want to create new products and innovate; if you want to make sure your advertising and marketing work effectively; if you want to strengthen your brand's overall health or improve the experience of the people who work for you or improve your customer experience; or make sure that the different channels you're using to go to market are being used effectively; build your corporate reputation, get ready for a very different world or indeed have an impact on the world, you need all the different services that we can provide people not just as consumers, but as it says there on the chart: citizens, customers, health care professionals, employees, we cover everybody. And I think what's interesting is when you look at our client base, our clients -- our major clients use this combination of geographic coverage and multiskilled, multi-service offering. So if you look at the bottom of that chart, the further to the right, you can see the number of markets that these major clients who make up a quarter of our overall revenue. More of them, lots of people in the top left at the top of the chart using over 10 of our different 16 service lines and many of them in more than 40 countries. On average, they're working with us in 29 different countries. And that's an offer that very few of our competitors can match. And also, of course, in a world of fake news, in a world of disinformation, it becomes all the more important and indeed synthetic data, which we can talk about in questions, if you like, in a world like that, becomes all the more important for you to know where the information you're using coming from -- exactly where it comes from. And that is something, for example, that the French election on Sunday showed us 0.3% average error for each party share of the vote. You can only achieve that type of thing if you have end-to-end control of where the data is coming from, understanding the problem you're trying to solve, bringing the data together and analyzing it effectively and then, of course, delivering impactful insights based on that information. So controlling the whole process is for us more and more important. And the majority of the data that we provide to clients is data that we have collected ourselves, that what we have controlled it from beginning to end. We also, of course, are well known, as most research companies are, for political opinion polls, but of course, the surveys are only a part of what we do. 1 in 3 people on this planet are not online still today. We want to be able to retain the ability to reach everybody in all corners of the planet. So we're doing off-line surveys, online surveys, but of course, also passive measurement. You will hear in a little while from one of our experts who specializes in the analysis of social media, where again, we can do more and more, faster, better with Generative AI. We now, of course, can also bring to all of that interview and perceptual data, physical samples, analyzing water samples, saliva, DNA, blood to actually build a full picture of what's happening. The prizes we have won from the World Bank for our work, looking at satellite imagery and drone photography to understand the impact of disasters in places like Libya and Turkey. The huge qualitative arm that we have, Ipsos UU. We believe that we are the largest provider of qualitative research on the planet where we're talking to people as I am talking to you now, one-to-one and face-to-face. Demographic data. If you haven't read my colleague, Darrell Bricker's book, Empty Planet, where he predicted before the United Nations started downgrading their population forecast that actually you would go into reverse in terms of growth of population, as has already happened in China, in South Korea, et cetera; and, of course, our use of client supplied sample, and we will have some other examples of that today. So some of the studies that we do now used up to 200 different pieces of data and having that total picture is hugely important. We have high client retention, most -- nearly all our clients buy from us every single year. And tracking customer satisfaction, we are, after all, a market research company shows that over time at the end of each project, despite rising expectations, despite an ongoing demand for things faster, cheaper, et cetera, we have actually managed gradually to increase the average score that our clients give us. And now the average, as I say to our people is over 9 out of 10. If you get less than 9 out of 10, you are below average in terms of the feedback that you get on your project. So our strategy remains focused in terms of some geographies, some key services. We're obviously, the United States, and I'll talk about that in a second. The major economies in our industry: the United Kingdom, France, Germany, China, India' some key services where we believe that we can build a competitive advantage, public sector and government, health care, data analytics and also digital assets that we want to add to what we're doing. And then, of course, all of that combined with the best people. We want to combine the best technology with the best people. So whether that's Generative AI, our platforms or our analytics, our strategy is built around those 3 strands, and we are making good progress. And in fact, a lot of that is because of the people at Ipsos. We have very high levels of employee engagement well above the average for a professional service company. We have a very high score in terms of recognition of what we do on diversity and inclusion, and that's important in a company that covers 90 countries. We have a whole series of programs; employee resource groups on things like gender balance, on sexuality, ethnicity, neurodiversity, you name it, we want everybody to be able to be their best selves so that we can be as diverse and as representative of the communities and societies that we're trying to measure. We are investing more and more in training and development, whether that's in the Ipsos training center, our knowledge center, where we're trying to make sure that everybody inside Ipsos is able to access all of the internal information that we have; a new global management development and assessment program; and of course, we continue with our Generation Ipsos program, where we're bringing in young talents all over the world and making sure they get some of the best training in the world, and that's hugely important to us. And one of the things that we've just done, we put out a press release last night, was get more of our senior managers to take a bigger personal stake in Ipsos itself. So we've launched a new partnership program this year. We've got around EUR 18 million contributed to that from around 330 plus senior managers. It means that the employees and managers of Ipsos now have an ownership of 12.3%, voting rights of 20.7%, and we intend to do this type of investment for our people every -- to give them that opportunity, every 2 years or so into the future. It's really important because the success of Ipsos is about making sure there is alignment between the investors, the people working here, and absolute alignment on what good looks like and what success is. So it's great to have so many managers willing to put their own personal money into the company as well. And one of the important new people who've joined very recently, and I saw Connor, I don't know where Connor is setting, I hope to hear from her today. It's a little bit too soon, Connor, to put her on stage, but very soon, I will do. So, Mary Ann Packo, is our new CEO in North America, hugely experienced in the industry, very, very strong reputation, both inside Ipsos and outside and has hit the ground running in terms of making progress in our U.S. business where we've been waiting really for her arrival in the summer of this year. So it's great to have Mary Ann on board. And in fact, of course, she's no stranger to Ipsos because she was partnering, I think, with Didier Truchot, back in the 1990s on building panels across Europe. So good cultural fit, understands what we need to do in our biggest market, of course. So that's all from me for now. I'd like to hand over to Dan Levy, your friend and mine, and he's going to take us through the numbers and just give you a bit more information on the top line. And of course, full -- even more numbers will come with our half year results in July. But Dan, tell us where we are now. Thank you.

Thank you, Ben. Good afternoon, ladies and gentlemen. So let me start with a short trading update. The landscape has not really changed a lot since we published our Q1 results in April. We see a globally good momentum with, as you know, 4.5% organic growth in Q1, which was mainly driven by EMEA, Asia Pacific, also all our consumer-related activities and our new services among which Ipsos Digital, which is growing by nearly 30%. We see a more contrasted picture in the U.S., as you know, with, on the one hand, some recovery on the big tech clients, but also a good resilience in CPG clients. But on the other hand, some wait-and-see attitude ahead of the presidential elections in the U.S. which, as expected, impacted our public affair business through the usual electoral cycle, but more globally impacted also the business climate in the U.S. Our health care business in the U.S. is also facing headwinds as it was the case at the beginning of the year, due mainly to restructuring and layoffs in the pharma sectors and also fewer drug approvals by the American authorities. So even if the U.S. market is going to remain difficult for 2024, we have, as Ben just said, appointed Mary Ann Packo and we expect performance to improve in the second half of the year. China. In China, we also lack macroeconomic visibility, and we are pretty much flat in China currently. On the profitability and cash generation side, we will update you more, obviously, when we present our H1 results in July. But I can tell you now that we are absolutely on track to meet our annual targets. So as a consequence, obviously, we confirm our guidance for 2024 of an organic growth above 4% and an operating margin around 13%. If we now take a step back from the quarter-to-quarter results and have a look at the more long-term view, I think we have strong structural drivers that all point towards some improvement in the profitability in the future, and we target a 15% operating margin for the medium term. What are these different drivers? Well, first of all, we have the shift from offline to online data collection, which has existed now for 10 to 15 years on the market and which is continuing. Our quantitative work, as you see on the chart, was 55% online in 2019. It is now 66%, and this is driving the gross margin and driving profitability. The second driver and probably the most powerful and the most important one is the streamlining of our operations. First of all, by optimizing our production platforms, both on data collections and on reporting. This will bring us more speed and also productivity gains. We have several projects internally to do so. and we are making good progress on that. The second point on streamlining our operations is about panel internalization. We want to increase internalization of our panels. Currently a bit more than 60% of the data we use is based on Ipsos internal panels. We want to increase this number. And Michel Guidi, our CEO, is going to take us through this idea, mainly for quality reasons because quality of data is absolutely key in our business, but also for cost efficiency reasons because it will also drive profitability up. And of course, we want to leverage Generative AI models. We have already started to embrace Generative AI, and Michel will come back on a lot of business opportunity Generative AI can bring to us, but Generative AI will also bring significant productivity gains, particularly on the operational tasks. And finally, we expect to have also a favorable business mix effect as we grow quicker in the medium term in the U.S., on Ipsos Digital and data analytics, which are, on average, more profitable than the rest of the group. This will bring us also some benefits in terms of business mix to improve profitability. We are in a very strong financial position. We have pretty much no debt with the leverage close to 0. We have delivered more than EUR 380 million free cash flow over the first 2 years of the plan. We have close to EUR 500 million of undrawn credit lines with maturity above 1 year. So obviously no issue with liquidity and our debt maturities are also perfectly manageable. As you see on this chart, we will, in due time, refinance our EUR 300 million bond maturing in September 2025. And as a consequent of that, we have all the means to finance our strategy. The group has created value for shareholders. Since last time, we saw us together in this instance and over the years. So since the last Investor Day, the share price has increased by nearly 30% from typically EUR 51 to close to EUR 65. And as you have probably seen, the share price has been at its highest level over the last few weeks and months. We have also increased the dividend for the shareholders to EUR 1.65 per share, representing now 32% of adjusted EPS. And we have started at the end of 2022, the share buyback program, and we have consol for EUR 60 million of shares, of which EUR 50 million in 2023 on top of our usual share buyback program for the free share programs for Ipsos employees. We have also gained visibility on the market. Earlier in 2024, Ipsos has joined the STOXX Europe 600 Index. We have now 8 analysts covering Ipsos stock, among which Berenberg, who has just initiated the coverage and our investor base is more and more international. 75% of our free float is currently international. It was 63% in 2021, and we see more and more interest from Anglo-Saxon investors on Ipsos' name. Obviously, all of this is very good for the recognition, for the visibility and at the end of the day for the liquidity. All our 8 analysts are in buy status and showing a strong positive conviction on Ipsos. I have put on this slide a few quotes from the latest notes we had a story deserving a greater recognition, a leader poised to strengthen its position and undervalued source of knowledge. And in average, the average analyst target price is now at EUR 78. So in a nutshell and before leaving the floor to [indiscernible], I would like to say that we have delivered over the last 2 years, a good level of growth given the global context and a very difficult macroeconomic and geopolitical context. We are in a very strong financial position that enables us to finance our strategy meaning acquisitions, but also investment in panels, investment in platforms, and obviously, in Generative AI. We have a resilient and diversified model, diversified both from a geographical point of view, but also from a sector point of view. We have proven, I think, our ability to adapt our operating cost to the macroeconomic context and to deliver a good level of profitability. The operating margin at Ipsos has now been around 13% for 3 years in a row where it was typically 10% before 2021. And beyond that, as I explained before, we have a lot of interesting drivers to improve profitability in the future. And we have increased the value for shareholders with a good performance of the stock price, higher dividends, but also more visibility of Ipsos on the stock market. I thank you for your attention, and I leave the floor to Jean-Michel Mabon, our Head of M&A. Jean-Michel?

Thank you, Dan. So good afternoon, ladies and gentlemen. So I'm Jean-Michel Mabon, I lead the M&A activities for Ipsos, and I will update you on our strategy, but I will also tell you where we are in our plans as of today. So a few words about our strategy. Of course, it's very consistent with what you have seen before. The objective of the M&A activity at Ipsos is to contribute to the growth, but also to strengthen some of our segment of business. We have identified 4 segments that we want to strengthen. The first one is buy market. So as you've seen before, the U.S. market is our first market. It represents about 1/3 of our revenue, but it's also the first market globally, where it's about 50% of the global market. So we have still space to grow there. U.K. and France are also a key market for us. We have a strong position there, but we want to strengthen some of our businesses within these two markets. In Germany, one of the largest market in the world as well. We believe that we could have a better market share there. So we are active in Germany to strengthen our position. And India, of course, one of the largest market as well in the world, many opportunities there. So we will be active in India as well. The second segment is by service. So as you know, Ipsos is organized with multiple service, as you've seen before. As far as the M&A activity is concerned, we have identified 4 key priorities. Public Affair is the first one. So we have today a very large public affair business at Ipsos. We are the largest public affair organization globally, but we want to keep strengthening our coverage. So we will keep consolidating our positions in some key markets to make our global public affair business stronger. Health care. This is another large service line at Ipsos. We work with all the big pharma, all the med tech companies, but we want to offer them more services. So the purpose of the M&A activity there is to identify services that would consolidate our offer to do more with our clients. And two other services, which are smaller but growing with good perspective. So advisory, where we help our clients to make strategic decisions based on research data and user experience, which is another area where clients are asking us to help them to make decisions about features on their website or on the apps. The third segment is more related to tech and data science. So as you have seen before, we control our end-to-end process. So of course, they are doing from the data collection to the data delivery. There are many steps there that we can improve in order to deliver faster our services or in order to be more efficient internally. So contribute to improved profitability. So it goes from access to respondent. So how do we identify and engage with the respondent, but also all the platforms that we have to analyze the data and prepare the data analytics for the client, and of course, data management and data analytics, which is the capabilities that we need to aggregate data from different sources and deliver insight to our clients. The fourth segment is more opportunistic. So we have key priorities, but we are also open to opportunistic acquisition, could be by geography, could be by segment of business, could be local, could be multi-local. So we will keep an eye on this as well as soon as there are good opportunities at a good price. So a few words about what we did over the last 18 months. So I took the job less than 2 years ago. So we have been very active over the past 18 months with the acquisition. We did 10 deals in '23 and beginning of '24, very aligned with our strategy. So I talked about Public Affair. We did 4 acquisitions in Public Affair in different regions of the world; so Australia, Ireland, New Zealand and Netherlands to strengthen this global network of Public Affair. We did 3 acquisitions in the tech and data management area. So one in the U.S., which is a B2B platform to access respondent; Germany and the U.K., which is data management and data analytics to aggregate data from different sources and help our clients to make better decisions. And Crownit, the last acquisition that we did a few weeks ago in India to access difficult targets in the Indian market where not everybody is connected through traditional online means, but through mobile. So thanks to Crownit. We are able now to access very different targets in India. One acquisition in the health care area in China. And as I said before, we have been opportunistic as well with 2 acquisitions. So one in the U.S. with the new vehicle customer study. So this is a study that track behaviors of car buyers and how they move from one brand to another one and so on. So the clients are the automotive sector. So we already work with this client in different part of the world. Thanks to this acquisition in the U.S., we strengthened the relationship with our client in the U.S., and it will help us as well to add more services on top of this study. Another, let's say, opportunistic acquisition in Africa. So Omedia covers both Senegal and Ivory Coast. And this acquisition will help us in the Francophone area to offer services to some Ipsos' existing clients. Some are French, by the way, very interested with these 2 countries. So all in all, these 10 deals will represent about EUR 60 million of additional revenue in '24, which is about 2.5% of the '23 revenue. So now I will give the floor to 3 of our new leaders. They will tell you about who they are, but also how do they feel after few months at Ipsos, what do they see as opportunities for them to consolidate the business at Ipsos. So it will be Luke for B&A, the CEO of Ireland; Gibbon from the Netherlands, who leads I&O; and Tom who leads Germany and the U.K. [Presentation]

Thank you. So three great acquisitions. We are working on the next one. So we will keep moving on the acquisition front. We have a funnel. We have identified targets, and so we will keep working on it in the coming months. Of course, I can't tell you more today, you know why. So we will update you when the time comes. And I'm happy to take any questions at any time after. Now I will leave the floor to Eleni who will talk about another very important topic, which is the clients.

Thank you, Jean-Michel and good afternoon, everyone. As this is the first opportunity I've had to actually meet most of you, although it's virtually, I did want to take a moment to just introduce myself. So I have over 30 years of experience in market research and advertising technology, running geographical P&Ls around the world, but also running large client organizations. Having joined Ipsos about 6 months ago now, I can sincerely say I have saved the best till last in terms of the best company and absolutely the best role. So in that context, it is very much my privilege and my pleasure to be here today as the Global Chief Client Officer. Within this role, one of my accountabilities is looking after our top 20 global clients. And as Ben mentioned earlier, they represent about 1/4 of our global revenues. They work with us across multiple industry sectors from automotive to technology, consumer packaged goods, luxury, public affairs, health care and so on. They do work with us across almost all of our geographies, and they leverage all of our solutions within our service lines. And again, as Ben mentioned, for these top clients, we also have a very impressive and consistently high satisfaction level of 9 out of 10. So that's not my only accountability. I have accountability for the entire client organization across the globe. So this is made up of some dedicated chief client directors who are dedicated to the top 20, and we geographically locate them near our clients headquarters, but they are complemented by a client organization in our markets helping global and regional and local clients deploy locally. The responsibility we have is to really focus on client relationship. We need to be experts in their business issues, their opportunities and their challenges, and we partner with Christophe Cambournac, my colleagues, teams across all of the service lines to really ensure we bring the right solution at the right time to drive client impact. And we are the champions of client first. You probably know that client first is an Ipsos value, but it's also a way of doing business. We really do drive the culture. That's how we behave to become trusted partners. We really do drive impact to ensure that our delivery is distinct and unique and end-to-end, and we work with our clients to ensure growth so that we win with them, they grow and we grow together as partners. So I spent the first 6 months. It was an opportunity as someone who was arriving to really talk to the clients and understand what it is that they value most about Ipsos. So almost without exception, one of the first things client tell us is that they really respect our integrity. We have independence, but they trust us and they like doing business with us. They do respect us as researchers, and we are the home of researchers. And they know we will always be thorough in terms of our methodology and approach, and they do recognize that we have the ability to understand the macro and complex questions within which consumers, patients, citizens take decisions. We are becoming increasingly unique in terms of having dedicated client teams, dedicated sector champions that work in partnership with experts to meet the client needs. And the fact that we are both global and local in terms of presence is very important because it means that we can partner with our global counterparts and ensure that there is global consistency and efficiency, but we are also present locally in the markets to ensure we understand the nuances and the culture of execution to be successful in market. And then finally, our unique matrix of technology and service line solutions and client dedication is unique and does allow us to really focus on impact. And that actually brings me to what I wanted to share next, which is sharing with you real life examples of impact that we have with clients. There are many of these, but we've picked 2 or 3 that we felt would be really worth sharing with you. So the first one is a brand called Ferrero, who many of you all know. They have a brand within their portfolio called Nerds Candy. And this is actually an example from the U.S. They needed to pick the best creative, the best ad and then they needed to deploy that ad, so that ad at one of the biggest advertising opportunities of the sports calendar in the U.S., the Super Bowl, where over $600 million is spent on advertising each year. So you can see the quote there from one of the insights leads. The testing we did for them was fantastic and enabled them to lock down the creative they were going to use. And in fact, the KPIs, the metrics here speak for themselves. Following that campaign, we were able to show that there was a 100% increase in people mentioning spontaneously the brand that they've seen in the advertising and over 40% said they would purchase the brand post seeing the advertising, and we all know the claim does translate to real sales. So a high impact with clear metrics. But actually, I'd like to hand over to Shannon, the client herself, who will actually hear you a little about the partnership and the impact we had. So roll the video, please. [Presentation]

Thank you, Shannon. So the next example takes us to a different client, Unilever, one of our other most important and valued global clients. We were working with them in Latin America and the premium sector of mayonnaise was increasing and growing materially and Unilever realized that they were losing market share because they did not have a product that could compete in the premium sector. So this time, we partnered with our innovation experts, and we help them to create and launch a range extension that could compete successfully in the premium sector of mayonnaise within the region. Again, the metrics speak for themselves. Volume growth was over 9%. There was also double-digit value growth. And in fact, Unilever was so delighted with this work that their own CFO presented at their own Investor Day, this example, an example of targeted growth that improved their market share overall. So another great example. You can see there the quote from the client from Reynosa saying that we help them increase both their volume growth and their pricing growth. So another real successful and tangible impact story. This brings me to the final example. I'm going to change different parts of the world. So now we're in Australia and now we're in the media and our advertising space. So I'm sure many of you will know that digital advertising is growing exponentially as complementary to linear TV and the traditional advertising that we're all used to. So the market in Australia, just the digital advertising a video across connected TVs and laptops and mobile phones and all of those devices is worth AUD 3.8 billion, and there was actually no way for marketers or broadcasters to measure and size all of those cross-device advertising opportunities. So we worked with our audience measurement experts to create the first ever objective measurement that allowed marketers, broadcasters and advertisers to size the audience and then know whether they reach the right audience at the right time. So this is huge in terms of currency, and this is huge in terms of impact. Again, I will hand over to Gai Le Roy, who is actually the CEO of IAB Australia, which is a global advertising body that helps advertisers and she will express this better than I. So roll the video. [Presentation]

Thank you, Gai. So in closing, I think you've just heard directly from our clients about our integrity that we're trustworthy, our technology, our innovation our global local partnerships and our expertise in end-to-end and different solutions. And that's why I passionately believe in client first because by placing clients at the heart of everything we do, it really does mean that in terms of our culture impact and impact on their growth, we are uniquely positioned to help our clients perform and succeed and win in the marketplace. And with that, I would like to now hand over to my colleague and friend, Michel Guidi, our COO, who is driving the charge on all things, technology and AI. Over to you, Michel. Thank you.

Thank you very much, Eleni, and hello, everyone. I'm Michel Guidi, the Chief Operating Officer or COO for Ipsos. This is my second Investor Day. I'm going to tell you more about our tech and AI road map, the progress we've made since we last spoke a year ago and also what's coming next. We'll start, obviously, with Generative AI or Gen AI. And I'll start with the clarification. It's important to distinguish artificial intelligence or AI and generative artificial intelligence or Gen AI. AI existed long before Gen AI. The market research industry has used it, and Ipsos has used and pioneered it for many years. We'll still use it today, and I'll allude to it in different parts of the presentation. It's still here and it's still very important. Of course, what's new is Gen AI, which has taken the world by storm 18 months ago with the advent of ChatGPT, which opened a new era. And we at Ipsos, immediately saw the trend. We do many things at Ipsos. One of them is to catch the trends when they start forming. And we immediately saw the potential for transformation of the technology. This is similar and unlike other technological buzzwords of the past, this one is a big one. Similar, we think, to the Internet when it happened or maybe even the personal computer. So we decided to embrace the new technology and to lead with it. We've published a lot of good content, which is available on the Ipsos website and without your Ipsos contacts. And we've done a lot of work, which I'll be presenting to you today. When I talk about Gen AI, I hope I convey the passion and enthusiasm we have for the technology, but it's also important to talk about the fact that we embraced it with our eyes wide open, wide open on two things. The first one is a reminder of Amara's Law. A new technology is often overestimated in terms of short-term impact and it's underestimated in terms of its long-term impact, and it's very, very valid for Gen AI. We are 18 months into it. Has everything changed? No. Is it really transformative? Absolutely. We'll show you examples of how we started the journey here. Something else that we keep in mind, alongside our enthusiasm, are the limitations and the risks. Gen AI is like a superpower. Used well, it can do amazing things. But if you don't use it well, it can be very risky or dangerous. Some of the risks we share on the screen. The first one is security. There have been problems with companies around the world in terms of leaks of intellectual property or IP or confidential data because people are using what looks like the black box and they don't know where the data is going or what it is being used for. Hallucinations is one of the tech words of 2023, meaning the technology looks amazing in how it responds to you. It looks very certain, very sure of itself. And sometimes it's totally wrong. It's fabricating things. There's an example on screen, a famous one, where a senior lawyer in New York City was pleading in front of a judge having used Generative AI to prepare his trialing in front of court and it was entirely fabricated. The laws he was referring to and the jurisprudence he was talking about was entirely wrong, but he had checked with Gen AI. Is it correct? And yes, it is. Unfortunately, the judge -- or fortunately, the judge caught it and it, of course, created a bit of tension. Bias. We, at Ipsos, work to understand, leverage and harness the complexity and the diversity of the world. Our clients of course wants too, but we'll talk about ESG later. There is a real risk with Gen AI as a technology to have a regression to the mean. And with that, the problem of potential mediocrity, but also bias. English as a language, gender bias, racial bias, everything we are trying to fight against. So there are real big identified risks with the technology. Enthusiasm is good. Method is also important. So what method have we applied to make sure not to fall into any of these traps. We've applied our evaluation framework, which preexisted Gen AI, around three nice words: truth, transparency and justice. What does that mean? It means that we're asking our people and we're telling our clients to keep the human in the loop at every step, to do what? To check for accuracy. Do not believe it because it says so. It's tech. Have humans check the outputs. Can you explain on transparency. Can you explain what this is telling you. If you can't explain it, do not leave it as being a black box. It shouldn't be. If you cannot identify the source and come back to the real human context, challenge it again. Last, are you sure? You are on the right side of justice? So ethics, fairness and sometimes just the law. Again, you probably have read or heard about copyright infringement issues, which are ballooning or people using famous people's voice or image without due permissions, which is creating in terms of problems or people using content without necessarily the right permissions either. So keep them in the loop and apply the right framework. All that said, and understanding that Gen AI is not magic. It's just tech. It can be amazing though, and when can it be amazing? When you apply human intelligence and artificial intelligence together, HI and AI. And good news, we at Ipsos are very well positioned for it. So last year, as the technology was fairly young, still is, was very young then, we were asked, is it a disruptor or is it an opportunity? Well, I'll tell you about how much of an opportunity is this for Ipsos. But you can't invent yourself being a specialist in the field. We hear from our clients today that we are way ahead of the competition head and shoulders. Why is it? Because we apply human intelligence and artificial intelligence concretely. What does that mean? Not everyone can do that because, first, you need a safe, secured and agnostic platform that leverages different models. I'll tell you about what we've done on that in a second. That means investment and a strong technology team. If you have that, you can move to the next box. You need data. If you don't have data to add to the system, you'll be relying on the same publicly available data as everyone else. So you're as good as your next competitor potentially. We, at Ipsos, are sitting on a treasure of data. Ben alluded to many different sources of data. We have offline data and online data, quantitative data and qualitative data. We have social intelligence data and [indiscernible] data and audience measurement data and passive data. Well, we are swimming in data, and it's curated high-quality data. I'll come back on that. I'll take 30 seconds to talk about offline data. So what is it? We talk to real people in real life. That's still a significant proportion of our business at Ipsos. Why do I mention it in the context of Gen AI because that is data that is not online. It doesn't exist online. So your next competitor doesn't have access to it. We historically and still today, go and talk to people, digitize that and make it available to our models, which very few other companies in the world have and even fewer at the level and breadth country-wise, and volume-wise that we do. Okay, you need the tech, you need the data. Then you need data science. You need teams of established data scientists, to apply the secret sauce, what is that? Apply our proprietary analytical frameworks, our proven methodologies, then select the right model. I talked about an agnostic platform, and I'll talk about ours in a moment. Pick the right model for what you're trying to do, not all models perform the same. There's a governance element into that as well that is increasingly important. Make sure you do things right, that you locate the data in the right place, for instance, given what your client is asking you or needs from you. So you need the data scientist to create the secret sauce. Now you can move to step 4. You have something that is robust and can be used. Then you need to iterate with it. It's only a tool. It's only be as strong and as smart as what you input into it. That where prompt engineering, which is another term that has become popular, comes into play. What is it? It's the questions you ask. It's the request you make to the machine in a way. We are, at Ipsos, experts in the art of the question, including the nonobvious ones. That's what we've been doing for a living for the past many, many years. We know what questions to ask. And when saying questions, it's not just a simple question once. It's iterating complex questions with a lot of context. I'll show you examples of that in a second. When you've done that, you can move to step 5, which is you need humans in the loop to verify before you can activate because, again, it is not magic, it is not a black box, it's just a very powerful tech. So make sure or that it produces a very powerful output in a positive way and we have the expert for that too, already at Ipsos. Then you can move to the last one. You need a brand that is a trusted adviser size to your clients. Why? Because of the magic haze or the magic cloud surrounding the technology. You have to trust the people you work with when you deal with powerful technology like that. We are a trusted adviser. We deal with serious business. We're known for how safe and sure our results are, and that's extremely important in that space. Together, the 6 points you see. They compose a competitive advantage, a composite competitive advantage for Ipsos and a barrier to entry. If I was to leave Ipsos tomorrow, which I won't, and start my own business with 3 smart technologies, could I create a new [ greedy ] inc., company that looks really good using tech, yes, I really could. Could I produce anything close to that? Absolutely not. Because I would use -- I would need the data, which I don't have. I'm only stuck on stage 1, assuming I've raised enough money to invest in stage 1. I don't have the data. I don't have the scientists for the many, many years and the methodologies and the property frameworks. I do not have the prompt engineering efficacy. I don't have the experts and I don't have the reputation. Would you trust me with your important data and big decisions? I would have to be really good in sales. That's why our clients are telling us that we're ahead of the game. This is an example from a client a few weeks ago. It's their questions. It's a client, by the way, that we've been working with for over a year on the matter. So they are a mature client. So I'm not talking about all the necessary first steps about security, all the things I talked about before. We are further down the road with them already. They've already established that we are safe to play and they are safe to play with us. This is what they're asking us. They are on the very advanced maturity scale of our clients when it comes to the new technology. They're asking us, Ipsos, what is your position and road map on AI? And they really mean Gen AI in this case. How is Gen AI integrated in your research process? Any learnings and any dos and don'ts? Tell us what we should do and what we should avoid? Do you have tools and solutions that are available? And what's your road map? Can we pilot together? And can you help us transform. That's the sort of conversations we're having more and more with our clients. And how do we answer. We answer leveraging Ipsos Facto, which is our Gen AI platform, which we launched just after last year's Investor Day. We announced it, if you remember, last year, and we launched it right after. We launched it with 3 objectives in mind. The first one is to democratize. So we made it available right away to all 20,000 employees across Ipsos to embed it into their day to day. Nowadays, well to this date, 72% of our colleagues globally have used the platform and adoption keeps going up. What do we want to use -- what do we want them to use it for? To operate productivity, to operate faster and better for things such as transcription, codification, extraction of themes, translation, summarization, generation of top line evaluation, et cetera. Speed. There's a third leg to it, which is maybe the most important. It's the innovation part, making the platform available to the service lines organization led by my dear colleague, Christophe Cambournac, so that they can do their amazing work of creating with our clients' new solutions and offerings, which we'll show to you in a second a few examples of. This has worked so well with the Ipsos that our colleagues naturally have showed it to clients and some clients we're like, "Wow, can I see? Can I try?" And so we're making Ipsos Facto now available to our clients as well when it was initially thought more as an internal platform. It's also become a household name, an expression in the common language. When you go around in Ipsos offices in any country, you hear Ipsos Facto come back in the conversation all the time, and that's usually a good sign. So what can you do again? You can do a lot of things. I will go quickly over that slide. What's important is to understand that Ipsos Facto is based on 15 distinct models, and it keeps evolving as the new models come around from 4 companies: Google, Open AI, Entropik and Mistral, French company, the last one. We also have, by the way, a version of Ipsos Facto in China for our Chinese colleagues with all the usual China requirements applied to it. People use Ipsos Facto in the day-to-day for things you see on the screen. There's a little button to the left, I don't know how visible it is, that says prompt library as well. So I talked about prompt engineering, and I'll come back on an example of a prompt in a second -- well, now. So when I talk about prompts, again, it's what you need to input into the machine, into the system, into the platform. What comes to mind maybe is can you create a picture of a nice cat with pink ears and something funny because that's what you usually people start with? Or can you create a recipe for -- we, of course, work in the insights industry and in market research, so we would apply it for market research. But my silly cat example can be translated into what we do. You can use it for a very simple request, but that's not what you do when you start using the tool. This is just one prompt after of 3 ended prompts that we delivered to our clients -- to one of our clients recently for a new piece of work we did around knowledge management and curation. So they gave us a large body of documents, and they asked us, "Can you make something out of it because we have all these documents, we don't know what to do with." And so we applied, again, 300 distinct prompts, this is just one of them, to take the substance out of it and give them something that they could actually work on. And then create insight or launch insights exploration against. You can see that it's quite complex. There are frameworks and ways to do that, but it's not easy. Our average prompt length currently at Ipsos is about 2.5 trending to 3. It will become 3 and more. Our longest prompt that I know of is 33 pages long. And again, it's trending upwards. So prompt engineering is not someone that -- something you can invent overnight. It's a real expertise. We think it's a game changer for the industry. And that's what we hear from our clients. And why? Because we're focused on our industry. We're a market research company, and we're insights experts. We've created the platform for ourselves to begin with and for our clients now with insights and market research in mind. And so when they use it, that's what they express getting back to us. It really speaks to me because it works for me in insights and beyond insights for marketers as well, democratizing access to our insights and data. I talked about innovation and the amazing work done by Christophe Cambournac and his team -- and his teams on the service line side of Ipsos. I'll give you a few examples. So this is something that we're working on jointly together from the tech, but also it's really the work of our service lines. In 2024, this year, we have launched or will launch 12 and counting, 12 new products and offerings that embed Ipsos Facto and GenAI as a technology. And you can see it's all across the spectrum in innovation, brand equity segmentation, market and consumer understanding. There's one more circle that will -- that is coming soon that will be communicated at the half year results by the way, in creative excellence. I could spend an hour just on this slide, I will not. I could talk about Persona Bot, for example, at the bottom right and how exciting that is to take data that is today still, it's on the shelf, and from segmentation data. So you're a company or a brand, you segment your customers in different subsegments. People that like this, people that do that, their perception of the world. So that's what we do for our clients. And usually, it's great work. It's expensive work, but it stays too confidential. It doesn't -- it's not democratized enough, and that's been a problem for the industry for a long time. What we can do with that now is create agents, which we call Persona Bots where we're going to make them come to life, and you can converse with them. We insist to our clients, it's directional. These people done, representing segment one doesn't exist in real life, it's just the epitome of segment one. It's a fantastic way though to democratize and to increase the return on investment of the work that we've done for our clients. And they love it. So we are working on the notion of agents a lot, again, in a conversational and directional manner. Again, I just took one example. I'm going to now leave the floor to two of my colleagues, Mathilde and Virginia. Mathilde is one of our best social intelligence experts at Ipsos and Virginia is the leader of our innovation service line globally, and they will tell us more about Signals GenAI that turns vast amounts of data into action-ready insights in less than 60 seconds. And then InnoExplorer, which enhances the speed and success rate of concept and product development. And with that, Mathilde, over to you.

Thank you, Michel. So good afternoon, ladies and gentlemen. It's my pleasure to share with you the capabilities of Signal GenAI. So what is Signal GenAI. It's a do-it-yourself platform, but we put in the hands of our clients to find relevant insight with a million of social conversation and this in just a few clicks and a few insights. So social conversations have always been a gold mine. So social conversation, just to recap, is basically what you're writing, what you're posting on your social media like Instagram, Reddit, X, Twitter or even TikTok. So social conversation has always been a gold mine of information for our clients to fit, for example, their innovation strategy, but it requires skills to use the social listening platform, and it requires time to find to get to the relevant information among the vast amount data and conversation. Now with Signal GenAI, our clients are able to find in just a few seconds and a few clicks relevant information in -- among this million of conversation. So Signal GenAI is simple, so simple that kids could use it, just a few click. Signal GenAI is fast. You know you can get to the information in a few seconds. And most important, I would say, Signal GenAI is secured. As Michel said, our prompts have been built by experts. Some of them are up to 30 pages. And even more important, I would say, the answers to this prompt are grounded in real social conversation coming from real consumers and real citizen. So how it works? So basically, what we've done is that we have a library of prompts, we have built a library of prompts that looks like buttons. So for example, all these buttons cover all the main business question of our clients. So for example, with one of them, you can deep dive into market understanding or consumer understanding. You can go to trends identification. You can even investigate all the questions related to innovation. Understand your brands and comp performance or even deep dive into specific frameworks like, for example, ESG frameworks, what people are saying about the different dimension of ESG for your brand, your market or your segment. And as I was saying, the magic. So I know Michel say, we can't say that because it's not magic. But sometimes you have a feeling of magic is that you just have to click on one of these buttons. And you will see emerging, for example, new popping ingredients, new format, new product or events, for example, emerging brand, but you have to compete with them. So to be more concrete, we're going to make a demo. So let's imagine you work in the food company and you want to investigate what are the latest trendy ingredients within the snacking universe, which is a pretty growing segment within our clients nowadays. So very easy. You just have to log on into the platform. You type your scope of investigation like you would do on Google Search basically. So here the [ gain ] snack. Then, let's say, we want to scrap the information from the conversation from the last 3 months, what people are saying on consumer media for the last 3 months, then you arrive to the business case. You click on this one, the ingredients buttons. And you will see in a few seconds, the tool is able to clusters, to groups, millions of conversation in 20, I think here, different ingredients that people are discussing right now online. Of course, each of them come with an explanation, like, for example, here with [indiscernible] and I would say, really, the beauty of the model is that we are able, that you can see on the right-hand screen -- on the right side of the screen, you can see the real conversation behind it, which is really important. Once again, we need to be grounded in real conversation. So it avoids hallucinations and no black box, really important. And this tool is so easy to use, that it extends the usage of data and insight beyond our traditional clients to be over function within our clients' companies. And I will be finished by that. I had the chance 2 weeks ago to share our presentation with one of our top clients in the beauty industry. And this project was about investigating what mental wellness means in beauty, big subject. And our clients said, if we could have used a Signal GenAI earlier, it would have saved us a year in our quest to understand the importance of mental wellness in beauty. And we have only scratched the surface, maybe 20% of the tool's potential. Thank you. I'm now handing you over to Virginia who is our global service line leaders for innovation, and she will tell us more about InnoExplorer. Over to you, Virginia.

Thank you, Mathilde. I'm delighted to be here today to talk to you about our new solution, InnoExplorer, which uses generative AI to really truly transform the innovation process. It allows us to create new ideas, concepts, products and packages in minutes. It accelerates the innovation cycles from months to days and very importantly, it boosts the breakthrough potential of new ideas and concepts. And this is a big source of revenue for our clients. It can be applied to any industry, including consumer goods, durables, tech and financial services. So what makes InnoExplorer different, better than the competition? Essentially, we give off-the-shelf models on the job training. And this includes using authentic consumer experiences, with fresh data that is relevant to the product category, representative of the target audience and traceable to avoid hallucinations. We then inject our innovation knowledge, our 45 years of innovation experience to still into engineered prompts. And then finally, we fine-tune the model algorithms with our vast database of 170,000 ideas and concepts so that the model can learn from success and failure. Most of the clients we talk to say we're way ahead of the competition. So how does this process work? Well, first, we decide what area we want to innovate in, what's our category of interest, then we collect the fresh consumer data. We discover unmet needs, we generate ideas to meet those needs. We then develop concepts, build superior products, generate bold packaging, and accelerate market transfers. And what this really means is we can determine if a product is successful in one market, the likelihood that I'll be successful in another, which has always been a very big business question that we've gotten from clients. All that is made possible by using the Ipsos Facto platform. But most importantly, what InnoExplorer AI really does is it enhances the speed and success rates of concepts and product development for better in-market performance. When we take the human intelligence and the artificial intelligence, and we put them together, the concepts tested have 9% higher trial potential and products tested have 10% higher overall liking versus not doing it. These percentages are significant and really game-changing for innovation. And our clients have certainly noticed they only have positive things to say. Today, we've conducted 60 projects across North America, Europe, Asia and Latin America. And we've gotten feedback from clients that says, it's mind blowing, take my money. This is gold dust. It looks mastered, and it goes on and on. So very successful service that's really delighting clients. And I have to say, in my whole innovation career, this is the most exciting time ever for innovation. So with that, I guess I'll turn it back over to Michel.

Thank you very much, Virginia. We've tried to show you what we're doing, what we've done and what we're doing and a glimpse into the future of Ipsos leveraging GenAI as a technology for both productivity and speed and also for new products and offering and growth, and we are very excited. So we hope you will continue to watch the space because there's more coming on the space. Knowing that, again, it's still early days. This is not magic. And knowing that we have built barriers to entry, not everyone can do this. Very few companies can do what we are proposing on market. I'll talk now about one of our [ fundamental ] data and the quality of data. Generative AI is a wonderful technology. It relies on data as we saw. You need the right data to feed the models and by the way, way beyond just GenAI. Data quality is of paramount importance, has been, is and will remain a competitive advantage to Ipsos and that is to be understood in a context where fraud is getting increasingly sophisticated and dangerous in part because of the other side of the [ medal ] of everything I just showed you because technology is getting more democratized and better at it. And now I'm going to show you an example of fraudulent behavior. This is not on the Ipsos panels, by the way, it's a client that was victim of fraud that centers this. What are we seeing on screen. A questionnaire in French, done by somebody not based in France, not speaking French, using technology to automatically translate a bit more than that to even answer open-ended questions. So what do you think about the brand, for example, once -- but not once, in batches, 6, 12, 50 times at once with different answers every time, which you needed a lot of humans to do in the past and you need just technology to do nowadays. Of course, that creates massive issues because none of this data has anything to do in our data sets as an industry. It's dangerous if we feed it to our clients. And that is why we are renewing our commitment to security and data quality at Ipsos by continuing to invest. And that is where everything that we saw before, our positioning as a company that we heard from Ben and our financial capabilities that we head from Dan are very important because you need to be able to continue to invest in that. We already do digital fingerprinting to identify the machines and the location, et cetera. We already were doing similarity checks to make sure that if somebody tries to enter in our panels multiple times, we spot different point of similarity that raised suspicion. We were already fighting against behaviors. And as I told you last year, I hope you remember, we are the only company in the space that is applying multi-factor authentication at the registration level in our panels. So we check you before you come in. That's great and that leads to us being considered, I believe, as the best or one of the very best panels in the industry. It's not enough, though, because as we saw technology is getting smarter and smarter. So we are enhancing our similarity checks by using Ipsos Facto and GenAI to fight GenAI and the potential of similarity checks that are less spottable because GenAI makes it better, still wrong, though. We have implemented activity and tenure because the new technology allows people to do things faster, including swarming our systems. We are going to implement in short order a panelist score to follow people's activity over time and make sure that we can build trust with the people that are responding to our surveys. We had something already today and we're making it better. And we're tending towards something we hope to do before the year is over towards ID validation because the only way to really stop fraud is to stop people at the gate. There's other elements of fraud, by the way, that I will not dwell on to. But we want to make sure that you really are who you say you are. You really are Michel, based in France, that age, that family composition, et cetera. So we renew our commitment to this. It's a financial commitment, it's an operational commitment, it's a technological commitment, and it's very important because none of what I told you before about our excitement and the opportunities offered by the new technology can work without the second box in my competitive advantage slide, which is the quality of the data we provide, and we always have provided. I'll finish the segment about technology at Ipsos. Progress and road map with speed. I talked about productivity and speed before. I'm going to focus on 2 important levers that we are continuing to pull to gain speed in the way we operate. The first one is our amazing SaaS platform, Ipsos Digital. The work done by my colleague, Andrei Postoaca and his team is truly outstanding. You've heard about it, Dan referred to it. It is very good for not just financially speaking with higher margins, it's also very good for speed. We accelerate the way we work with it. Good news. We are continuing to invest and to develop Ipsos Digital. You can see a few numbers on the screen. There are more countries covered than ever. The number of client users on the left-hand side keeps increasing. The number of unique companies using the platform on the right-hand side. The signs are green. The revenue is growing 36% year to date on Ipsos Digital. It's going well, and we're continuing to invest in it as shown by what I'm mentioning below the chart, we just, this week, launched FastFacts Light based on Askia's New Generation Interviewing Engine. Askia is one of our recent acquisitions, as you may remember. And that is a new generation platform that will enable us to increase the speed and scalability of the platform and also to improve respondent experience for better quality and for better financials too. The scalability sounds like why does it matter to me as an investor in Ipsos? Well, we have clients coming to us now asking us, can you do 500,000 interviews in one day, one hour, sometimes concurrently? In the past, any company, Ipsos included, would have had to throttle that and take it easy. And when you build a modern platform like we're doing, you can take the load and that opens new opportunities for our service lines and our client organization to go and get bigger deals. So it's a small step, but an important one on our way to being able to do a lot more on Ipsos Digital. Finally, we are also investing into our modern data stack in 3 different ways. First, we're investing into our global data hub. Again, clients expect us and we want to be able to connect our different data sets. Are we there? No. Will it take time? Yes. No company today, as it figured it out entirely, we're making investments in that space, and we're in a good position with our new global data hub, which will not only enable us to connect the data sets together, ours and our clients, but also decrease the cost of data at Ipsos. We're investing into a modern data pipeline to accelerate the speed of delivery of data. That's something that the market research industry as a whole has suffered from for a long time, Ipsos included. I'm giving you examples at the bottom of early wins. This is extreme projects, by the way, that we run for some of our largest clients. One of our global tech clients. One process, in particular, for our largest study, I think we run was taking 24 -- 25 hours end-to-end, too long, way too long, very complex, lots of countries, massive data. We are -- we've reengineered the whole thing, and it's now 15 minutes. Another example, with a global financial services client of Ipsos, where there too, one process that used to take 10 hours is now 10 minutes. It's early wins. It doesn't mean we're doing it at everywhere, that's our ambition. We're investing into making that reality across Ipsos. And finally, we bought a company called Infotools a few years ago. It's a great company with a wonderful data visualization product called Harmoni, and we are working to further develop the product, but also make it available within Ipsos and to our clients in order to have a better visualization platform, which is good commercially, but also enables more automation in the way we work and our clients work with us. So that concludes the tech and AI segment of this Investor Day. I hope you come out of it with more information on what we've done and what we're up to. And I hope I've managed to share the excitement that exists within Ipsos with applying new technology to our benefits, again, for productivity and speed and for growth. And with that, I'll hand back over to our dear leader for ESG segment.

Thank you very much. So you've just seen a lot of slides. I'm going to be brief, but I did want to end this by just talking about life and all of us on this planet. Our latest global trend survey shows that 84% of people around the planet believe that company should do more than just make profits. We know making profits is important, and we're doing our best, as hopefully, we have demonstrated. But actually, we're also here because we're not investment bankers. It's nice to see some investment bankers in the room, but we're interested in people. We're interested in the planet. We're interested in what makes people take. And we're interested in making a difference. So on that, ESG and just what our impact is and our relationship with the world is something that's pretty fundamental to who we are. We study societies. We study people. That's what we do. And we were one of the first companies to sign up to the United Nations Global Compact on sustainability back in, I think, 2008. And for us, ESG works on 3 levels. So the first actually has commercial benefit for us for the hard-nosed investors in the room. We can help our clients with their own ESG progress. They are under more and more pressure from regulators to change what they do to make sure that their operations are more environmentally sustainable, and we are working with many of them. We are also, of course, as a market research company and as part of democracy these days, one role we have is to actually let the world understand how ready humans, citizens, people are to make the changes that they will need to make in the next few years if we are to avoid irreversible temperature feedback loops. And finally, of course, as a company, we need to do what we can to reduce our own impact on the planet. So on that, for our clients, we have built a set of solutions that help them understand the things that they really need to focus on in terms of their business to orientate themselves, to help them understand how they should act, but also how they should talk about it because you can easily get yourself into deep trouble with things like greenwashing and consumers are more and more skeptical of claims made by brands. We can help them mitigate risks and focus and understand the areas where actually there is opportunity in newer, more sustainable products and services. And we can say, well, what is actually going to -- what changed or the different changes you might need to make will have the biggest impact on your business. And finally, we can see how all of this is impacting on their brand as both an overall -- in terms of how consumers see the company, but also how investors see it. So we track that too. We have solutions for each element of what people are trying to do. And for us then some examples working for the U.K. government on how the country will make a transition from where it is now to net zero. What would that actually mean? What does it practically look like if you take out all fossil fuel production from your energy supply. For L'Oreal, we're looking at -- it's a huge -- beauty is a huge and growing industry on the planet. It uses lots of small plastic containers. What does that actually -- and huge quantities of ingredients in manufacturing? How can they be more sustainable? For the Indian government and Minister Modi, of course, launched this survey the first time last year, we continue with it. Can we get clean? How do we get clean water sustainably to the largest country on earth for global tech clients all over the world. How can we help people who've got disabilities. They can't see necessarily. They are deaf. They have various challenges to use technology in the same way as everybody in this room, how can we improve and optimize accessibility? And for many of our clients, we're working on what are the materiality impacts of some of the changes that they need to make. So we're doing this all over the planet. This work grows every year, and it's one way that we can help make a difference. We publish regularly for our clients both on -- whether it's social things around, for example, gender, a hugely contested issue in countries like America, accessible UX, looking at the power of ESG for brands. What people's -- how people's attitudes are shifting, celebrating and understanding Earth Day, you name it. We are looking at how we can help clients, governments, brands, take people with them as they make the changes that they will have to make over the next 20 and 30 years. And so in terms of what we're doing, the things that where we're actually making a change at Ipsos, just a few things. The first is this is our progress towards net zero by 2050. We have committed to the science-based targets program. We've seen a 36% drop in the last 4 years in our Scope 1 and 2 emissions that we have a fair amount of control over, a 15% drop in the Scope 3 emissions. Those are, of course, the larger part of our emissions produced by our suppliers. But clear progress, lots of heavy lifting to be done after 2030, but we are on the way and it's something that's very important, in particular to our many, many young people, of course, who are going to be on this hotter planet for much longer, for example, than me. In terms of what we're doing for society, we're making a series of contributions. One is through the Ipsos Foundation. We've already funded over 120 projects helping disadvantaged kids in terms of their education. We also support some of the most disadvantaged people on the planet who are refugees. Many of our staff around the world are either children of refugees or were actually refugees themselves. I think there's one estimate that the 700 million refugees on planet earth at the moment. So we've recruited people into the business and upskilled them and trained them where we will actually be celebrating together with the United Nations, the United Nations Day for refugees later this month. And for all our people in their local communities, everybody at Ipsos has 2 paid 4 days a year to go and volunteer to raise money for charity. We have matched giving. Again, hugely important to our teams. We want to be a responsible member of the different communities we're part of all over the planet. In terms of our suppliers and our supply chain, we've obviously got our supplier code of conduct. We are aiming to comply completely with the United Nations Global Compact. We've got a supplier engagement program. And when it comes to our panels, one thing that we can do there is make sure that our panels are representative of the different societies that we are examining as possible. And that means making sure that everybody is fully and accurately represented in those panels. It's what we do as researchers, but it does mean for different groups in different societies, you do need to make extra efforts to reach out and get the right people there. So we're doing all of those things, which together, we believe, help make a difference in terms of dealing with our own supply chain, dealing with the world as it is, and representing it better. And that's good because it's been recognized by the United Nations, but also by EcoVadis, by Moody's, by Sustainalytics, by the Carbon Disclosure product -- project as we're making progress. We're heading to the left-hand side of this chart, which is great because it's -- again, it's important to us. We are, of course, about -- we're a business here to make money, but we also care about the overall planet we live on. So I will end today's session by just saying, I think, hopefully, what we have shown you today is a reminder that Ipsos remains well positioned and unique, I think. And as I said at the beginning, increasingly distinctive from many of our major global competitors. Because of the range of what we do and the adaptability and agility that we have in our -- both in our business model, but also with our people and our teams and how we're able to deploy technology, we're able to deal with the most complex of challenges our clients have. We've embraced AI with human intelligence. We have growing productivity, both in terms of how quickly we get things done, how much an individual person can do. And of course, that all helps with profitability. And finally, as I hope I've just shown you, we are also making a difference. Thank you very much. We have time for -- Conor, yes. We have time for questions. And for those of you in this slightly warm room, we will now open the doors to improve the air conditioning. We will take questions from the room first. And then if there are people who are watching online, we will take those all via the platform. We have a platform for people watching online who want to raise questions. So Conor, off you go.

Conor O'Shea from Kepler Cheuvreux. Three questions, maybe one by one, I guess. First question, just on the political situation in France, the elections. Does that have any impact on your business for 2024? Any disruption from the unexpected election and dissolution of the parliament?

Okay. Short answer, pretty minimal, actually, probably. I think our people, who have been very busy working up till the end of the Euro elections on Sunday, were hoping for a short break, but they're now, of course, even busier working up for the next set of elections. So no, I would say on this one, probably net neutral. But it's a good question because in other countries where the government practice is bigger, then you will see some -- you might see some disruption during an election.

Great. And then second question on the U.S. business. Just a little bit more color on what's happening to the market. I think it's quite soft. Is there any examples of clients internalizing AI work because we've seen that with some other in the marketing services business of agencies and so on, that they're experimenting with these tools? Is that a factor? And also maybe comment on whether some of the competitors. I think YouGov made a big thing about their competitive positioning, improving with the acquisition of the Panel GfK. Do you see that? Or is it just a change of ownership and same kind of competitive dynamics?

Okay. So I mean, I think the issue with America, obviously, Dan has touched on some of the macro in the slides earlier. What we're seeing is just a very uneven business performance by different sections of our business, lots of growth in some, less so in others. We'll give you more detail on that at the half yearly results. I think in terms of AI at the moment, to be quite honest, there is more even in the United States, which arguably is the most technically advanced market, we are seeing as much opportunity as threat in terms of AI and the same -- along the lines of what Michel has been describing. So people, I think -- and it's -- if you roll your mind back to Google surveys, I think, which was launched in 2012, I can't remember. Google launched Google surveys, [indiscernible] they're going to do all the surveys themselves actually, our clients want to focus on what they're good at and leave market research to us. And as you've seen in those examples, they often want help or advice because, of course, we've got thousands of people across Ipsos working together collaboratively on the application of AI to consumer insight. And many -- that's going to be many more actually than a consumer insight department in all but the largest businesses on earth, to be quite honest. So that is that. [indiscernible] we put a rather derisory bid in for that asset because it's not a direct -- it's only really in one country that we operate panel that's even equivalent to that. So I think I'm going to wait and see how my friends at YouGov are actually making that acquisition work, but that is not a particular challenge to us at the moment.

And then last question, maybe following on from Michel's comments that Ipsos Facto has been rolled out to the client base beyond the employees. Are you billing that as a premium product? Or is it more keeping clients on [ side ] through this technological change?

I think it's early days. We're certainly billing them to maintain and improve our profitability, but we're not -- it's -- so it's not -- but we'll see how it evolves because if we are able to enhance it and turn it into something that is an absolute must have, then, of course, we would make use value-based pricing, Conor.

Emmanuel Matot from ODDO. Two questions for me. First, do you still plan to spend EUR 500 million to EUR 700 million for acquisition before end 2025, that was your road map. It is still the target. It seems to be a lot compared to what you have done until now. You have done many bolt-ons, but we are still waiting for mid-sized acquisitions. And second question, you are investing a lot on technologies. We spent some time about that, also on panels, et cetera. When are we going to see the leverage of those investments in your EBIT margin. Is that going to be the case starting next year? Or it will be later?

So on acquisitions, as you have seen, we have done 10 acquisitions since the beginning of 2023. As you say, Emmanuel, it's true that it's more bolt-on and small to mid-size companies, but we are quite active, actually. We are still sticking to our target EUR 500 million to EUR 700 million towards 2025, but obviously that would need bigger acquisitions on which we are working on, but as you know, on M&A, sometimes it goes faster, sometimes it goes slower. And we have constructive discussions, but it could come earlier, it could come later. Obviously, we are very selective, and we are not going to do acquisitions for the sake of acquisition, but we are still sticking to our target currently. And we have discussion, which shows that it could happen at some point, but obviously, it's too early to be more definitive on that. On the investment, it's true that we are investing a lot on, as you say, generative AI on panels, on platforms as well, which is actually one of the reasons why the different drivers that I showed, which are pointing towards more profitability in the future shows that in the short run, we are sticking to around 13% because when you look at the drivers that I explained before, you could expect more and actually we are targeting for more in the future. But we are investing, and that's the reason why we are sticking to this guidance of around 13% for the time being. And again, as Michel clearly stated, I mean, the impact of technological changes, in particular generative AI, tends to be overestimated on the short run and underestimated on the long run. So typically, it means that it will take a few years before we actually see a full impact, and it's probably early day. It's a bit too early to say -- to quantify to an extent what the impact of productivity will be. But on small things that Michel has developed before on, for instance, the way we compute our survey for our clients on reporting and translation. We can already see some productivity gains in the way we compute our data on a day-to-day basis.

This is Grégoire Hermann from Berenberg. Three questions, please. The first one would be on the chart you've shown on the customer satisfaction. We see that it's increasing since 2014. I think that's coincided as well with somehow more or less with the launch of new services. Can you please tell us what has driven this increase of customer satisfaction over the years?

When I've crunched the numbers, what we find among the people who are giving us the highest scores, I mean, it's a gradual increase. So you're talking about decimal places each year, but it's -- it was -- it's funny. We were looking at this what we were going to talk to you about today. And Didier Truchot pointed out, well, hang on a minute, go back and look at the trend. And of course, actually, it's a very good point. But when we look at the drivers and we run the regression, it's based on the competence of the teams, the fact that they're really responsive and they really understand our business. And those are the things, of course, that Eleni, who you met for the first time today, is pushing our teams to focus on. So hopefully, what we are doing is servicing our clients better. We are able, as you've seen with some of the investments we've made to deliver data faster, the cycle time on projects is going faster. But at the same time, it's not just about getting the numbers faster. It's also actually about making sure that they're getting useful insight from people they really want to work on. And one of the things that I'm personally very passionate about is making sure that our people are great people to do business with, precisely because the data you've just pointed to shows that where we get that to work, where they get that to work, where they're on, that's where we have client first. When they're on their side, they're more likely to be satisfied. They're more likely to buy again.

Maybe the second question would be on the panels. You said that you wanted to increase your proprietary panels. Can you please explain how you plan on executing that? Is it like you have targeted panels you want to acquire or?

Well, it's partly about just you can just do it organically. You can go on to Facebook and recruit more people. It's not that complicated. But what we want to do is make sure we've got the right size panel. We'll never have every single type of person we need at every single moment. Pipe smokers in Bulgaria, we probably don't have enough on our panel. We may have to go and find some with others. But overall, what we're trying to do is, first of all, make sure that our panels are large enough in all of the major markets that we're operating in because there's always churn. But we're also trying to make better use of them. And a lot of the -- maybe one of the things that we can talk about or we're very happy to do a presentation for people, if you want a longer presentation, we think we've already taken you a couple of hours. But is to explain in detail what we're doing to make better use of the people who are already signing up, and that's what we're already doing, and that is what is driving some of our productivity gains, if you see what I mean. So it's both making sure you've got the right coverage but also using the people that you've got in a more satisfying, more engaging way for them, which is about what we call access but also, yes, more -- so we can go to a smaller number. The same number of people with more surveys in some ways more effectively. And all of those things are going on simultaneously. So long answer to a short question, my apologies.

Maybe just a last one on Ipsos Facto and basically all the solutions that are AI-driven. Can you provide us with a rough estimate of how much the client-facing workloads is reduced with that?

In some of the businesses where we're deploying it, we're seeing savings in time of standard processes of 4% to 5%. And that gives the staff more time to spend more time talking to the client or arguably, they could do another project. But it's very early days. It's very varied. And we're not -- we will obviously produce figures. Some of you in the room are familiar with our new services analysis, which Ipsos Facto will be part of, but that's probably as much as I can say at the moment. We're seeing, as you've seen from Virginia's slides on the InnoExplorer, I mean, those are [ verbatim quotes ] from clients, and it's gone from astounding start to a substantial amount of money in no time, but it is very early days. And I think for all of us, and everybody in this room to be quite honest, one of the things I'm really interested in is how do we change our basic way of operating? How we train trainees to work so that Ipsos Facto isn't something you play with on a particular project where you use -- when you're using Ipsos -- when you're using InnoExplorer, but it's actually how you make your day, you get through your standard tasks faster. You've just seen Apple, of course, pitching their AI, which is meant to do that. And that's the really big -- one of the really big goals. If I can make the 10,000 or so basic analysts here, make them all 10% more productive than we're talking. But it's early days.

Marie-Line Fort from Bernstein. I've got two questions. The first one is about generative AI, sorry to come back on that point. Would you say that it will be more leverage for your margin on an upside on your revenues? That's my first question. The second question is about your strategy in the U.S. to know what would be the Mary Ann's road map, what shall we address in the first in your organization?

Okay. So on AI, I mean, I suspect that ultimately, AI will be, as with some other changes in this industry, deflationary. When we move from face-to-face interviewing to telephone interviewing, the price went down because you didn't have to send people around the streets to interview people. You could just phone them. When we move from telephone to online, the price went down, but our business grew and our margins grew. And so in the same way, I would expect actually, AI should both it will be deflationary in the cost of doing things, in the same way that you've seen Ipsos Digital, which is growing rapidly, costs much less than other forms of market research, but the margin is roughly double, our standard margin. And I think it's going to be more like that. So it gives us the opportunity to sell more to find new segments of people who perhaps wouldn't have bought from us in the past and actually to also increase margin at the same time. So I see it, and I may be wrong. But at the moment, I see it as one of those evolutions that I've seen over the last 37, 38 years that I've been in the industry. It's more like that at the moment. It's something that changes the dynamic completely. On what Mary Ann is doing in the U.S., the poor woman has only been there about a month. So she's got a number of things to be dealing with. There are a couple of our major businesses that need attention there, and she's already working on those. But it's probably -- we will bring her in to explain exactly what's going on, but when she's had a little bit more time.

[indiscernible] from Pascal Investment. Just a follow up on the price and volume quantification from GenAI. You're probably enlarging the opportunity with existing clients. So can you try to quantify from scale 0 to 10, how much business you do today? And how much basically you're enlarging the opportunity? And on the price, I get your point on the deflation. But given, as you say, you provide services which competitors can't do, you have an edge and you provide better payback for the client, why you're not able to pricing? So is there a price benefit that you don't want to say so far?

Well, we're not -- we never want to be the cheapest. That's never been our strategy. Let's wait and see. What was the first question again, sorry?

Volume benefit, if you can, I mean existing client.

Yes. So well, there's a separate discussion about volume benefits, generally with our bigger clients because actually, although as you've seen, 1/4 of our revenue comes from our 20 largest clients. When you actually look in detail, as Eleni and her team have been doing at how much of the spend of those clients we have, we still have -- we're still well below, in most cases, half their spend. Perhaps in only 1 of those 20, we're at the majority of their spend. So there's still a lot of room just with our existing clients. One of the things we always say is we don't actually need any new clients. We already have 5,000 or 6,000. We can actually do better with just doing more with some of our existing clients and developing new services and offers and AI, of course, in a competitive market makes us more attractive. And yes, I think where we can -- one of our major challenges as an industry is not just charging cost plus margin, but actually doing value-based pricing. And that's an evolving journey. And while I know that my CFO is very interested in. Okay. It's getting -- it's quite warm in here. There are no questions online, thankfully. Thank you so much for coming to see us today. We're around for cooling, drinks and refreshments, if you would like to talk to us now. Thank you.