Good afternoon, welcome to our friends, colleagues, investors, potential investors. It's great to see everybody here. Thank you for making your way over to Gentilly on a hot summer's day. We hope to have things to keep you interested and engaged over the next hour and a half. It's my pleasure to update you on our progress at the Investor Day 2023 for Ipsos. What we're going to be covering today is just some more background about the company, because I think it's important for you to understand how different Ipsos is from many of the other major competitors in this area. I'll take you through the growth plan that we announced last year at this meeting and our progress on that.
My colleague, Dan Lévy, our CFO, who joined us a year ago, will take you through the numbers and talk about the guidance. Then importantly, and one of the big changes, both inside Ipsos and outside, of course, the changes in technology. My new colleague, Michel Guidi, who joined us at Christmas, our Chief Operating Officer, who leads on IT, data collection, operations, will take you through what we're doing on generative AI and technology in general. We'll talk a bit about ESG, which remains absolutely fundamental to our DNA, then, of course, we will have time for your questions, and perhaps talk a bit more about the future. To start with, what makes Ipsos special? I thought it was important to just be clear about this, actually.
The first is that the world is a pretty complex place. Some people believe that you can understand the world by sitting at a laptop in Paris, or in London, or in New York, or in L.A., and everything becomes clear. I think the world is a little bit more complicated than that. We're living through a series of crises. You can call it a polycrisis, the pandemic, the inflation surge that follows, the energy crisis in many countries after the Putin's invasion of Ukraine. All of those things, obviously, sometimes feeding back on each other. It's a Mona Lisa world, which is what The Economist called it. Like the painting in the Louvre, depending on which way you look at it, is the economy not quite as bad as we thought? Is it getting better?
In France, 44% of people believe we're in recession. A third think things are getting better, go figure. I think that's a reflection of those mixed signals that we're seeing. We are definitely in a world with more data than ever, which we need to understand. We're definitely in a world of fake news, and generative AI, and the arrival of generative AI on top of the social media revolution of the last decade, I think massively accelerates the ability of people to get fake news. Fake news has always been around, now we're able to distribute it perfectly in real time, you see some of the results of that in some societies like the United States. We suffer as a species from collective misperceptions.
We have real issues around malicious interference in all sorts of things, but particularly in elections, and of course, geopolitical tensions are back with a vengeance. The world is, if anything, getting more complex. To understand that world, I think you need a business like Ipsos to be able to do that. Companies and the governments that we work for around the world need data that is accurate, that is relevant, that has been analyzed, and above all, is actionable, something you can do something about. Karl Marx said, you know, "Philosophers have only interpreted the world. The point is to change it." In our business, it's not enough to produce data.
Our job is to provide answers to our clients to help them with their business questions that they can actually use. That's what we are aiming to deliver here, and we'll talk about the impact that we have on our clients' businesses today. Our raison d'être is fundamentally about giving people reliable information, not just on consumers and markets, but also on societies, on individuals, parents, people. It's fundamentally about helping the world make better decisions, which means this information we are producing is valuable, and it's what we do with that information that particularly makes it valuable. We're built to make sense of complexity, and you'll see this in terms of the range of services that we provide and the way in which we provide them.
The first thing is, you know, we are digitizing not just online surveys, which is, of course, a huge part of our work, but actually a wide range of other data sources. Satellite data, mobile phone data, passively collected mobile phone data, all the clicks that you make on your laptops, on your phones, we collect those in many countries, social media data, also physical real-world data. One of the reasons why Ipsos had a good pandemic was our ability to pivot and use our infrastructure to take actual physical COVID samples from tens of thousands of people a day, move them in a cold chain across Europe, and provide that data daily to a government.
Our ability to take water samples in 27,500 locations across India, which is fundamental to one of the biggest challenges the Indian government faces, which is getting its people toilets and clean drinking water. We even use drone footage, as you will see today. We are digitizing all of that data, and we are linking that with the best human intelligence that we can find. I think that is really important. Online, online-only companies can't do that, and that is a massive advantage to us. We are aiming for the best of online, We will talk about our panel quality, the fact that we have multi-factor authentication for all our respondents in a way that I'm not aware about any other major company doing, but also our ability to reach real people around the planet.
One in three people on this planet are not accessible online. The only way you're going to speak to them is by going to talk to them somehow or somewhere, and we have some of the largest abilities to do that in terms of face-to-face interviewing, in the market. If you look at the world, you know, there's the overall level of people who are connected. It's very high in some parts of the world, but also, if you go to the largest country on earth, less than half. If you go to Sub-Saharan Africa, less than four in 10. In order to speak to those people and understand those people, these billions of people, you need the ability to be there on the ground. A quarter of the Indian population is illiterate.
They're not connected, and if you send them a questionnaire, they couldn't read it. They also speak 37 different languages. You're gonna really have to work hard to get accurate data about what is going on in the world's largest country. Because of that, our infrastructure allows us to do both offline and online interviews. Yes, the vast majority of our work is online. The ability to go the last mile, to reach everybody on the planet, is something that is hugely important to us and to many of our clients. That's why we maintain that ability. It gives us unique opportunities and unique capabilities. Even in Europe, you will observe that in countries like France and Germany, and the U.K., a proportion of interviews, of our overall interviews, are still being done by telephone or face-to-face.
If you think about it, I suspect not many people in this room are going to go home and answer a lot of questions about washing powder. You may be people with more things to do with your lives. As a result of that, you need sometimes somebody nice to come and see you, charming, to perhaps talk to you as a, as a real person. Perhaps they will persuade you to take part over the telephone. We interview millionaires in, what's it called? Monaco, sorry, to talk about their on behalf of one government, to talk about their taxations. We actually flew out to yachts to speak to them. We can reach everybody. It's really... I think it's just a reminder that everybody is accessible, but you need the ability to be able to reach them.
We are about the whole human being or, and all of human beings. People as consumers, but also as clients, as citizens, as employees, as patients, as professionals. Business-to-business research is more and more important and reaching everybody, and that includes the ability to reach both very young people, very old people, but also underheard minorities. For example, in many European cities, we may do surveys in foreign languages, not the language of the country, simply because there are now some of our populations are so diverse, and there are groups in our societies who will not be found in ordinary online surveys. You know, in Australia, we have an Aboriginal and Torres Strait Islander Research Unit. They work for large mining companies on the impact that those companies are having on particular communities in remote parts of Australia.
They work on social programs and how effective they are for the government. We are able, with the infrastructure that we have, to reach everybody, and that's something that's really important to us, and we should celebrate that alongside best-in-class digital data collection, which I want to talk about as well. We've also got expertise in all spheres of market research. We obviously have a strong focus on consumers, on helping our clients grow their brands, test their advertising, make sure their communications are working, find new innovations and new markets for them to develop, testing new products. It's an important part of our work. We also, of course, have a significant practice looking at healthcare, which I want to talk about.
A significant practice working for governments, who are great because they can always print more money if you need it, and of course, customers and employees, the CX world. That range of things is something that our clients buy into. It's really, really important as part of what Ipsos brings to the party. We also control our own value chain. Whereas some of our larger competitors have effectively outsourced all their data collection, we want to control data from designing the surveys at the start, collecting the information and understanding exactly where it comes from, and controlling the quality during that process, analyzing it, applying data analytics of our data scientists, and then, of course, working with clients on the implications and the activation of that information.
We want to be able to control our destiny across everything we do. Again, I think that is important in terms of making sure we've captured value at each part of that process. Above all, in a post-truth world, it's about us. When we stand up on television and talk about data, we know where it's come from. We know we can be confident about its quality. It's really, really important to us. We are in 90 countries, 250 offices. We are in, obviously, the major countries in Ipsos, so places like the U.S., France, China, Germany, the U.K., but also Ukraine. We're still in Kyiv, Kuwait, Zambia, Panama, Costa Rica. Why are we there?
To really understand the world, you need people on the ground from that country who can actually tell you what is really happening, and you can't find that out easily on a laptop. You might think you can, but things are very different when you're actually there. That's really important to us. Again, when you look at what we're doing, we are trying to deliver truth to decision-makers. We have the highest standards in statistically representative panels. Nature magazine in America did a peer-reviewed piece of work looking at data on vaccine take-up in the United States, and they looked at our KnowledgePanel in the United States, which you'll hear about shortly. The gold standard in digital data collection, now present in the U.K. and being rolled out across Europe. What does that do?
It's a digital panel providing real-time data, but it is designed to include all households, including those that are offline, where we actually install the internet and give people a tablet so that the little old lady in, you know, down in the France profonde somewhere can take part, and that is proven to be the most accurate type of digital panel. We have some of the best forecasting tools on the planet, which are in our innovation business. Our polling results at every election are some of the most accurate in the industry, and our clients basically trust us. It was interesting recently in Lisbon, we gathered together at our annual conference.
We brought some of our clients in. You know, they're sitting there on the stage, and they're saying, "Well, how do you judge a company like Ipsos?" One of the biggest drugs companies in the world just says, "Well, do I trust you?" There's a lot of trust involved. Where does this data come from? Do I trust you? "Would I trust you with my kids?" one said, perhaps joking a bit, but it's that point. We get nine out of 10 satisfaction after every survey we do. This is important to us. When you listen to our clients, Novartis, what do they say?
Outstanding, thought-provoking, heart-wrenching." Snap, "High-class insights that will have a strong impact on our business." Campbell's, one of our big CPG clients, "The partnerships delivered significant value to our business. We had the best possible information, guiding our decisions." Havas, "We're able to guide appropriate platform budgeting to reach audiences more effectively." As Google said last year, why do they partner with Ipsos? They partner with us to understand emotions in a way machines cannot. It matters. We've built, over the last four years, a large organization sitting over the service lines of 240 senior professionals, a little bit like relationship managers in banking, whose job is not to sell work to clients.
Their job is to really understand the business of our clients, the challenges that they face, and help them navigate the solutions that they need to make a difference and to have impact. That, again, has been part of our growth recently. The client organization making a difference. You know, this matters because we're able to be, for many of our clients, effectively a one-stop shop for people who are making decisions from all sectors. This shows you know, the healthcare sector and our competitor there, the major competitor, IQVIA. In government, often local players, but maybe Kantar Public. CPG, Nielsen, Kantar, GfK. Quite different competitors in many of these different areas that we're working in.
We are able, with our cross, service line expertise and these modular and unified offers that we provide, to be able to really give our clients what they need, and that's reflected in the money that they spend with us and how they spend it. If you look at this chart, what it's showing you on the horizontal axis is how many markets these top 20 clients of ours... There are some seats down here if you need them. How many markets our clients are actually buying services in. For many of them, it's over 40 different countries that they're buying services in. Being there matters. On the vertical axis, the number of different service lines they're buying services from.
You can see most of them buying from over 12 of our 16 different services, major service lines, which is important. It's making the connections between these different act issues, understanding context, which is something that we will be talking about at the Cannes Marketing Festival next week. How important it is in the world today to really understand your consumer's context, to manage their expectations, and to show empathy to them. Our structure allows us to help our clients do that. I think it's I just wanted to share that with you briefly at the start because it's really important to who we are as a business, and it's different to the models being pursued in other parts of the industry. In terms of the growth plan, how are we doing 12 months on?
Just an update in terms of what we said last year. Of course, an investment in technology and data analytics and generative AI has arrived on the scene. We have currently 190 different people working on different aspects of that. A huge amount of things happening. Michel Guidi will update you, but there's a lot going on. At the same time, combining the best technology with the best people, and Ipsos is nothing if it's not about its people. It is something, for those of you that know the company well, something that distinguishes us. Those are our enablers, our growth drivers, some key geographies, some key services, and of course, acquisitions. I'll update you on those in a second. We're keeping the same overall priorities.
In terms of geographies, of course, as we said last year, a lot of it is about the United States. It is the majority of the global market for market research. Over half of the money spent on market research is spent inside the United States, where probably it's about 33% of our revenue. Relative to our overall global market share, we're relatively under penetrated in the U.S., so we'll keep going there. The other major markets, even where they're mature, often have major opportunities to grow. We have good growth in France, good growth in the U.K. Obviously, China and India, we'll talk about, have huge potential. In terms of sectors, we've identified government as an area that we want to build and strengthen in, and my colleague, Trinh Tu, will talk about Public Affairs in a second.
Healthcare, important sector, I think, in the economy as a whole, you would agree. Ipsos.Digital, our SaaS platform, we'll update you on that today, obviously, our focus on tech and media clients. Yes, some of them may have retrenched recently. They thought that the curve they saw during the pandemic was going to keep on going like that. It's returned to that instead, still growing, but not at quite the same rate, but a vital part of the economy. We'll combine that with people and technology. Those are. We're moving on all of those things. Just to start this off, I'd like to welcome to the stage, Trinh Tu, who leads our Public Affairs work in the U.K., to talk about the global business to government. I've worked with Trinh for 30 years.
I don't know if we ever imagined we'd be standing here today in Paris, Trinh Tu, but life is full of surprises. Over to you.
Good afternoon, everyone. I'm really happy to be here. As Ben says, I head up our Public Affairs team in the UK, where we are the number one market leader. We work with every government department, and we provide them with insight, on social research, and also we do a lot of program evaluations. My focus this afternoon is to actually take you through our global Public Affairs offer. The first thing I'd like to say is that Ipsos Public Affairs is the global market leader, and I want to take you through this little journey this afternoon to show you some of the reasons why we are the market leader. One reason is because our market is very fragmented. It's dominated by a lot of local players in the various countries.
We are quite unique in that respect, one of the reasons why it is dominated by many local players is because the barriers to entry is very high. We are unique in having 43 Public Affairs team located in the countries that you see there, in a number of these countries, we have an incredibly long heritage, including in my country, the U.K., where we hold over 50 years' worth of trends data on the politics and on society. This is really special to our clients because it means that we have this unrivaled understanding of citizens and of society. Our teams are pretty unique as well. We're quite different to other organizations in that we have multidisciplinary teams.
That means that when our clients come to us with a problem, we can give them access to researchers, evaluators, economists, data scientists, behavioral scientists, and most important of all, we have got in-house specialists covering every public policy domain. We're incredibly well-networked as well. We work with a lot of partners, and we have these kind of very well-selected institutional partnerships with academics, with think tanks, and many of our senior people, including Ben, are visiting professors at a number of these institutions. This is really important because it really helps to amplify our impact, and that gives us the license to sell. It helps us to attract the best people to our business because we are a people business, and most important of all, it really helps with engagement from our respondents, who are the foundation of everything we do.
These are some of the things that are special and really unique about us. What does it all mean? It means that we're seeing this double-digit growth in a number of our markets, including in the U.K., where the market is really well established and where we have more than double our growth in the last decade. The market itself has got incredibly high uplift side still, and I'm going to show you why. Ben talks a lot about these kind of like challenging macro trends. These macro trends means that our clients are increasingly needing insight in order to manage their citizens' expectations. We've seen in recent years how governments have expanded during COVID, and now they are much more involved in citizens' life in terms of the cost of living.
The other thing to note about is the market is also changing in terms of the client types. We're seeing really a big expansion of NGOs and private sector foundations who are increasingly wanting to measure the impact that they're making in some of the countries where Ipsos has a very strong presence, namely Sub-Saharan Africa and India. Most of all, our clients' demand for speed during the pandemic has really stayed. They are so used to speed now. They want everything really, really fast, but they want accuracy as well, and we are able to deliver both because of the global infrastructure that we offer.
Just to bring it all to life, I want to show you a couple of examples of real-life work that we're delivering at the moment, just to kind of illustrate why the evidence we provide is so important and why we are so unique. The first example is a project that we're doing for the European Union Agency for Fundamental Rights, and it's a project to actually estimate the prevalence of different types of violence against women across eight European countries. The reason why we're doing this study, and a number of similar studies to this one, is because there is a lack of internationally comparable data, so they need to commission research to get it. It's also very, very complex.
The data that we collect has got to be robust enough for the agency to use it to actually estimate the prevalence, as if we've interviewed everybody across these eight countries. It's also incredibly complex. The subject is complex, and the audience is really, really vulnerable. Ben mentioned the importance of face-to-face earlier. It's not just about who it is that we're interviewing, it's about the topic that we're interviewing. This topic is not suitable for anything but face-to-face because it's so sensitive. We need high-quality interviewers on the ground who are especially trained to be trusted with this type of work. There are not many suppliers out there who can do it in one country, let alone in multiple countries. The second example is Ofcom.
Ofcom is a U.K. regulator, but it is dealing with issues that are common to regulators all around the world, and that is the spread of misinformation and disinformation. Ofcom is trying to experiment with different interventions to try and broaden our diets of news consumption towards more high-quality news, but they want to be able to assess this accurately. This is, this project is quite unique in that we're talking to respondents, we're doing our traditional interviews, but we're also measuring consumption passively using technology in real time. We can see what they're watching, what they're doing, in order to estimate the accuracy and the effectiveness of the different trials, and this is really important. Passive data is going to be a huge area in the future.
On this topic, it's because all of us, we underestimate, and we underreport our news consumption, because half the time, we don't even know that we're consuming news. Passive data is very important for this reason. Hopefully, those two examples illustrate why the work that we do is really, really important. In summary, it is the combination of these three things that we offer that makes us incredibly unique. The policy expertise, really deep policy expertise. We cover health, we cover education, we cover skills, we cover welfare, we cover energy and environment, we cover all of the policy areas, and the special thing is we overlay the policy expertise with methods expertise. We carry out some of the most challenging national and official statistics for government, which they use to monitor their progress.
We also do some of the really highly technical stuff in the program evaluation space, where we're using randomized controlled trial and experiment to actually isolate the impact and value for money for government. Then we round it all off with a global data collection. We are the one-stop shop for our clients' every needs. On quality, Ben mentioned, quality is incredibly important to the public sector, mainly because all the data that they do, policies have to be made on the ground, that it represents everybody in society. Can't miss out anyone. The KnowledgePanel is really special in that it is our solution to give clients that rigor, but at speed. The KnowledgePanel started in the U.S. and the U.K., and now we're rolling it out across Europe.
As Ben said, it's really special in the sense that if you are offline, we give you a tablet, we give you Wi-Fi, and then we help you to take part in it. you could only take part in the Knowledge Panel if we invite you, and that's its distinctiveness compared to panel, carried out by our competitors and also our own access panel. That way, our clients are reassured that the panel data is accurate and it's representative. In the U.S. and the U.K., it's really fueling our growth. It's really important in its own right, but it's also helping us to win some of those multi-year large contracts where the Knowledge Panel is a key selling feature, but it's not, the whole thing. Just to kind of like sum up, are we on target with our growth plans?
These are some of the reasons why we are. Policy evaluation continues to be one of our highest growth areas, and in some countries, we are now so advanced that we're expanding into new submarkets, such as health economics. Some of the global trends I shared with you earlier, that is really fueling clients' demand for some of our new adjacencies, such as policy advisory, behavioral, and citizen engagement. Governments want citizens to be part of the solution to some of these challenges, and we are delivering high growth in these areas to both our new and existing clients. The other area is the fusion of data. Clients are increasingly looking to us now to triangulate data, so its primary data collection through survey is not enough anymore.
They want to be able to link it to the incredible amount of data they hold themselves and also to the passive data that we have. We're really special and unique in that because we've got a data lab. The data lab is a trusted source where the government gives us remote access. We are the only agency in the U.K. that has this ability to tap into government data from our own offices in order to add value to the work that we do for them. Of course, we continue our expansion of the KnowledgePanel across Europe and the rest of the world. Thank you for listening.
Okay. Healthcare. This is part of our strategy in the sense that there are not many global players in this space. It is a structurally growing industry. Just to refresh you, we expect growth in this sector to exceed GDP growth consistently over the next 20 or 30 years, reflecting the massive increase in older people in Europe, for example, rising demand and access to healthcare services, but also biotechnology advances, things like gene therapy, med tech, et cetera. If you look at the amount of investment in this sector, it grows every year. We are providing a range of solutions to the industry, including our syndicated studies, led by Jackie Ilacqua. These are multi-stakeholder studies on areas like oncology, repeated over time, where you're collecting one set of data and selling it to a large number of clients.
We have a lot of work around helping our clients with drug discovery and launch their commercial strategy once they've started that process with their launch solutions and getting the product right and working out how to optimize it. Then a new area in the last 18 months or so is clinical research itself, where we will start to use our offline capabilities for real-world evidence that we will link for regulatory submission for publication. Healthcare, you know, I think is a fundamental part of who we are. Actually, there is a strong link to our work in government healthcare systems. Although we have obviously very clear Chinese walls between the two, it means that we are an agency with in-depth understanding, both of the payer side and actually the government and healthcare provider side as well.
Healthcare remains absolutely part of our strategy. In terms of our key markets, I thought it was important to update you, in particular, on America. The potential remains enormous, a $62 billion market size, very roughly. We think our addressable market there might be $46 billion. Our turnover, of course, under $1 billion. We're still a major player in the United States, which reflects how fragmented it is. There are fewer than 10 players in the United States with turnover over half a million, over $500 million. There are real opportunities there for more consolidation and to get, as I said earlier, from number six, perhaps to number three or beyond. In terms of what's been happening, we grew by 12% in the United States last year in organic terms.
This year, we've definitely seen some headwinds. You can see that in numbers coming out from across the industry. Our American business has a greater proportion of tech clients. That has affected their growth. We can talk about that if you want. I think the key point there is that the structural demand is very strong, and we have really good growth in things like business-to-business research, which I'll talk about in a second. This very close focus on client understanding, real interest in things like AI, ESG, this very differentiated offer in terms of client service that we're providing. You know, our model is not about just giving you data. Our model is about data linked to human beings. You'll see that time and time again today.
Even when we provide a SaaS , and where we actually imagine that many people might just want the data, as my colleague, Andrei Postoaca, will show you, 83% of the clients want do it with me, and that also applies to our work in the United States. Business- to- business. We're working with 16 of the largest B2B companies in the U.S. Most businesses in America are actually primarily business- to- business, interestingly. Even, you know, if you just look at something like the cloud computing market, that itself is bigger than the entire drinks market in the United States. Just understanding, cloud computing, potentially larger than beverages.
We made, of course, an acquisition earlier this year with Xperiti, a young dynamic company, focusing on digital data collection, focusing on using AI to recruit professionals from 130 different countries. It puts us at a and it gives us an advantage over traditional research firms, over tech-enabled experts, networks, and the traditional ones, because of the use of generative AI in that process. What Xperiti are doing is screening and also now validating people, because in all of these digital industries, there are always challenges of verification. Are you speaking to the people you think you are, or can anybody pretend to be anything? We are validating it, both with our consumer research, but also, of course, with our B2B research. It's very important. Over 50% growth.
It's a small business, but very exciting in the United States, China. We are number one in China, a French company that is number one in China. Business in China has gone through, of course, some pretty tough years. I was in Wuhan and Shanghai, earlier this year, just to really sort of talk to the people on the ground. The business there has been stable, despite really heavy lockdowns. I mean, we think we suffered during COVID. You have no idea what the Chinese have just been through. Things are now rebounding. We've acquired a company in Shanghai, Focus RX, to build and strengthen our healthcare offering as part of our strategy, a particular focus on a few sectors. We will keep going. We can see business accelerating.
Still, some headwinds and some of the expectations of some players in China perhaps not met, but our business there has outperformed its local competitors during the period of the pandemic. We're very optimistic about the future. We can see real opportunities in Chinese brands going global, like Huawei. B2B research, as in the United States, offers huge opportunities. There are many large exporters in China who have never, ever done market research before, and they end up working with our clients. We have to actually brief them on what it is, but the number of opportunities is enormous. Research-based advisory services, again, is a major part of our business there.
China wants to be the world's number one, not only auto market for the consumption of cars, but also manufacturer of them. We're heavily involved in the electrification of cars in China. I've already talked about our work in pharma. Then with the Chinese government's focus on right raising consumption, we see more and more opportunities in terms of CPG, consumer packaged goods, FMCG, et cetera, in that market. We've got some very clear sectors that we're focusing on and an incredibly strong team across China. In India is a market where the industry, our industry, grows by about 10%. Ahead of GDP growth in India. We are growing by more than the market average in India. We've become recently the second-largest player in the market. We will keep going.
We are number one in government, in Public Affairs, in healthcare, in automotive, and in innovation, in things like product testing, et cetera. Again, a really exciting but also very complex market, where you really have to be there on the ground to be able to deliver and to use, you know, complex hybrid solutions, mixing digital and offline, to really understand what's happening. The next steps in India is to scale up Public Affairs, advisory and healthcare. We will continue to invest in tech and the media. We are building an integrated access strategy. Looking at, we have, you know, tens of thousands of people's contact details. Can we get those people online somehow and then be able to reach them in real time?
We have a number of acquisitions that are in play at the moment to keep building our scale because there is so much growth potential in that market. There are, you know, what, hopefully, what I'm conveying is this, you know, is the parts of the world where we can see really strong growth. Actually, even in, you know, as I say, in markets like France and Britain, we are still growing well. Acquisitions, we haven't done a lot, you may have noticed. We've done, you know, but we are focusing very clearly on our major markets, some key services, but also technology and science. As some investors have said to me, it's really important you don't overpay, and valuations expectations remain high.
We are expecting them to come down in the next 18 months. We are in conversation, as you will see, with the, you know, with many, many different players. We also, number four, do have an opportunistic approach. We are about to become number one in another European market through some people who came to us and decided it was time to sell their privately owned business. Great synergies, great chemistry, hopefully, we will announce that soon. You know, a very clear focus on the major markets, on the key services, and on technologies that will enhance how we can access respondents, strengthen our platforms and our data science. We've strengthened the M&A team over the last year. We've been looking at 150 different targets in the last nine months.
As I say, we've completed Xperiti and FocusRx. We have a number of LOIs out. We have 12 serious discussions at the moment, including with, you know, businesses that are turning over hundreds of millions of dollars. We hope to have more to announce, we don't want to overpay. You know, it's a, it's an interesting period, I think. Everybody's mindset has not shifted from the free money, low inflation world that many of them have grown up in or built their businesses in, to a world where actually money really does cost money and may go on costing money for some time. It's a, it is a process of adjustment. I'd now like to hand over to Dan Lévy, who will take you through the financials.
What I'd say is the same thing that I said on February the 15th of this year, and that is that the profile of revenue acquisition this year, as we said in February, will be different than in 2022. This is much more like a 2018 or 2019 type of world, where we build through the year rather than starting with an amazing jump, and then with that slowdown that we saw in many markets at the end of last year. Dan?
Thank you, Ben. Good afternoon, ladies and gentlemen. I'll start with a reminder of our financial targets for this strategic plan. As you see, we target EUR 3 billion by 2025. This will be done obviously, with acquisitions. Ben, I just mentioned our plans on M&A, but also obviously with organic growth, we target 5%-7% organic growth over the period. We have delivered 5.6% last year, and as you know, we are targeting around 5% for this year. We are also targeting an operating margin above 13% in 2025. Beyond this profitability target, I think it's very important to understand that we have also on the top of that, structural drivers for profitability improvement.
The first one, which we have seen on the market for years now, is the shift that we observed from offline to online in terms of data collection. This is the thing that we have seen for more than 10 years now. It has clearly improved and increased during the pandemic, but we are still seeing a shift from offline to online. Currently, our own work for our network and our quantitative studies, 65% of the work we do is actually done online. This brings about, obviously, a higher margin over the longer over the long term. But it doesn't mean, of course, that we do not need any face-to-face capability.
As Ben said, I think it's quite important to recognize that 1/3 of the world's population is offline, and there are actually countries like India, Ben mentioned it, where actually most of the market research is face- to -face. The second driver for profitability in the future is what we are going to do and what we are already doing in our operations. Michel Guidi, Ipsos CEO, is going to take us through this into a bit more detail later on. Basically, what we are starting and we are trying to do and to improve, is to build an end-to-end platform from panels to the results reporting. We also want to increase our panels' internalization.
Currently, the majority of the panels data that we use are internal panels, are Ipsos panels, but we want to increase this share in the future for many reasons. The first one is that it's more efficient and brings about productivity gains, but also because in a world where generative AI is going to develop, we are pretty much convinced that it is the businesses which are going to produce their own data that will be successful in the future. Obviously, we have a lot of opportunities in terms of productivity gains with generative AI, in the way we code and we script the questionnaire for our surveys, in the way we report the results of our studies to the clients. Think about translation, for instance. We are at Ipsos in 90 countries.
We do a lot of multi-country studies. We have to translate a lot of surveys and questionnaires from one language to another. This is typically something which is quite easily done by the new generative AI model. That would bring, for sure, for the future, some productivity gains. The third driver in terms of profitability improvements is our business mix. As we grow quicker on more profitable businesses like Ipsos.Digital, which is growing by 35%, on the first quarter, like the U.S., in the U.S., typically, the margin tend to be higher than the rest of the world. Our advisory business, our data analytic business, these are all businesses that we want to grow in and which have a higher margin than the average of the group.
We are also a diverse and resilient business model, as you can see on this chart. On the left, you have the breakdown by geography, and you can see that we have a nice geographical breakdown, which is good because it means that we can more easily accommodate a macroeconomic shock. When you look at the current macro outlook, for instance, it is more the advanced countries which are struggling, but other parts of the world, like Latin America, like a lot of countries in Asia, for instance, India, typically are doing extremely well. We are also well-balanced in terms of sectors. CPG, which is our historical sector, still account for 24%.
A quarter of our revenue, but it used to be 1/3 , five or six years ago. Since then, we have grown in other sectors like healthcare, like public sector, and it is actually good for resilience as well, because these sectors, healthcare and public sector, are sector which tend to be less sensitive to the cyclicality of the microeconomy outlook. Finally, we are well-balanced in terms of client portfolio. As you can see on the chart, our top 20 clients represent only 1/4 of our revenue. We are a very strong cash-generative business, and we have a sound financial profile. We delivered last year, EUR 214 million free cash flow.
We have hardly no debt, so leverage, which is now close to zero, which is good news in many ways. The first one, obviously, is that it means that we have all the financial means to finance our strategic plan, to make the acquisitions and the M&A plan that Ben detailed just before, to invest in our panels, to invest in our platform, technological platform. Also it means that we are less sensitive as compared to many of our competitors, to the increase in interest rates because we have no debt, which is quite important when we see how some competitors struggle with indebtedness. In terms of liquidity, we are also in a very good position.
We have nearly EUR 500 million of undrawn credit line with a maturity above one year. Just a quick reminder also on our cash allocation plan during the strategic period. As you can see on the bottom left, we forecast to deliver around EUR 900 million of free cash flow. We have already delivered EUR 214 million last year. In terms of usage, you have the second column on the right. We want to buy between EUR 500 million-EUR 700 million of enterprise value of acquisition. Again, Ben detailed the plans. We want to invest in our panels and in our technological platform, and obviously, we want also to reward our shareholders. We do this in two ways.
First one, obviously, is paying dividends, we want to pay between 25% and 30% of adjusted EPS in terms of dividend over the period. The second way is our new share buyback program that we have on the top of the usual share buyback program that we had in the past to finance the free share programs for Ipsos employees. In terms of return to investors, we have increased the EPS between 2019 and 2022, from EUR 2.8 to EUR 5.12. We have increased the dividend from EUR 0.88 in 2018 to EUR 1.35 in 2022. As I said, on the top of that, we have started a new share buyback program at the end of last year.
We did EUR 10 million at the end of last year, and this year, we have already done EUR 27 million so far. A quick update maybe on the current trading on 2023. First of all, we confirm our guidance for this year, which is an organic growth of around 5% and on operating margin of around 13%. Of course, there are macro uncertainties. Ben mentioned some of them. Inflation has started to increase in many countries, but inflation in the past has also meant that most of the central bank in the world have tightened their monetary policy, and this obviously is waiting on the macro outlook. We have seen a clear rebound in China numbers at Ipsos after the end of the zero- COVID policy at the beginning of the year.
Now some economists are questioning the strength of the recovery of the Chinese economy. Some other countries, like Europe, for instance, are doing well, given the macroeconomic context, and particularly given the fact that, as you know, the Eurozone is now in a technical recession, and particularly Germany, there is a war in Ukraine. Given this context, our European business is particularly resilient, and we have other place in the world, like, as I said, Latin America and a lot of countries in Asia, which are doing very well with double-digit growth. As you know, and as Ben has mentioned, the slowdown that we have seen at the end of last year and at the beginning of this year is mainly coming from the big tech clients in the United States.
These clients have been going through a restructuring period since last summer. As you know, as we communicated, for the Q1 results, we lost EUR 7 million in Q1 on this big tech client, as compared to last year. With these big tech clients, it's important to understand, I think, that we do mainly traditional market research. We do brand tracking, we do product testing, we do mystery shopping. This is clearly demand that will come back at some point, and that has already started to come back for some of them. On the top of that, we have new opportunities with these clients, mainly around generative AI.
We have, a few weeks ago, signed a contract with one of these big client to help this client understand the context around generative AI. We are pretty sure that structurally, the demand will come back. The current landscape is quite diverse, so for some actors it has already rebound. For some other actors, the demand remain still low. In terms of the order book, as you can see on the chart, we have currently 66% at the end of May, of our revenue for 2023, which is visible, which is in line with our guidance of around 5% growth, and which is in line with what we observed in the past, as you can see on the chart.
Obviously, when you look at the numbers, 2020 was an exceptional year because of the COVID crisis. 2022 was the other way around on exceptional years, because you might remember that we did a very strong first quarter in 2022, which explain the 68% that you see on the chart. We always said, and Ben has just reminded us, that the pattern for 2023 was going to be the exact opposite of 2022, which is a slower growth at the beginning of 2023, and then an acceleration of the growth during the course of the year. This is actually exactly what we are seeing in the numbers. We said in the Q1 results that we had an acceleration in our order book during the course of Q1.
February was better than January, March was better than February. When we look at the April and May number, we can still see this acceleration in the order book. At some point, this acceleration in the order book obviously translate into an acceleration into revenue, with generally a few, a few months lag. We are expecting, obviously, Q2 revenue to be significantly better than Q1 revenue, with Q2 revenue being in the growth, being in the positive territory. We are probably not going to close the gap between order book and revenue during Q2, which means that revenue is going to accelerate in Q3 and Q4. Thank you very much.
I now leave the floor to my colleague, Michel Guidi, who is going to take us through Ipsos' technology roadmap and speak about generative AI.
Thank you very much, Dan. Hello, everyone. My name is Michel Guidi. I'm the new Chief Operating Officer for Ipsos, having joined the company six months ago. I come to Ipsos with 23 years of experience in the insights industry and expertise in data, operations, technology, and innovation. My role is to execute our technology roadmap and to accelerate the digital transformation of Ipsos. For that, we are cranking on building our new generation, end-to-end proprietary platform, as was presented to you last year. We are integrating the access to people, the data collection, and data analysis, reporting, and storytelling. We choose to control our destiny when it comes to data production along the entire value chain.
Our common tech stack is an enabler for us to connect our different systems, and increasingly, all of our data streams, and also to accelerate our development of new product cycle. One great illustration of that is Ipsos.Digital, our software and service platform, for which we're now going to listen to my colleague, Andrei Postoaca, who heads Ipsos.Digital and Ipsos Synthesio globally, and who's going to tell us more about the latest and exciting developments around the Ipsos.Digital platform. Andrei, over to you.
Thank you very much for the opportunity to present Ipsos.Digital. This is Ipsos' self-service research platform that we have built. We conceive it as an umbrella for a multitude of systems that we have been able to interconnect. We, in total, we have 11 systems that we have been able to interconnect, full end-to-end connection, either directly or through APIs. We're looking at availability in 48 markets. What we have seen for this year, especially, more of a localization. We have now the platform in Chinese and Japanese, and soon launching it in Spanish and Korean. We have seen a very strong adoption in those markets. Once we had the local language, once people can do it in their own language, it has been a positive, a very positive impact.
To give you a little bit of a flavor in terms of what is this platform, how does it look like, I encourage you to go in and sign up. It's very simple, 30-second process. One of the uniqueness that is worth speaking about is also the fact that we do not have any subscription fee. You can utilize the platform, and we charge only for the projects that are actually launched, and exactly for what you get. We started with the innovation and the creative excellence service line and built the main products within those service lines. Creative|Spark, which is a copy testing, so copy testing product, Duel, InnoTest, so concept testing products, and we also build FastFacts, which is the more ad hoc, build your own study as you find relevant.
To give you a flavor, you go in, you decide which is your field market, number of tests or number of interviews, what are the target groups? Here, everything is connected to our sampling algorithms, our panelist databases. All the time, you can see on the right side, what is the price that I'm expecting for this study? Depending on what kind of ad you're testing or what kind of research you're doing, you can upload your stimuli material, your ads, in this case. You can upload brands, you can decide the logos, tailor-make the research exactly according to your needs. This is done either by the client directly or in collaboration with our researchers. The client can always choose to do as much or as little as they like.
The complexity of the questionnaire is at a level that is quite high, so you can do quite a lot of things with the questionnaire. We try to keep it to the level that it can still be done by a researcher or a client. That is the overall logic behind it. Automatically, you get your results. You can play around with the data, you can analyze the data from many perspectives. You see here verbatims in this case, automatically creating the PowerPoints, SPSS files, Excel files, and much more. Everything fully interconnected, when the last interview is done, you get everything reported, which makes the life of the researcher and client much easier and also much faster.
If we look at some metrics, 11% of our clients are now on the platform, with thousands and thousands of users using the platform continuously, both internal and external. An interesting number is also that 83% of the studies are done together, so our clients, together with our researchers. To some extent, this is the power of Ipsos in a tech world, is the fact that we are able to be local. We are able to give you the right researchers for your necessity, for your kind of project. Everybody working in the same platform, in the same interface. Revenue-wise, we have grown nicely for the last couple of years.
We started it a couple of years ago, we have grown from EUR 35 million to EUR 65 million, now EUR 100 million, with a very strong profitability. We're looking at double the profitability of Ipsos that we can have on the platform. Looking forward, what are we focusing on right now? Of course, country expansions, language expansions, but also looking beyond the innovation and creative excellence service line and building products together with more of our service lines. Market strategy and understanding is one of them that we are looking in building products and enabling them to move some of the business on the platform. Customer experience and employee experience, the same thing there. A very important part, to be able to upload your own sample and work and do the study on the platform.
Automotive, mobility, and development, and Public Affairs. Step by step, looking at what are the products, what are the businesses that we can standardize, and offer them the opportunity of the full end-to-end automation. That being said, thank you very much for your time, and I encourage you to go on Ipsos.Digital and try the platform. Thank you.
Thank you very much, Andrei. This is a great illustration of the power of the platform we're building and the ability it gives us to do more, to connect more, and to launch new products, again, on Ipsos.Digital and beyond this. One of the fundamentals of our integrated platform is our access to people. It's one of our pillars, obviously, at Ipsos, and I'm going to focus here on our online access to people, our online panels. We are continuing to invest into our the growth of our online panels. You can see three numbers on screen. The first one talks about panel size. We've grown 18% since a year ago, and we'll continue to do that. There is an acceleration of our investment into our own data assets, online data assets, in the second half of 2023, as I mentioned.
The second metric is about panel health. We recruit people that are better quality, quote, unquote, "and stick around longer." 84% monthly retention, as opposed to 80% last year. It was a good number. It's a better number this year. We've been investing and will continue to invest in the retention of our people, because obviously, the best respondents is, like the best clients, is the one you keep. The third metric is also very important to us, panel quality. 99.66% of the projects we have delivered to clients were flawless from a sample quality standpoint. We'd love it to be 100%, that world doesn't exist. We're trending better, last year was 99.46%. It's very important because people come to Ipsos expecting that sort of quality level.
Ben alluded earlier in his presentation to multi-factor authentication. This is one of the key things we did in the past 12 months that explains part of the panel health and panel quality improvements. Multi-factor authentication is well known. It's used in many different online systems and applications today. It is not very used in the online panel world, and as Ben said, to our knowledge, at the registration level, we think we're the only ones to apply it at scale around the world. It was an investment that was very much worthwhile and paid dividends. This is important because we want to control our data and know where it originates, as Ben told us earlier. We've also heard Dan talk about the opportunity that bigger and better online panels represent for Ipsos in terms of drivers or profitability improvement.
It really matters to us from both a client value standpoint and also from a financial performance standpoint. It's also, of course, extremely important because that's the base of the next topic, which I know is a popular one, generative AI. The more data, the better in a generative AI world, of course. On generative AI, we have received several questions from the investor community, really revolving around, is generative AI a friend or a foe to Ipsos? Should we be excited or worried about generative AI associated to the Ipsos name or brand name? Let's rewind the presentation a little bit, if you don't mind. We've heard Ben articulate our raison d'être, and then we've also heard him talk about the type of work we do for our clients and how we do it. It's almost all there.
We produce data that is unique, digitizing all data sources. I'll come back on that because it's very fundamental in a few slides. We do that applying the best human intelligence, and that's another differentiator for us in a generative AI world. We do so for what? We do so to deliver accurate, relevant, analyzed, and actionable insights for impact. What we do is very relevant, and we see very little risk towards generative AI. We actually see a lot of relevance and opportunities. We are very excited. Selfishly and half-jokingly, I remember interviewing with Ben and Laurence and DJ, several months ago, and we were talking about artificial intelligence, and at the time, natural language processing.
Generative AI was not yet a coined term, and the advent of generative AI is making my job a little bit easier in terms of executing and accelerating the roadmap and the transformation. An opportunity, how so? First and foremost, it is an opportunity for speed. We can use generative AI, and I'll show you examples of that in a moment, on how we can increase the speed of our delivery to our clients, something that all of our clients will be very happy about. Of course, we will use it for productivity savings, too. There are parts of the tasks that are currently done by humans that can be done partly by generative AI, and that will help. We also see opportunities to create new revenue streams, and Dan and Ben mentioned a few already, and I'll come back on a few more.
We also see an opportunity with generative AI to differentiate Ipsos, and to do so by our unique approach towards responsible AI. It is a great power, and as with every great power comes great responsibility. People come to Ipsos today to be sure and to have truth and actionable insights. They trust us with the data. We've all heard about hallucinations and sometimes painfully wrong, outputs from generative AI today, including in the legal space, where it can sometimes invent laws that don't exist. That's what you come to Ipsos for, to be sure. We will lead in, again, responsible AI.
You will find on the tables in Paris, and so those of you connecting remotely, you'll find it on our ipsos.com website, one of the two points of view that we've made available to you, called Exploring the Changing AI Landscape, where amongst other things, we talk about the evaluation framework we apply to generative AI. We published it in April, around three key themes for us: truth, beauty, and justice. We talk about accuracy, explainability, and data ethics. Those three are very important, and that's a framework we'll be applying, and we are applying when using generative AI today. Again, it will be a differentiator for Ipsos. We haven't started with AI, with generative AI. Ipsos has a head start in artificial intelligence altogether, and I thought about what could be the best way for me to illustrate that to you.
I could have used the 10,000 data science causal structure models that we've delivered to our clients in 2022. My colleagues told me this was too complicated. I picked another example, among a long list, amongst a long list, the World Bank Awards. I will not be talking today about the social impact that this has had, because Lauren is going to do that in a few slides, and she does a much better job than I could. I'm gonna talk to you about the technical or technological prowess that is with this.
We work for the World Bank, and what we do for them is combine machine learning and AI models with multiple data streams, satellite footage, drone footage, anonymized mobile data, media analysis, including social media analysis, survey data, and other types of data, and we combine all that and apply, of course, our domain expertise to produce meaningful, impactful insights for the World Bank. Not only do they like what we do, as shown by the fact that they continue to work with us, they gave us awards since 2017 and all the way to this year in 2023, for the quality of the work we've done, again, blending AI and machine learning with domain expertise and different data sets. That's one really good and powerful example.
Another one of our pre-generative AI life is the fact that before it became a thing, we were already applying artificial intelligence in eight of our service lines. We were doing so for internal purposes, such as process automation or open-ended question management. Also, we were using AI and machine learning externally to build AI-powered offers, blending machine learning with the knowledge of our researchers. A great illustration of that is our innovation service line, which in 2022, has launched several predictive applications to predict innovation success, to predict product trends, and also predict image or pack or imagery success.
I'm gonna show you now a video of one of them, one of the offers called InnoPredict, which is a global and scalable offer that blends machine learning with consumer reactions to come up with better predictive results. It's already available on the market and being used by clients of Ipsos today. Let's play the video. So this is just one example of an offer already out there in the market, leveraging predictive AI. With generative AI, we're now working on implementing generative AI to further accelerate the delivery of that product, predicting, again, future success for innovations. In a generative AI world or era, there are two key ingredients that will make you win. Good news, there are two big advantages of Ipsos. The first one I alluded to in my introduction, it's about the data. We, Ipsos, produce unique, high-quality data.
We produce it in many different ways, as Ben said, online and offline, in a qualitative and a quantitative manner, in over 100 countries. We store that information to make it available to our clients. Think about 50 years of social and political trends, or 150,000 innovation use cases, or 35,000 ads that we've analyzed, or the 12 years of our Global Monthly Advisor Study, or our Global Trends Study, our Generation Study, all the data we own on our panels, on all the data. Okay, I'll stop. The list is too long for me to go through. We own so much data that is unique. Our data sets are a unique capability and value add to language models that use publicly available information. That's it. We go beyond online. We go specific.
We collect that data real-time for specific answers. It's curated and analyzed. It's a treasure trove for language models. The second ingredient that you need to win in generative AI is expertise, beyond having access to the models themselves, of course. It is humans that operate and need to evaluate the models. Good news, we have the best human intelligence at Ipsos, and our researchers across the world are experts at asking the right questions. The art of the question is becoming even more important than it was in a generative AI world, because it's all about querying in an iterative way with the artificial intelligence model of your choice.
There's a second point of view that we produced two days ago, which is available on the tables in Paris, and again, on our ipsos.com website, where you will find a glimpse of the training that we're starting to have with our researchers on best practices around prompt engineering for a generative AI world. It's not gonna be easy because they are already experts at the art of the question. It's critical to make the most of the models. It's a differentiator there, too. We have data scientists, statisticians, and data modelers, which are adept at swimming in data and very compatible with training models. That matters, too. We at Ipsos have decided to use...
To sign agreements with the tech leaders of the space to leverage their models, and also decided to have a model-agnostic approach in order to adapt the models to the client's needs precisely. Not all models will work the same, in the same situation, and we'll be able to switch from model to model in order to provide the right outputs. We won't stop there. We will also have our own extensions, and we'll train models with our own data to make it available to clients. We have the people that know how to do that very well, our data scientists, statisticians, and data modelers. We also have operational data specialists to benchmark and validate the data, and that's also extremely important. Why? Because what we do at Ipsos matters. This is a serious business.
We impact the world in the public sector and in the private sector. Again, back to the hallucinations and the craziness that can come out of models when they're not controlled. We have people whose job it is to validate data and to evaluate models. We firmly believe that no data should be exposed to the outside, being published or exposed to clients, of course, without being checked by a human first, and we have the humans to do that in our data specialists. All that is really meant to create insights, activation, and impacts for our clients. Again, we have the teams and the models. Beyond the theory, again, the two big advantages that we have, what are we doing now? We have a strong community of practice.
Ben mentioned the number before, about 190 people, and growing, that are working together to come up with ideas. "Okay, we could help with that part of the value chain, the production chain. What about we get on it?" We've currently identified 60 applications that we'd like to test and put in production, maybe someday. 12 of them we think are low-hanging fruits. The three main topics are translation, for now, analytics, and production. Let me take two examples. The first one will be the translation of open ends. I'm not going to dwell on it because Jacquie, my colleague, Jacquie Matthews, is going to talk about it more in a subsequent segment of the presentation. There's huge potential there. Dan alluded to that before as well. Let me take another one in the analytics row. Automated data processing.
DP stands for data processing. We do a lot of work cleaning data and processing data before it can be used by our clients. A lot of that is manual work. It needs human intelligence to use tools that are fairly rudimentary. They're okay, but they're not very different from the tools that we or anybody in the industry, you know, has been using for the past 15 years. The new models behind generative AI are going to enable us to go a lot faster and automate parts of the process, still with validation from a human, but to go a lot faster on that. We're very excited about those use cases, and again, we'll go down the list and prioritize them as we go. I'll talk to you briefly about the roadmap that we're following at the moment.
Let's get closer to the frontline, and to do that, I'm very happy to introduce you to Jacquie Matthews, who leads our qualitative research service line, and is going to talk to us about practically, how is generative AI impacting her service line within Ipsos in 2023. Jacquie, over to you.
Great. Thank you, Michel. Hello, everyone. I am Jacquie Matthews, the Global Service Line Leader for Ipsos UU, which is a qualitative research division of Ipsos and the largest qualitative research company in the world. The HI on the screen there stands for human interaction. Everything we do at Ipsos UU is to do with interacting with people and with the human. Think about, in context, ethnography, spending six-eight hours with a consumer, a patient, people in society, in real life, and filming all of that. Think about interacting a researcher with a consumer digitally across a screen, as I'm doing right now with you, and videotaping that conversation. Think about the interaction of in-person focus groups, and again, we capture all that via video for later viewing.
Think about interactions we have with all of our research respondents as we give them homework to do prior to research, so that we can get a glimpse into their world, asking them to upload mobile videos and mobile photos. Think about our online community business. This is both qualitative and quantitative methodologies, but this is a business that has, on average, 1,000 online community members, all talking on the same topic over the course of, on average, 52 weeks, so a whole year. From a number perspective, we're talking about 160,000 people that we talk to in-depth each year around the world. Each of those people, we talk to for at least one hour. In addition, we have 125,000 online community members interacting with each other and with Ipsos.
It yields 100,000 hours of video recording every year. That is a lot of unstructured data in video format, in picture format, and all that text from people writing digitally in the online communities. This area was just waiting for something like AI to come along to help us to translate when we have a multi-market, multi-language study, to help us to transcribe video to text so we can analyze it, and to help us to theme all of the content that we have. When somebody says to me, "Do you think generative AI will transform your business?" Absolutely. AI will be an enabler for us, and I'm going to share with you three different areas in which it will be an enabler. The first area is all about speed.
It will be an enabler in getting speed of insights to our clients, and then our clients being able to get the speed of their ideas into the market. What I'm going to do now is I'm going to start off by showing you a short video. This video has no sound, so I am going to talk over the video. It's two different communities that we're allowed to share. The first one is an older age group of people who were quite nervous about in-person grocery shopping at the beginning of COVID, and our retail client wanted to help solve their problem. The next community is a younger community called the Tech Insiders, and what you're seeing here is just answers to our questions, our activities. They actually comment on each other's comments.
Sometimes we ask them to upload pictures of their devices, in this case, so we can get a glimpse into what their environment and what their real life looks like. Sometimes we ask them to upload a video. Here we ask them to upload a video about breaking up with one of their devices that they don't love anymore. We even do online focus groups in the middle of an online community. Why I showed you this one-minute video is because I wanted to show you in just one minute time, look at how much data we capture. Focus group data via video, we capture uploading videos, uploading pictures, and all of that text in the online communities.
AI is going to be such an efficiency help for us, again, in translating the multi-market, multi-language studies into one language, in transcribing things like all of those videos and the focus groups into text, so I can then analyze them, and in theming all of that content. Of course, the role of Ipsos will be to ensure that the data we get from AI is indeed accurate. That will be a new role that we have. The first enabler is to do with speed and efficiency, improve productivity for us. The second enabler is to do with powerful collaboration, providing inspiration, and also, you know, speed to market of new ideas, new early-stage products.
This is a real study that I'd like to share with you that we just did two weeks ago, where we actually had generative AI as a member of our research team. The client here asked us to ideate some new ideas and concepts for their category. As a research design, we chose to have in-person workshops. Think about having three different workshop tables, similar to the one that we're seeing here on the screen. At each table, there was two Ipsos researchers. There were four clients with different functions, like research and development or perhaps marketing. There was a super consumer for the category, and then there was generative AI.
What we did with generative AI, as ideas started to bubble up, we would ask questions like: "Is this idea present anywhere around the world, and can you bring it to us?" It helped us to be inspired, it helped us to articulate things better, it helped us to collaborate, and it also helped, again, from a speed to ideas and to concepts. We have speed as an enabler, we have powerful collaboration and inspiration as an enabler, and the third example is that it will allow us to open new revenue stream, in this case, advisory services. We have a client who commissioned a study with us, it was called an empathy program, where they were having 300 of their clients, of the client itself, talking digitally over a screen to 300 of their consumers.
Here you can see the client speaking with their consumers digitally, 300 different clients, 300 different consumers, all talking for one hour. We videotaped the whole thing. Our role at Ipsos was to connect our clients with the consumers, to help them understand what they heard in their conversations, and to help them know, so what, and what now? How do we action this? We did that research project. Done. We filed away our 300 hours of video. The client has recently come back and said, "Can we please revisit those 300 hours? Because we have some different questions we think those 300 hours can now answer." To do that today, it would take 300 man or woman hours to rewatch the videos. Too much time, too much expense for the client.
AI can be a real efficiency play here. It will open new revenue streams for us, and again, be very efficient in being able to source out those new questions. Our role then becomes storytellers and finding those very sticky insights that make the client sit up and take action. Three enablers, speed, speed, and efficiency, collaboration, inspiration, and opening new revenue streams. We have had, just recently, a client come to us and ask for their transcripts for a recent project because they want to put it through an AI engine to see what is the report that AI would deliver versus the report that Ipsos actually delivered. Of course, clients are looking for efficiency in spend and to do it themselves whenever it makes sense.
We have an amazing relationship with this client. They're actually having us do this analysis together to see the output of what Ipsos has versus the output of what AI might generate. What AI needs is AI needs to be prompted with the right questions. To do an initial prompt, any one of us will be able to do that, and there's a whole industry now called prompt engineering. An initial prompt will be a very basic requirement of all of us, such as knowing PowerPoint or Excel. Our clients will be able to do initial prompts as well. Simply saying to generative AI, "Here's some transcripts, can you tell me some insights?" Will not be satisfactory to anyone. That's where Ipsos steps in, because we are, as Michel said, experts at the art of the question.
Where we go further than what AI does is in the well-crafted, non-obvious, iterative questions. The specialty here is not in the initial prompt or the initial question that I ask AI, it is in the subsequent iterative questions that we keep getting deeper and deeper, and deeper with AI. It's the non-obvious questions that have tensions in them, that yield unexpected results. As Ipsos researchers are conversing with AI, prompting and asking the right questions, we're also going to be leveraging our proprietary and proven framework to layer on top for analysis and interpretation. How are we different than a tech company that might have an AI engine? We know how to write and ask the right questions, and we also know how to layer our proven framework on top of the data.
Bottom line, at the end of the day, we need to have an expert human, and we need to have AI in order for clients to be sure in the world today. Michel, back over to you.
Thank you so much, Jacquie. What we've tried to do here is to start with the question we got from the investor community, and starting from a more theoretical stance, then really get closer to our client work and the frontline where it really happens. I hope you found, like me, this example very interesting. We could have picked other service lines, frankly, to illustrate the same, the way generative AI can help with everything we add to it and how it helps us, again, gain speed and productivity and differentiate. What's next for Ipsos in the realm of generative AI? First, we are continuing to develop a unique approach towards responsible AI. It's a very powerful technology, no doubt. It's a bit of a black box, too.
You're throwing things in there. You get an output. Today, it's very hard to, you know if it's accurate or explainable, you know, what are the sources, and is it, is it true? By the way, is it sometimes legal, or is it right from an algorithmic fairness standpoint or data rights or data privacy? We need to lead because our clients ask us to lead with them on the topic. We just published internally our AI policy, which is going to be available to clients upon request, of course, and we're gonna be training our staff on it.
We've built a team of AI review committee for internal questions and the ones from our clients as well on, again, good practice of generative AI, because there is a lot of interest from our clients to use it, but they don't know how to use it best. Leading with, again, our technology leadership, but also our responsible AI. Two, we'll continue to work with the generative AI leaders on our enterprise agreements. We have signed MSAs to make sure that the data that we use and the extensions we leverage to train with our own data is a safe space for ourselves, Ipsos, and for our clients. We'll continue. We have access to their beta programs. We'll continue down that road. I suspect more and more interest, by the way, because of the unique data we bring to the table.
We'll continue down the path of partnering with the generative AI leaders on the technology side. Third, I mentioned briefly, our strong community of practice. It keeps growing virally within Ipsos. It's 190 strong. These are the people that come up with the ideas from the front line, a grassroots approach to, "Okay, I'm doing this today. I think it can go faster if we use generative AI. Can I? Is there a problem with me doing that? Am I allowed to do it? Does it fit with our responsible AI?" We are approaching that today, and we'll continue to do so. I mentioned 12 low-hanging fruits amongst a current place of 60 applications. We need to orchestrate all that.
I'm not sure we'll do all 60, by the way, probably not before the end of the year, but we are looking at them every step of the way. Last but not least, generative AI is not only an internal opportunity for Ipsos, as we heard from Jacquie, it's also an opportunity for us to sell to our clients. Why? The world is changing. You know, we are masters at making sense of a complex world, as Ben mentioned before. Well, generative AI certainly makes the world more interesting, but not less complex. As a company or an organization, whether you're in, whether you're in the public sector or the private sector, all of our clients have questions about the technology, how it changes the way their constituents, their consumers, their employees use it to do what?
What impact does it have on path to purchase, on the way people search for information? There's big election cycles next year in several major countries. How is generative AI going to impact those elections? There are questions everywhere from clients around the topic itself, and all of these, and Dan mentioned one recent example with a tech leader company. Those are as many business opportunities because people come to Ipsos for that reason today, to try and make sense of a complex world with actionable insights. I'm briefly showing you this. I'm not going to dwell into it. It's way too crowded a slide, I know. The idea is we entered generative AI with momentum because AI is not new to us. We are very good at it.
We have a clear roadmap and are very excited and equally aware of the responsibility that comes with it. We're very excited about generative AI and the ability to create opportunities for Ipsos and for our clients, particularly around speed. We have a clear roadmap. We have a team working on it. We have a clear leader internally as well, and we're pursuing with speed. It's early days. It's early days, as reflected by your LinkedIn page or you're reading newspapers, where every day there's a new, like, top 20 applications in AI that you must be using. They're never the same. It's a buoyant space at the moment. It's fluid. It's fluid also, frankly, on the compliance side of things, which makes it interesting. It's information. It's not mature yet. It's early days. We're on it.
We have the right teams focused on the right things. We'll keep you updated as we go. This is my last slide before I hand over to my colleague, Lauren, for our ESG segment, which is going to be fascinating. These are principles of how we operate at Ipsos. They are not new. We didn't create them after generative AI. They were there long before. Security, simplicity, speed, and substance. We won't change them. We're gonna be using generative AI across the board on all of our service lines, and the biggest impact we'll have, as you hopefully understood, is on the third one, speed. It's a big lever for us to more hastily, with more diligence, without compromising on security, simplicity, and substance.
If you're in Paris, at the back of the room, you see posters on the wall, which you see small images as well here, of our marketing campaigns about Be Sure. You come to Ipsos because you need to be sure. You need to be sure, why? To move faster, to act smarter, to go further. That is still very true, if not truer, in a generative AI era, and we're here for our clients for that. Thank you very much. I'll now pass the microphone virtually to my colleague, Lauren, for our ESG section. Thank you.
Thank you, Michel, and hello, everyone. I'm Lauren Demar, Chief Sustainability Officer and Global Head of ESG. Ipsos is evolving in many ways, as you've just heard, and of course, the world is also evolving to address environmental and social issues, which is opening up new commercial opportunities for us, while also it's impacting our actions as a company. This is the world of ESG, which is a key element of our 2025 growth plan. We see ESG as a driver of value creation moving forward, and as Ben mentioned earlier, Ipsos is raised on deck. It's to deliver reliable information so that the world can make better decisions. And this puts us in a unique position to bring true understanding to a world where ESG challenges are more and more present.
Our mission for ESG is to give businesses, governments, and public bodies the confidence to take the right decisions for the benefit of people and the planet. We're increasingly seeing ESG as an important commercial opportunity. There's a growing list of client needs and requests around environmental sustainability and social issues that we're getting from clients. We are uniquely placed to help our clients to address their ESG business questions and to help improve society in general. Our ability to understand the whole human, as Ben discussed earlier, in this case, not just as a consumer, but importantly as a citizen, is key. Our expertise in both the public and the private sectors, and of course, the breadth of our offer, enables us to help clients with a broad range of their ESG business challenges.
Importantly, ESG is uniquely Ipsos is uniquely placed to inform citizens, businesses, and the media about how society is evolving on ESG. In a world where fake news and unreliable information are more and more present, Ipsos contributes to the betterment of society because we're providing accurate data and reliable information on which to base decisions. To that end, we share our data and our public opinion generously on many different ESG topics. Of course, ESG is relevant to multiple stakeholders, including current and prospective employees, to our clients, and of course, to our investors. Our achievements have been recognized by leading rating agencies and the press, including by EcoVadis, the world's most trusted provider of business sustainability ratings, who put us in the top 5% of the more than 100,000 organizations that they evaluate on ESG.
Ipsos has been measuring and reducing our own carbon emissions for 10 years. We are now accelerating our carbon commitment. We've now formalized our commitment to the SBTi, which is the Science Based Targets initiative. As you may know, SBTi is a nonprofit organization that signs off on companies' carbon reduction objectives in order to ensure that they're compatible with the targets of the Paris Agreement, whose goal is to ensure that the rise in global temperatures is ideally kept below 1.5 degrees Celsius compared to pre-industrial times. We're already on our way, given that we've already cut our own carbon emissions per employee in half since 2019, across Scopes 1 and 2, and the business travel portion of Scope 3 emissions.
We're also implementing new policies to ensure that we sustain and accelerate these reductions, and we're also initiating a supplier engagement program around their ESG commitments, so that we're tackling head-on a reduction in our full Scope 3 or indirect emissions. Our clients are increasingly asking us to do so, and we will increasingly demand the same of our own suppliers. Now, on the S side of ESG, we are caring for our people, the communities in which we do business, and society at large. We have made great strides on improving gender diversity at the senior levels of Ipsos, with 46% of women now among our leading executives. This is up from 42% in 2021, and women make up 54% of our Board of Directors.
We've also joined the United Nations Women's Empowerment Principles, and we've made a commitment to the Working with Cancer Pledge to build a supportive workplace for those who are living with cancer. Our own Ipsos Foundation, which was established in 2015, has funded 112 projects in 43 countries. The mission of the Ipsos Foundation is to provide access to education for disadvantaged children and youth. Not only are these initiatives good for the world and the right thing to do, but they're very important to our employees, and they drive engagement. We know from our own pulse survey, which is our internal employee satisfaction survey, that employee engagement scores are much higher among employees who believe that Ipsos is committed to protecting the environment and who feel that Ipsos has a diverse and inclusive culture.
Our work on ESG is important for Ipsos to be able to retain and attract. We're also making a difference in the world by helping clients on their own ESG journeys. Our ESG Risk Monitor enables clients to leverage digital data and AI to help them manage risk in almost real time. Michel mentioned this earlier. We're doing this for the World Bank to provide them with an early warning system, so that when they're providing financing to invest in projects such as building a road, they're ensuring that the people and the environment and communities around that project are protected from potential adverse impacts. We've just completed an important project on damage assessment following the earthquakes in Syria and Turkey, which was recognized by the World Bank with the Sustainable Development Global Practice Collaboration Award.
Importantly, we're now selling the ESG Risk Monitor to private sector clients. For example, in tech, CPG, energy, and financial services. We are helping governments to manage the effectiveness of their programs around protecting the environment and improving society. Ben alluded to this earlier. For the government of India, we're helping them evaluate the progress in providing clean, safe drinking water. We're interviewing 5 million households, and we're conducting lab testing of water samples from 27,000 locations. We're also doing a lot of work around sustainable packaging. As an example, for a major beverage company, we're helping them to evaluate the appeal of returnable packaging. Our ESG economics offer helps clients understand whether they're getting a return on their ESG investment.
We completed a project for Braskem, a Brazilian petrochemical company, to help them understand the impact of their ESG sponsorship on Braskem's public image and reputation, and the economic value that was generated from those sponsorships. We have a comprehensive offer around ESG brand performance. For Coca-Cola, we're measuring the impact of their marketing initiatives at the product brand level and are providing them with ESG brand performance metrics. Communications, as we know, is an important area for ESG. We're doing a lot of work here. Google partnered with Ipsos for research and advisory engagement to help them understand how their client base should communicate around sustainability, to get people to actually engage in more sustainable behaviors in different categories.
Now, a key point about Ipsos, which distinguishes us, is that we're not just a commercial organization, but we also have a role and a responsibility to share our accurate data and reliable information with citizens and the media, to inform them about how people and society are evolving on ESG. We share our data and insights here on our website, in webinars, on social media, and with our clients directly. We've had a consistent voice on the market around ESG, with releases on various environmental, sustainability, and social topics such as inequality. These materials, and many more, are available on our new ESG web pages, which are accessible through ipsos.com. Ben, back to you.
Thanks so much, Lauren. A lot of stuff, I hope that gives you a sense of the diversity of what we're doing, the speed at which we're moving in new areas, the blend of the human and technology. All I'd say finally is, the world, I think, is becoming more complex, but I really believe this company is one that is built actually to master complexity. We've got experts all over the world on the ground. We have technology combined with numerous examples of real-world evidence, and you've heard about a few of those today. We are on a three-year journey, which we started last year, looking at tech, changing. Looking at our people, obviously, mergers and acquisitions. That's underway.
There's a clear strategy, and fundamentally, this, in this sector, is a unique company with unique capabilities, and I hope we have demonstrated some of that today. Now, Dan and I and others are very happy to take any questions you might have. We'll also try and get you some air.
Meanwhile, pardon the interruption. There's the operator, meanwhile, for those from audio, if you would like to ask a question.
This is Investor Day, Emmanuel Matot from ODDO BHF. Two questions for me, please. First, to come back to generative AI. I understand it should improve your productivity, but do you expect more pricing pressures, more competition in some part of your addressable market? I'm sure there are already huge investments from startups in that segment. Do you see already that pressure from new competitors? My second question is regarding M&A. We are waiting for a long time now. You talk about that.
Yeah.
Ben. In case you will not achieve any sizable acquisition this year, would you be ready to increase return to shareholders as from 2024? Thank you.
Crudely, we aren't yet seeing any changes on pricing pressures as a result of generative AI, but I think only the paranoid survive, and we're obviously acutely sensitive to that. I think what we can see is that we can see very clear, quick productivity gains that we can deploy. How that will evolve in terms of pricing, we'll just have to wait to see. To be honest, it's too soon to tell.
I think although we've covered a lot on generative AI today because we knew there was market interest, I think it's also worth saying that if you look at major technological advances and how long it takes them to actually release overall productivity gains in an economy, if you look at something like the electrification of the economy or the invention of motor cars, it isn't. Now we're talking digital, so of course, everything moves faster, and we know how many people started using these tools so much more quickly than anything else. I think we need to be quite cautious about that, but we are certainly ready for that.
On M&A, I think very crudely, I think Dan would agree, if we, if we don't do something substantive, we will return more money to shareholders through share buybacks.
We said at the beginning of the year that we would target EUR 50 million in terms of share buyback with the view of consolidation, if nothing happens specifically on M&A. Obviously, our target, as you understand, is our strategy to keep ongoing on M&A, and we have a lot of discussions going on. It is true that if nothing happened within the next few months, then we will do a bit more share buyback, yeah.
I think it, and I'm, I understand everybody's impatience, and I'm impatient, too, but it's also so important to get these right. As other investors message me and Dan all the time, "Don't overpay." There's a, you know, if you want us, if you want us to spend money, we can do it, but we're trying to get that balance right, and particularly get things that are right for the business that will deliver value over time, and when we're not just gonna burn something.
Yes, Conor O'Shea, Kepler Cheuvreux. Three questions, please. First question, just to follow up on the AI question. Your peer, Kantar has cited weakness in the U.S. I think you referred to it in terms of the TMT client side. Is that the only factor? Obviously, that sector has seen slowing growth for.
It's a major part of the American. Many of those businesses are American and a lot of their spend is in America, the slowdown in that sector spend will disproportionately affect the market research industry in the United States, I would say, and that's pretty clear, to be honest. You can see that in the Kantar numbers. Hopefully, you'll be pleasant. Well, you'll be maybe slightly pleasantly surprised by our numbers, but, you know, we'll go into that in due course.
No AI effect yet?
No, there's, I don't think there's.
Okay.
There's no sign of that. There are, you know, other challenges around, you know, using cloud, et cetera. Generative AI is still really in its infancy, and I think there are still real challenges around hallucination. I don't know if you've practiced asking the same question on four or five platforms. I even went on something this morning. There are now many apps you can use to test if an image is made by AI. This beautiful photo, you've seen, is that AI or not? I took one of the ones that I created on DALL-E, of my, a car like mine in front of an Italian palace. I load it up, and I say: Is this, you know, is this AI or human? Oh, it's human. It's like, so, you know, AI can't spot AI.
There are some real issues that here that we need to you know, we need to deal with.
At Kantar, also, they lowered the guidance. I think they have similar sort of full year targets, but at a range of 3%-5% full year, and I think now they're going for 3% because, you know, particularly the U.S. is a bit weaker-
Yeah
... than expected. Obviously, you were off to a slow start in Q1 because of the multi-year comp and so on. Are you still confident that 5% is, you know?
We said around 5%.
Around 5%.
At all of the numbers that we've looked at, Dan Lévy and I, and we've interrogated it every single which way, at the end of May, what we can see is a very clear path to that, which is why we haven't changed the guidance. You know, there's seven months to go.
Obviously, hear that there are uncertainties and macro uncertainties. We said that one of the most important factor for a slowdown was the big tech in the U.S. As I said before, some of them are rebounding. Some, for some of them, the demands is still low, so the question is more or less the timing actually. On the other hand, we have a lot of good signals in terms of acceleration of the order book, as I said before, and it's been keeping accelerating since the beginning of the year. On the chart, I saw the revenue that we have visible towards the end of the year, make us quite comfortable with the around 5%.
Obviously, you know, we have a average maturity contract, which is a few months, so the visibility we have towards the end of the year is not absolute. There are uncertainties, but we don't have any reason to move the guidance now.
The last question, Ben, you referred to your ability to touch the offline world.
Yeah.
One out of three. I mean, to what extent is that, insights from there demanded by the private sector, or is it more sort of social programs and government?
No, it's both. I mean, it's both. If you want to sell cigarettes or beer in Nigeria, and you, again, you think you're gonna do it online, you've got some issues. If you want to do that in, you know, in Pakistan, if you want to do that in India, in large parts of Asia, you need to be able to do that. There's a major financial institution at the moment that we're working with, which is why, which is just coming to mind. They were working with a digital-only American company, and then they said: "Well, we need to understand where these money flows are going, so we need..." They asked their American digital company: "Could you go and do some surveys in Mexico, in India, in Pakistan, in the Philippines?" They couldn't.
They came to Ipsos because their, the data that they were going to provide would just not adequately represent the market in those countries. That, you know, the world is complex, and I think this, if we think of it as a purely online world, we're missing a large part of what actually happens.
Being offline doesn't necessarily correlate with low purchasing power?
Well, obviously, the people who are offline are poorer than the people who are online as a generalization, but there is a market for very high-quality data. I mean, it's one of the reasons why some of the biggest tech companies are actually buying mixtures of offline and online data because they have a lot of data. To understand the country, many countries, you need to actually have offline and online data, because if you only look at the online section of the population, you will get a very strange idea about what that country is like, what attitudes are about in that country. It's a mixture of both, of governments, obviously, the development projects, et cetera, you know, it's huge interest, billions being spent.
Also, of course, actually, yes, you know, in lots of Africa, in lots of Asia, commercial research has to rely on offline data collection because the population, you know, the, you've got the largest.