YouGov plc ($YOU)

Hello, and welcome. Thank you very much for coming. Now I realize you're very much focused on the numbers this morning, but I want to start with this key message. Next one, there we go, that I realize, as I say, you'll focus on the numbers, but because I think so many people in the market believe AI is a problem for us, I want to emphasize this. We certainly believe there is great disruption in our industry, but this will favor YouGov, favor us, not harm us as we are investing in AI to maximize that opportunity. So it's against that background of expected disruption that we have shown incredible resilience. We've delivered reported revenue up by 2% to GBP 194.8 million when most companies are flat or down. We're showing a statutory operating profit up by 14% and statutory profit before tax up by -- sorry, [ 14% ] for the revenue and profit up by 4%. Data products up 2% like-for-like, that is excluding a discontinued product and adjusted profit is down for the half year to GBP 24 million, reflecting essential investment in Shoppers data gathering methods and also investment in AI. Earnings per share is 11.4p and the balance sheet is solid at nearly GBP 33 million positive with a net leverage of 2.1. Profitability, of course, is affected by investment partly to -- in shopper to keep it competitive and partly in AI to transform the company. Given what we consider to be pretty extreme undervaluation of this company, we do intend to have a share buyback after refinancing. And while the macroeconomic environment remains uncertain, clients are continuing to prioritize the high-quality human data and strategic research projects that are where YouGov continues to be strongly positioned. This is a quick reminder that we had 15 years of growth driven by innovation. That is the DNA of this company. And as I will show you a little later, AI is driving some amazing innovation at YouGov. We've launched an add-on to brand index. Voices, renewals have been steady. And in the new year, as in the old year, we've hit record highs for a single subscription in this case, to a major tech company. We've -- in research, we've had strong performances, especially in banking and retail and in all regions, especially in America. A particular favorite of mine was a major study of perceptions of AI across the U.S. for Anthropic, which they released as a major public report. In Shopper, we've expanded and upgraded our panels, and we've added a new method for data collection, which was essential to be competitive in a changing landscape. And given the severe undervaluation of this company, we are conducting a strategic review of how Shopper best fits. I now hand over to my colleague, our CFO, James Davies, but let me first say what a pleasure it is to be supported in our transformation by such an experienced, talented and determined individual. James?

Thank you, Stephan, and hello, everyone. I've enjoyed my first months as CFO of YouGov. This is a great business with passionate people around the whole globe. It has an exceptional platform and brand, although we are not optimizing its potential, and there is a lot of work to do. We are not as efficient and streamlined as I would like us to be. 5 weeks in, and I am seeing many areas of productivity and efficiency improvements that must be executed. These are all within our gift. We are not yet embracing modern ways of optimizing the value we can drive from our market-leading panel and trusted brand. This doesn't just relate to cost cuts, but also more efficient ways of working, collaborating across teams and countries and with a much sharper focus on how we actually optimize margin across our diverse product range. We're in the process of improving the discipline across the business in all of these areas as tools, market dynamics and competition have evolved. My road map is filling up nicely, although my main priority is to focus on growth in the right markets, margin enhancement on a group-wide basis and to bring back agile ways of working. Yes, we are a listed business, although this doesn't mean we can't be slick and entrepreneurial, focusing heavily on maximizing returns on every penny of investment. I don't plan to spend too long looking backwards today. I want us to look forward, and I will be announcing shortly our value delivery plan. The value delivery plan is a full reset plan. However, I will spend a few minutes going back through the segmental P&L performance, along with some data points around our balance sheet and cash flow before moving on to H2 '26 and beyond. As Stephan alluded to, the market reset and ag sector is in a tough patch at the moment and growth is being compromised for many reasons, including budget constraints. However, our product and reputation means we remain highly relevant, and we continue to grow top line by 2% year-on-year. There are not many businesses in our sector or adjacent sectors that can say that. Many are declining in high single digits, let alone growing in single digits. I particularly like the graph Stephan ran through earlier, showing 15 years of growth. That is very impressive stat for a business of our age in the sector in which we operate. Research has been a star performer with Data Products and Shopper stable from a top line perspective. So moving on to adjusted operating profit. There are 3 themes. Firstly, when we compare statutory measures, not adjusted measures, operating profit is up a healthy 14% year-on-year. As a principle, like many of you, I do not like adjusted measures. It can mean a reader can't see the wood for the trees, although I am reassured that our statutory measures have performed well. Also separately identified items are much lower this year relative to last year. There will, however, be separately identifiable items as we progress through the value delivery plan, which I will talk through shortly. However, I commit that the threshold around adjusted measures will be higher and communication clearer going forward. So theme 2, the Data Product division continues to perform very well with an even higher margin than an already attractive margin of 30% that we achieved last year. In this half, we were 35%. This is very encouraging. Theme 3, material incremental investment in Shopper of over GBP 3 million in H1. This is the main reason for the adjusted operating profit reduction versus last year. This investment does continue into '27 and '28. However, we do see revenue benefits hitting us in the '27 financial year. This investment is substantial, but it's imperative we introduce this now to enable semi passive and passive data collection as well as to expand and upgrade panels across Europe. This investment is intended to support its growth trajectory and competitiveness, and we're already seeing early success in improved client delivery and new opportunities. I'll come back to operating profit in a bit more detail shortly. Moving on to revenue by region. 3 of our 4 regions drove revenue growth. Americas sector focus is proving fruitful. And despite the political calendar comp going against us this year, 2% growth was still achieved. U.K. and Asia Pacific also grew nicely. Mainland Europe is our largest revenue and it was flat year-on-year. The larger most strategic European countries performed well, although there are some of [indiscernible] for example, in Nordics. And as we go through the later we focusing on areas that we need to improve performance and some of the smaller European countries definitely fall into that category. There is a slight timing difference for Shopper, but I'm confident that H2 growth will be faster and if not better than expected for the Shopper business. And we will return to year-on-year growth for the European region by the end of the year. Moving on to customer base. Our customer base remains resilient. And from a sector perspective, nicely diversified. Technology remains our largest segment and continues to deliver growth. On some of the detail on this page is the [ FMCG ] aspect of Shopper. This page just shows YouGov core and exclude Shopper. So when you add on Shopper, this is an even more attractive picture. Cash generation. Cash from operations is GBP 3 million times lower than last year. That is less than the reduction in adjusted operating profit. Cash conversion efficiencies are supporting this cash conversion measure to. However, I've been here 4 or 5 weeks, whenever I need to focus on more is to improve the cash conversion and focus more heavily on working capital improvements and smoothing as we go through the next few quarters. Our business is very seasonal from a cash flow perspective, as many of you know. Moving on to the financial sheets. Since H1 '25, we paid down EUR 20 million of our term debt, which occurred in October of last year. The net payment is for the same amounts due in October of this year. However, I'm delighted to say that we have already commenced our refinancing exercise for both the term loan and our fully drawn RCF. Our debt ratios we net debt to EBITDA and interest cover are both well within the terms of the current loan docs. I'm not concerned about liquidity for this business, although I am looking forward to doing a refinancing which will better reflect the structure and needs of the balance sheet as to where we are at the moment. Moving back to operating profit. Yes, I was keen to view 2 bridges. The first one is the H1 bridge. This is basically earning story H1 this year versus H1 last year. The [indiscernible] shows a GBP 6 million reduction year-on-year. Half of this is due to the additional Shopper investments already communicated. The GBP 3 million represents the incremental level of Shopper investments in H1 '26 versus H1 '25. In H2, a further delta is expected to be of getting up circa GBP 3 million to GBP 3.5 million with further investment in '27 and '28. But as explained, we do start to see returns coming in, in 2027. I could divert a narrative, which says of a further year-on-year value also of GBP 3 million is due to a clear and specific investment program in YouGov core which would be delivered very clearly defining returns in [ 8 ] years. However, I won't do this, I see this earnings variance has been broadly in the normal course of business. Yes, there is investment in our platform and tax footprint, but there is also GBP 1 million additional cost to the significant progress we've made in the Shopper work stream. However, I class is actually GBP 3 million as ordinary course activities. So to conclude, [indiscernible] we are GBP 3 million down due to extra costs that have been incurred. Moving on to H2. So I'm now starting to look at the future following the escalation of the past. Last year, H2 was flat with H1. This year, we will see growth in adjusted operating profits in H2. We are guiding to an adjusted operating profit outcome for the full financial year between GBP 52 million and GBP 56 million, which gets us into the bottom end of the consensus range. Let me talk you through the chart below on the seen, and I can explain why I've communicated the range. Shopper investments have approximately GBP 3 million to GBP 3.5 million will be incurred in H2. We also have GBP 2.2 million of additional costs in the second half. This is a consistent theme in H1 that the year-on-year delta is due to shoppers TSAs and platform investment in the core business. Again, not specific enough for me to take out as an investment variance. Plus, we have a benefit of no less than GBP 1 million, which I'm calling margin improvements or value creation in H2. What do I mean by this? I mean, as of now, I can identify this uplift is relative to last year. I have very high conviction of achieving this number. This number gets us to the bottom end of our range. My work on a more specific year-end outcome -- outturn is ongoing. And there are initiatives in flight, which could mean value growth margin enhancements could be up to GBP 4 million higher. This will get us to the GBP 56 million number. We will be working hard over the coming months to optimize where we end up. However, I'd like to be very clear that we are not solving for an optimum FY '26 operating profit. We have sold them for a value optimized in 2027 and beyond. They'll be very clear at the year-end along how and why we ended up with the prime position within this stated range. There at the year-end alone how and why we ended up with [indiscernible] imposition within this stated range. I also wanted to highlight that by simply taking out the Shopper investments, the guidance for the year would be between GBP 58 million and GBP 63 million. I want to reiterate that I could take out GBP 3 million YouGov core to add to this number, but I won't. This is normal to business under my definitions. Moving on to the value delivery plan. We have covered sort of numbers and the outlook is actually here, but I want to talk about 2027 and beyond because this is where it gets really interesting. I'd like to follow some of the details as Stephan will go through shortly. There are 3 ways to our value delivery plan. This is a plan that we've kicked off in the last few weeks. Second session below is going to mainly cover Wave 3 and part of Wave 2. But let me go through my version of each wave now. Wave 1, as I mentioned, I've been at YouGov over 5 weeks, and I was very keen to ensure that we could land some key messages with conviction stay. I want us to be able to provide confidence that we are taking all appropriate actions to optimize value and earnings I do not want to sit here today and say, "I was going to do this. I will do that." I wanted to say that I have done something already with the full support of the board. So we've already actioned annual run rate margin optimization of GBP 2.5 million where we have eliminated terms of costs and several inefficiencies predominantly in nonrevenue-generating areas. This is real, and I'm delighted to say that Wave 1 of the [ VDP ] has been completed. We are just starting this new focus of executing your face, and this will continue. Wave 2. This kicks off next week, which is needed to get the business back to where it needs to be. [indiscernible] today is where the solution to every problem is to increase headcount and further inefficiency and so that focus on accountability. This is stopping and will stop in its entirety very soon. I can [indiscernible] provide you an update on Wave 2 at t he Capital Markets Day in the summer. It won't be complete by then, although it will be reasonably advanced. The earnings uplift from Wave 2 will be a multiple of the value of the Wave 1 uplift and the shrink of this benefit will hit the financial year 2027. The combined impact of the first 2 waves is expected to deliver an annualized adjusted operating profit margin uplift in excess of 350 basis points relative to the margin achieved in H1 FY '26, 50 basis points relative to the margin achieved in H1 FY '26. This is once it is fully executed. For example, if we manage to fully execute this before the year-end, the full benefits of this uplift will happen in FY '27. If it doesn't all happen for the end of this financial year, there will be elements of this investment within FY '27. Again, looking forward to giving more updates on this in the Capital Markets Day and at the full year results presentation in October. In Wave 2, as you see from the graph, that's where we start to implement the AI transformation program. Wave 1 with nothing was over to do of AI, Wave 2 is partly to AI and Wave 3 is fully to do with AI. So moving on to Wave 3. This really is a exciting piece, and Stephan will go through this in detail shortly. This wave represents a material evolution in how we win, operate and serve clients. Wave 3 is expected to deliver a margin profile aligned with an AI-led data business, making a step change from our historical margin structure. I can't yet unify the specific financial shape this will result in. However, I'm starting to get an incline of what this could look like, and it's very exciting. We will be sharing further detail on this in due course. I now like to pass back over to Stephan to go through Wave 3 and many other topics in more detail.

Thank you very much, James. Yes. So look, the AI disruption is not only about a big operational savings. The bigger picture is that the age of AI demand is exactly the kind of data that YouGov deliver so efficiently. And I'd like to remind everybody that it was YouGov that first build an engine for automated research data at scale. And by the way, we're still the only ones that do this. Since 2007, for 19 years, brand index has been selecting some running daily surveys, processing the resulting data and delivering into an analytics platform without any issue intervention. It's all automated that has never broken. That data goes into our data lake, which structured and corporate the cube and it generates a variety of products, including brand index profiles and rates. So the single source of human data is our mode. Nobody else has it, and nobody else can do it quickly. And that's why I mean data disruption in AI will disadvantage most companies, but it will massively benefit YouGov because the advantage for us is inherent in our data generation engine. A couple more specifics about that disruption, which sit is very welcome. On the left, individual samples will be commoditized, people safe on those are panels just becoming a commodity. But we have to differentiate between supplying near samples, but having an engaged pan. These are different things. And engaged panel builds constant stream of connected data is not just like the [indiscernible]. [indiscernible] panel is a [ concall ] and behavioral single-source data system. The human operations that are required for our conventional research with YouGov, this will be 100% end-to-end automated. We already have [indiscernible] right now, and LLM has direct access to our system and can create its own machine research have any human intervention on either side. The only human been develop panels -- has talking directly with [indiscernible]. And then there's a consultancy part, which is, of course, a huge part of the market research industry that has always been scalable and like most forms of consultancy will be directly replaced for the most part by AI. I guess some teams in using the system. Yes, expect deploying the data, yes, for consultancy whole. This Is a dynamic engine for premium connected structured universal human data. It's a utility for everyone to use. I'll be quick on this slide. You already saw it 6 months ago, and it's operationalized now. So it's tens of thousands of people every day, not only cooking on brand index questions but then talking about the markets and adding the quality why data to the quality of what data. So you get the moving lines and the transcript of the [indiscernible] from the same Qs. And the first [indiscernible] for this was a [indiscernible] that we launched a few weeks ago and it has aroused initial interest, but it's only scratching the surface of what's coming. This is the new operating model for generating the most relevant human data made dynamic by AI. It's a learning. The left-hand side of the diagram is the flywheel of the data generation, which has been going through '19. A global panel automatically serves surveys the ocular global panel automatically service surveys that going to the structured data lake. Number two, is the dynamic, the inflection. Number three the data lake cube, which is structured and contains all of that data in a form that generates trusted data products. Is number four, and that's served up into dashboards and an analytics platform. Now there are recuts as well and to AI systems. That big middle of the left, the 5 pillars is the YouGov system for generating our connected data. But now we have the dynamic part on the right-hand side, a learning rule. So the right hand side -- data seek agent, which is -- which looks at the data and did what takes needed and an AI analysis system that works out what the uncertainties are in the data and what then is needed in order to fulfill that. And that puts us back into the flywheel as it well. It changes the dynamics, daily collection and means that an updated too. So let us put it this way. Imagine you have a number of new straight trend line. And then new data starts to look at it both as a move away from that trend line to load assets is away from that trend line. That will be the detected, and it would mean we will increase data collection to make sure that our imputation model and in the consistency of the trends is understood and then corrected or update should be in the model. That will generate automatic alerts to users and will be a trigger for new research about. This is a way of a new way of seeing research, not as a series of surveys not as a series of a disconnected [indiscernible] of insight but a model of your markets that is continuously updated in an intelligent way, without you having to do anything and getting the alerts for what's happening at what's changing. So that's the revolution that's coming, and it will be pioneered by YouGov, the Internet revolution in research was pioneered by YouGov. And this is what excites me and what is coming down the [indiscernible] faster than you might expect. But it all exists, and we're going to the next stage with it. So finally, I'd like to introduce our new Chairman, and man with so much experience in our industry that I wanted him to succeed me as CEO when the time comes, but sadly he turned me down. Our new Chairman Ian Griffiths.

Thank you, Stephan, James. Good morning, everyone. I just want to close the few comments, recognizing where the business is also putting together some of the key themes that we've announced in our RS this morning. But I would just like to say that when I joined the Board back in September, I was genuinely excited about the opportunity be part the team, the guest's is business back to delivering sustainable growth. Stephan has traded something rather unique at YouGov. It's a panel of fast and breath driven by data from real people. And as, as you said, it's almost impossible to replicate and is very different than anything else in the industry. Now I recognize that this business and the shareholders have had a challenging couple of years. However, I hope you agree and almost [indiscernible] today's share price reaction. But what you've heard today does starts to set a different tone and sensor direction. Under James, there's increased grip and focus on our financials. We have a new value delivery plan, which we're executing at pace. We're getting much clear about how the business can be positioned as a leader or using Stephan's phrase as a pioneer in the age of AI. And we've kicked up a strategic review of our shopper business. As the next phase, the Board has asked the team to put all of this together into a new business plan that will be reviewed in May and it will form the basis for a Capital Markets Day later in the year. And I believe probably the most important role for this Board is how we allocate capital to create value for our shareholders. And it was very clear even before today that there is a material disconnect between the underlying value of the business and the share price. As a result, the board has agree that annual return to shareholders this year will be by way of a share buyback rather than the usual dividend. I also want the board with more of a private equity mindset, one that really aligns the team and to shareholders. We've already reduced the number of net [indiscernible] and by the time of our AGM late this year, we'll be down to 4, which feels about the right size. And then finally, for me, let me deal with the CEO succession because they were announcing the start of the process to find a permanent successor to step up. That process will take as long as is needed for us to find the right person. As you've heard today, there is no rush. When Stephan will step back into the CEO role, he sends off a series of objectives. And as you've heard today, we believe there's real progress delivering against that. And the pathway is an exciting value-creating future is becoming clearer. There's clearly a lot to do. There's new initiatives underway, a view on the direction that we've set. And I hope you agree, a real positive sense of momentum Pace is important but so is executing brilliantly. The Board is genuinely excited about what's possible for this business, and we look forward to updating you as our delivery progresses. That's what I want to say at this point. I'm now going to hand you back to Stephan and James to lead you through the Q&A. Thank you.

And we do have quite a few questions coming in. Could I just remind anyone wanting to post questions to please put it in the Q&A. So the first question coming out is from Sean Thapar. And we have a few questions from Sean, thank you very much. The first one is you are pursuing investment in FY '26, but are expecting the value delivery plan to deliver enhancements to profitability in FY '27. What gives you confidence that this level of investment is sufficient rather than requiring a sustained investment period over multiple years. Relating to that, what details can you share that give you confidence that there will be an immediate improvement in profitability from next year?

[indiscernible] First of all, thank you for your question, Sean. So is the investment in that -- it's like only been here a month or so, I think the investments in the Shopper business has been structured well. It's been -- it didn't start in the last few weeks or months, it started last year. So it's very well structured at has come through the required due diligence and asset from the central stage. I'm confident that the work that the investment is done is in the right way. And as we explained, that investment is predominantly a Shopper with the secondary identifiable pieces, the investment within the rest of the business at the moment on I'm keeping that in the sort traditional operating profit category. I now move nicely answer your -- the second part of your question on margin. So yes, I've got high conviction that we will get to that 350 bps improvement during [indiscernible] but during FY '27. And that will be after to any further investment we may choose. So for example, Wave 2 to is not going to be about pure cost cutting. There will be some cost cuts, but we'll also be investing in different areas as well. So the net benefit will be the 350 increase in on margin, but we'll be thinking about how the business reshapes as we enter Wave 3.

Thank you very much. We have a few more questions from Sean. On the balance of investment, you have earmarked 6 million for Shopper and 3 million for YouGov platform. Given how ambitious you are being on the platform side, AI research, delivery, et cetera. Why is the largest portion of investment going into shopping now rather than -- why is the largest portion of investment going into Shopper now rather than allocating more to the YouGov core?

Well, I think that the investment in -- to the core part is the large amount. I -- our teams know what they're doing. It's actually very efficient where we generate efficiencies as we move forward on that. So about that. I think the investment in Shopper is about competing in a new environment, and it does mean some important new pieces are being built. That does feed back into our core data pieces, which is not entirely separate. The new method of methodology for gagging Shopping data that generates not only purchases in the shopping basket, but in fact, all kinds of [indiscernible] across a variety of different outlets, everything from gas stations to [ fit ] tickets, to visits to department stores. It isn't limited towards hoping supermarkets. Therefore, it provides a constant stream of verified purchases that can be that can lead to instantaneous interviews or interview people who just made those purchases, and that can be not to by [ Quanta ] but [indiscernible] as well, in other words, by conversation. So it does actually open up new closely related products to the main zSpace product. I wouldn't see it as a sample altogether, I would say as part of the total.

We have a few more parts to Sean's question. Your strategic review of Shopper seems to suggest 2 potential outcomes for the business. either disposal or deeper integration. Your current investment appears to be a bet on expanding the business. Can you indicate which outcome you favor for the business? And how the investment you're making today aligns with that?

To be clear, we see some of that as important to the strategic direction of zSpace in terms of data that generates more and more updated generation. And one way that needs some certain is part of our strategy, but there are ways that we can do that with our Shopper like if that's as necessary. It wouldn't be something we'd be doing where we now so undervalued or nothing the shareholders must want us to consider. Do you want to add something to that?

No, I just really want to reinforce the last point Stephan made that if it wasn't our share price position, we would not be doing a strategic view Shopper. The rationale when we bought the deal is exactly as strong now as it was then. And it would be wrong of us not to assess the potential returns of sale of shopper could do to the shareholder base. So the question you raised, Sean, was about further integration or sale. That's spot on. We've integrated it to a certain level. We've cleared out pretty much all of the transition services agreements from the previous owner. And it is now in a nice clean self-contained space, and we will be making a decision over the coming quarters whether we do go for the big full integration or the value creation is better with another owner. But I would like to reinforce that we're not in active discussions at the moment. So this is not going to happen over the coming weeks or months. It will be a slightly longer burn. And the investment we're doing, we've got half -- it will enhance the sale should that be the route we go down. If we do keep it, it will obviously enhance the ownership of our ownership going forward as well.

And final part of Sean's question, could you share what programs are underway to deliver the value creation component of the FY '26 profitability plan? And what drives the variance behind the GBP 1 million to GBP 5 million contribution range?

Very good question. Yes, if we're doing this results presentation in 2, 3 months' time, I could probably be much more clear on not having a range, but there's quite a few initiatives we have actioned in the previous weeks, which will dictate where we end up in that range. But as I alluded to, and I think this is an important point, and I will -- please do hold me to this in the summer in October. We're not solving for a 2026 number. The waves are very exciting, and we want to make the right decision with investment as well as efficiencies over the coming quarters. So yes, there are a few decisions we'll be making as a Board going forward. But as [ Ian ] said, we are going to be thinking more private equity like and returns based than maybe we have in the past. But yes, looking forward to updating you in more detail over the coming announcements.

Thank you. And now we have a question from [ Leo Mansor ]. He has asked in reference to underperforming products, what are these? And are there other products like that still?

This specifically refers to the residence or free wall piece that we were using for recruitment. We now have a much better way of doing that. We are recruiting. I think I neglected to look at my last slide where I talk about -- apologies for that. So I just fit in this very important part of it. But we are now using AI as well in the recruitment process and the recruitment process generates data, whether someone joins a panel or not. This is one of the really big advantages of our methodology that I don't know how I missed that, but it's alluded to in that last slide. So we have something better, and we don't need a free wall part. It was not generating the value that we wanted it to. Things have changed, the way people interact have changed. and we are obviously modernizing our systems with that. That's what happened there. There are a few often strain -- what's the word stray little data products that have grown up over the years. Nothing consequential. But here and there, there are some things that we discontinued as they come out of contract, but they're trivial. It's nothing significant.

One thing I would just add, and I alluded to this in my section is we are very global. We have 52 offices around the world. And there are some regions that I think we can be more efficient in operating in such regions. And I think that's part of the reason for our European performance in H1. I'm optimistic -- very optimistic that growth will be there for Europe as a whole for the year, but we are doing a deep dive into how we operate on a regional basis for accountability and efficiency purposes.

And now we have a question from Jessica Pok. Do you have the skills within the business for Wave 3 of the value creation plan? Can we expect reskilling or replacing the staff?

Well, I definitely think that automating research operations must lead to, I wouldn't say significant, I would say, huge efficiencies. I think this is fundamental when other companies are talking about technical companies, tech companies are talking about this, and they're talking about 40% reductions in workforce. I'm not going to put that out there, but we are a data generation company. That is the most important thing we do. And that has involved quite a lot of heavy lifting inside those processes, which will be done by AI. So we are expecting really significant change in that. Now obviously, we hope to grow in other aspects of our business such that it doesn't necessarily mean that all of those people are lost, but I do think it will be a skinnier business. I think it needs to be. I think that is the effect of AI and [indiscernible] to talk about it, but we must be realistic. We see great efficiencies in this.

And the second question from Jessica. Can you provide more color on the progress of YouGov voices and how it has been received by clients? What percentage of your panel is interacting via chat for YouGov voices?

So I'm not going to give a blow-by-blow account of that a couple of weeks after launch, but it was a very generative launch, lots and lots of conversations and better than that, but I'm not going to give that right now. But I would say that something like 1/4 of our active panelists that are actually doing a survey are going on to do conversations. That is within the countries where we're doing this, which is U.K. and U.S. I don't have the exact number because it differs from day-to-day.

And now a question from Michaels [indiscernible] . Michael, apologies if I haven't pronounced your name correctly. Regarding the buyback, consensus is expecting you to generate just under GBP 30 million of free cash flow. Is it reasonable to expect a buyback of this size? Or will it just be the equivalent of what the dividend payment was expected to be, i.e., closer to GBP 10 million?

First of all, Michael, that's a good question. The answer is there's an element of flexibility. One thing that I would like to reiterate though is we are doing a refinance. We are not going to be doing a refinance to lever up to do a share buyback. But as you correctly pointed out, we do have generate free cash flow. So there is an element of flexibility, but I would use as your base case as a starting point, an amount broadly in line with the annual dividend payment. But I wouldn't say if not, it couldn't be potentially margin higher should we feel comfortable with the cash flow at that point in time. And in terms of the analyst modeling, I would assume that this happens in H1 2027 in terms of timing.

On August onwards.

Yes, I'm still getting used to our year-end, yes, so August onwards.

And now we have a question from Steve Liechti. It's in relation to the VDP. And the question is, why no revenue growth targets? And one related to you specifically, James, are reporting budgeting systems good enough to deliver cost savings? And then the final part of this question is related to the AI learning loop. When is it in place and revenue generating? Is this a product thing or changing or changing all existing operations?

Great. First of all, thanks, Steve. And as always, great questions from yourself. And so first of all, revenue, I'm deliberately focusing on margin on disclosure of the plan at this stage in the summer and in October, we'll talk more about the revenue side because I feel at the moment with the AI piece in Wave 3, that's still bit still getting sort of finalized from our thinking. So at the moment, we're focusing very confidently on the margin improvements, but the revenue picture will become more clear as we shape out Wave 3. In terms of your second question about systems, I work in many companies now and systems are never perfect. The systems that you got are not perfect, but do I have enough MI and real-time information to enable to deliver these benefits? The answer is most definitely yes. I don't believe we need to invest heavily in our back-office systems yet, but there will be some investments going through in the coming quarters to make sure we do have what we need to go through Wave 2 and 3.

And Steve, on the third question, a really interesting question, and you've clearly got what's happening there. So that divides into 2 bits. There's the learning loop that the AI that the LLM can do and that is now. So in other words, if somebody is -- if an AI or indeed a researcher looks at the data coming in and says or their own data that they have and says we need more, that system defines what is the data that is to be collected. They generate that through their systems into our system and get that data back. So that learning loop exists with the client AI, they can learn in real time, they can update their models in real time using our engine and that is by them. And that is today. That is not -- I'm not saying I talked about a prototype. It's a working prototype that's functioning. But obviously, we're learning from that functioning. Now the second part is when do we do it for ourselves because you have that LLM bit in the other pillar, the other part of the pillar is when we do it for ourselves. And that really goes hand in hand, I think, with our research partners. They are our client partners. I can't really say more than that at the moment.

And here we have a question from James. It's in reference to Wave 3. And the question is Wave 3 references a margin profile aligned with an AI-led data business. Can you give us a target range for what that means? Are we talking 20%, 25% or higher?

That's a question I was hoping I wouldn't ask from being honest at this stage. I don't think I'm ready to give any number there, but it will be more than we're getting at the moment. So I know that's not overly helpful. But please do bear with us. And as soon as we feel comfortable giving a range in that level, we will do so.

A question from William Larwood. Are you seeing pricing pressure in the market? And if so, which areas are you seeing it?

Let me start off and then Stephan can add to. So I think the one thing which I was pleased that when I first joined was I double checked what our order book look like and our pipeline look like from a client perspective. And I was particularly pleased that -- and I think we got this in one of the releases that we are 80% covered for the full year already. And when I say covered, I don't mean order book, I don't mean pipeline, I'd be contractually covered. And that is 100 basis points higher than where we were at this position before. So that gives hopefully an element of comfort that from a revenue perspective, we're feeling pretty comfortable. We're not being complacent, obviously, but we're feeling comfortable. And obviously, within that number is both price and volume. So I think hopefully, that gives you comfort that pricing pressure is not something that's hitting us from all sides at the moment.

I'm just going back to a few additional questions from [ Leo Mansor ]. So we have a question related to Shopper and the investments we made. What are these investments?

So these investments and Stephan can probably answer it's basically changing the way that our panelists collect data because one thing I've learned in my very limited amount of time in the market research sector is you've got to make it easy and clean for certain cohorts to provide the information you want when you want it. And we all know from previous sectors I've worked in user experience is absolutely critical and consumers now compare [ UX ] across all sectors regardless of what that sector may be. So really, this is about introducing semi passive and passive data collection, upgrading panels across Europe and the investment is to make panel more effective, more efficient, more timely and deeper. And that obviously can create many benefits to the big customers that we have in having them in Shopper.

Yes. I think we went over the key part of this, which is the methodology for data collection moving from scanning to receipts and downloaded data uploads and downloads. This is a new methodology, new technology that is in the market and that replaces or adds to the other methodologies that we have. And that requires not only the tech of that, but also the interpretation, the translation, if you like, of the data from something that is incomplete to something that is more complete. I was mentioning, I think, at the SKU level stuff, there's an investment there. And of course, a significant part of it is also in the cost of the new dashboards, I should say, platforms on which the data is delivered.

And 2 more questions from Leo. The first one in relation to our business in Asia Pacific. And the question is, can you explain why your business in Asia Pacific is not profitable? And what is your strategy there? Is it the same offer as in the other geographies?

Well, it's always been a difficult market for us. I can't say anything particularly smart about that. There are parts of it which are, in fact, very profitable and other parts -- other countries where we haven't done well and some difficult countries, as you can imagine, some large companies where we are struggling to not to be so involved. So it's -- I haven't got to say there other than the strong parts are what we will focus on. And as James has already mentioned, there will be geographies that we may not want to be in physically anymore.

And one last question from Leo related to panel development. What is your strategy there? When will you have enough panelists? Didn't you reach the right size with 34 million panelists?

Well, it becomes a very different calculation when the process of recruitment itself gathers data. So this is actually pretty important because a panel must always have fresh minds in there since quite a lot of what we do has to do with awareness. And you can't test people for awareness of things that you've already talked to them about several times. And so you are constantly refreshing your panels and growing them as you have more to do and as you're selling more, of course. So the process of the recruitment is now part of the value generation of our data. So it can't be viewed in quite the same way. We want to keep having new people involved. We also are changing the experience of being a panelist. It won't all be about paid panelists. It will be a lot about unpaid panelists doing it for the experience of being part of a panel, and we found that in some areas, that's better in other areas, panels than necessary. This is actually a core strength of YouGov and a core challenge all the time is to change with the times. So it isn't now exactly as it was 5 years ago.

Thank you. And we have a question from Johnathan Barrett. It's in 3 parts. The first part, can you tell us what price uplift you achieved in data products?

Well, we have not been increasing the price of data products significantly over the last few years. And we feel we need to be doing that, and that's part of James' review, and that certainly is an important part of that. The new data that we're adding, however, will, of course, bring extra revenue to those products as well.

Second part of the question, how many sales of [indiscernible] have been made? And what is the pricing?

Well, as I did say, we just have been there for a couple of weeks. We'll wait for full year.

And the final part, when will the end-to-end product be released? And I think you may have touched on that.

The end-to-end, meaning the end-to-end automated research.

I'm assuming [indiscernible] product.

Yes. So that is what I mean. Well, that is developing all the time. I mean we are running end-to-end [indiscernible] on a case-by-case piece and the will take -- it will keep developing. It's an ever developed product.

Thank you. I think that brings us to the end of our questions for today. So thanks very much. Thank you to everyone who submitted a question.

Thank you very much. Bye everybody.