Ovh Groupe

Ladies and gentlemen, welcome to OVHcloud FY 2025 results. Today's speakers will be Oktav Klaba, Chairman and CEO of OVHcloud and Stephanie Begna CFO. I now hand over to OVH Management to begin today's conference. Thank you.

Hello, I'm Oktav Klaba. I'm the Chairman and CEO of OVHcloud and I'm really glad to be here with you today. to review the financial results with Stephanie. We can start on OVH. OVH Cloud relies on three pillars. The first is that we are in the vast and rapidly growing cloud market. The second is that it's scaled globally. We want to scale worldwide. And the third one is that we just are launching our next strategic plan FY26-30. So on the key highlights of the FY25, we delivered our guidelines that we announced one year ago. 9.3% of revenue, a growth of the revenue, more than 40% of EB day, adjusted EB day, and they've had a free cash flow that's doubled from the last year and the capex about 33% of revenue. We also achieved significant milestones, more than 1 billion revenue. Also in the corporate market, corporate customers, we generated more than 200 million revenue. In public cloud, more than 100 million revenue and also to give you the overall highlight about the US, United States. We generated more than 100 million revenue. Also, we improved significantly all our key financial numbers. Yesterday, board decided to unify the role of chairman and CEO And I've been appointed as the CEO with the main objective to shape, drive our FY30 plan. For this new plan, we need tight governance and the bold execution. Really unify vision, strategy, and execution to make that fully aligned. I would like to thank a lot Benjamin for the work on the last 12 months. Our focus will be on the restoring growth marathon, boosting free cash flow and improving growth share. We have the five years plan investment and also, you know, last 10 years we invested heavily. It's time really to have the money back on all this investment. This is my mindset for the next five years. So just to give you also the the app site that we have. So I've been talking about the 200 million in corporate. Yes, but it's just in France because we have 80% of this revenue is in France. So our, let's say, our takeaway of that is that we have playbook for the corporate market. And now it's time to upgrade in France because we have new opportunities to go after the new kind of customers, bigger customers. but also to deploy this playbook that we had in France, in Italy and Germany. So this is what we will do in the next quarters, years. The second significant upside is in public cloud. As you see, we just generated 20% of our revenue in public cloud based on PaaS. Usually, you should have 50%. We still have a lot of upsets. It's really just beginning of the story on the public cloud. Last one, it's US. We generated more than 100 million in US. So we have a success in US. It's rare for a European company to have a success in US. This is our case. But what you can see that our 100 million revenue, it's mostly private cloud. So we didn't even start our story of public cloud in the US. So this is also what we will do in the next quarters and years. Let's see about the financial results, the financial key numbers. About adjusted EBITDA, you see the cycle of five years investment. we increased adjusted EBITDA over the last five years, but also what you see, that you see this percentage of adjusted EBITDA, it's now growing. We just restored the minimum level, and we continue to grow on this adjusted EBITDA. And you can see that also on the elevated free cash flow, that it's just positive this year, and it's just one step ahead that we will go back to you next year announcing that we are free cash flow and then we continue to generate cash for the next years. This is the first time that we would like to share with you Roche since IPO so we start average Roche in OVH. Keep in mind that Roche is based on the two different activities. The first activity is investment in the hardware and infrastructure. And the second is investment in the software. So this is average Roche that we have in OVH. And if you come to our investor day, we will show you, we'll split that and we'll show you how it's really involved in the next years. So let's have a look on what we call go-to market. And this is a new way of showing you our activity. Of course, we have to keep universes of product, but we have also in the mind that all these products, they are going to the market in a different way. So the first way is digital startups. It's all small customers that say, less than 25k annual revenue. Here we have less than 5% of revenue of Groupe. I will not tell you that it is fantastic and we probably go back to that. The second is digital scalers and here you can see what is our growth. It's more than 20%. We totally succeed in the growing, helping our customers, digital customers, still digital customers growing with OVH. Upselling the product, having different geographies. So we have a playbook and we know how to do that in the next years. And the last one is corporate. It's totally new, let's say. It was totally new five, six years ago. to go on this new market, corporate, big companies, banks, pharma, governments. And here also you see that we are growing more than 12, and generating more than 200 million. So if now I'm looking just at some numbers, we have more than 1,200 customers that generate more than 100K RRR and it's growing. It's going to show you every year how it's growing from this perspective also. I think we should spend a little time on the digital starters because less than 5% of growth is not good, especially when it represents more than 50% of our revenue. So let's have the small going zoom on that and to see what's inside. So on the digital starters, we have three different kind of product. Update cloud is 60%, private cloud is 40%, and web cloud is 90%. So what we will do on that in next years Of course, there was the issues that on support that we need to solve. This is our feedback on our customers. And we will do that. We already started work on the AI, work on the new experience of the support. You probably have seen that on my Twitter. We want to have improved the price performance on the private cloud and public cloud products. So also you probably have seen the last weeks that we launched VPS new range when we started to announce the new product that it's better price performance on this range of product. There was another topic that we, because of the sovereignty market and the business isolation between the U.S., Europe, et cetera, we totally cut access to the web cloud from U.S. customers. So we need to restore that, and we will work on that also. We've been talking about this hosted private cloud in VMware that was bought by Broadcott last year. and we have some impact on this entry level of private cloud, of hosted private cloud. This is why we launched public VCF, but it's not enough. We will launch the new version with the additional features in January and we'll continue to help this part of the customers that they want more managed services to host very small, small scale of cloud. And of course, AI. Okay, now our customers also, you can see that web cloud market, bar metal market, private cloud and public cloud is impacted with AI. We need to innovate. We need to bring to market the new solutions. We already started. And in the next quarter, we will launch a new product specifically on this AI solution. And it's not just that. I just highlighted five small topics that we started to execute. And we have more ideas how to restore this growth in this digital start-up go-to-market for OVH. So on the next page, just to explain why we unified leadership and vision strategy and execution. So it's not enough to have a good vision and good strategy if the execution is not aligned with the vision and the strategy. And the main focus why I was appointed is that we want to be really sure that the vision and the strategy that is decided in board, it's really executed in the small details everywhere and not just on the part of the business. This is why I'm here and I have both jobs to align board and align execution and to bring to OVH the better performance growth in the next year. So also we started to work on this five-year strategic plan, FY26, FY30, to really to explain to our customers, to our teams, to you, to our partners, financial partners, where we want to be in five years. What is our direction? What OVH looks like in five years? And then also explaining the way, the path. that we will use to go over and to execute all the strategies step by step, year by year, quarter by quarter, and to deliver that in the free cash flow positive way of thinking. This is what I've done for 16 years before 2018. This knowledge about having really focusing on the growth, but having focusing on the free cash flow. It's something that is now mandatory for us after this 10 years on the Havia investment, first in the data center, and then in software. And this is why we announced in upcoming investor day, early in 26, we want to also to explain to you in the details what does it mean in the numbers, financial numbers. What does it mean on the growth? What does it mean on the margin? What does it mean on EBITDA? What does it mean on CAPEX? What does it mean on free cash flow? And also, Roche. Roche is something really important that we are focusing on. But also, it's too global. I would like to explain, go deeper, and to show you what does it mean on the infra, what does it mean on software, and how we will improve these financial numbers. Let's talk about the FY26 guidelines. So this is what we announced on the growth, 5% to 7%. This is what we wanted to announce. Of course, we are not happy with that, and we will work on that. Adjusted EBITDA, it will be more than FY26, so it depends on the growth. It will be easier to deliver more once we have more growth. but at least we will deliver more than we had before in 2025. CapEx, 30th of June, too. And of course, leveraged free cash flow positive. And it's not just a little positive. It will not be one million. It will be more. So thank you very much. I will now let Stephanie to present the details about the financial of this year. Thank you very much.

Thank you, Octave. We can already feel the energy today. Thank you all for being with us. I'm Stéphanie Besnier, CFO of OVHcloud. And I will begin with the FY25 results. And to start, we'll look at our performance by product segments in Q4. So first, Web Cloud. We posted revenue of 46.8 million for Q4, up 5.8% life online compared to previous year. The performance is mainly supported by domain names, growing year on year at double digit thanks to the successful rollout of our new offer, the multi-year renewal in several geographies. And as I've said previously, we plan to boost the other segments of this business of web cloud with AI and to expand it abroad. Second, Public Cloud. Our revenue reached 61.9 million in Q4, up 18.1% like for like. We had a strong ARPA growth, and this was also supported by a solid increase in YAS and PaaS offering, including AI. And then on the right side, Private Cloud segment registered revenue of 168 million in Q4, up 4.8% life-alike. Our performance suffered here from a high comparison basis in Q4-24. You'll remember that it was boosted by the price increase of VMware licenses. Now, on the flip side, this new pricing policy impacted significantly our digital starters, on which we plan to focus in the short term. So now, let's have a look at our results. with our key financial figures for the full year 25. On the next slide. So what you can see is that we achieved all our financial targets and we delivered a profitable and cash-generated growth with an organic revenue growth of 9.3%. Second, an adjusted EBITDA margin of 40.4%, up a strong 200 basis points compared with FY24. Third, our adjusted EBDA in million euros reached 435.8 million compared to 301.5 million in 24. Fourth, our capex was 33.3% of our revenue, down 120 base points compared with FY24. And in absolute value, we invested more than last year with 361.4 million euros in FY25. And last, an unlevered free cash flow of 57.6 million euros, which has more than doubled compared to FY24, 2.3 times exactly. So now let's take a closer look at our P&L on the next slide. So in FY25, our profitability significantly improved with a 15% increase of our adjusted EBITDA and a margin of 40.4%. So we had a significant improvement in our EBITDA margin, 200 basis points. And this comes from, first, reduced electricity costs as a percentage of revenue compared with FY24. We are at around 5% of our revenue this year. It was 6% in FY24. And we have also a strong operating leverage, thanks to a higher volume of service produced with contained operating costs. So this strong cost discipline led to a clear improvement in EBIT, which increased to 69.4 million. It represents a margin of 6.4%. at 380 basis points compared to FY24, which is almost twice as much as our EBITDA margin increase. Year-on-year, our EBIT increased by 2.7 times. This demonstrates management focus on driving profitability through the business. Now, below our EBIT, the financial results reached 65.1 million euros. and it includes the fees related to the previous debt for 10.1 million, an increase in our interest rates over the period, and a higher net debt. After including a tax expense of 3.9 million, we recorded a positive net profit of 0.4 million, a significant improvement compared with a net loss of 10.3 million for FY24. So for the first year, a positive net profit. Let's now look at how this increase in profitability translates into cash generation. So, this strong growth in our profitability is also reflected in our gross cash flow from operating activities, which rose from 378 million in FY24 to 422 million in FY25. So after a strong H1 boosted by a failing effect, as you will remember from our discussion in April. The change in operating working capital requirements has now normalized and is slightly positive, amounting to $1 million for the full year FY25. Our CAPEX now amounted to $361 million, so again representing 33.3% of our revenue. We invested 21.4% of our revenue in growth capex, and we invested 11.9% of our revenue in recurring capex. All in all, we are generating a non-level free cash flow of 57.6 million in FY25, 2.3 times, like I said, our FY24 level. Now, our level free cash flow amounted to minus 67 billion, and I can reconfirm to you today that our target for FY26 is to deliver a positive level free cash flow. Let me give you on the next slide a reminder of how flexible our model is with the split of our capex. So during FY25, we reduced the capital intensity of our infrastructure capex, and we significantly optimized the component inventory management. to increase, at the end of the day, the availability of our assembled servers in our data centers. So to achieve a positive level of free cash flow in FY26, we will continue our efforts to optimize inventory management. In FY25, our hardware capex represented 21% of our revenue. So it's five points higher than FY24. Why so? This is linked to the proactive push on the assembled server and to reduce the time to delivery of our servers, but also to prepare the growth for FY26. And we focused, by the way, on the entry-range servers that we mentioned also for the digital starters. As planned, we reduced infrastructure and network capex, which are three points below last year. We focused on the data center occupation rate. That was a 66% at the end of FY25. We continue the usual infrastructure and network work to prepare for the next phases of growth. We have currently around 250 megawatts of installable power capacity, which is a strategic asset for future growth. Then, as planned, our product and software development capex stabilized in absolute value, and it represented 7% of our revenue. We continue, obviously, to develop and enhance our products, particularly in our public cloud offerings and new sovereigns offers. Finally, the other CAPEX declined in FY25 compared to last year. It includes mainly the cost to open new local zones, compliance costs, and the proceeds of a sale of a legacy data center in Paris in H1. On the next slide, let me remind you in details of financial structure. So as you know, in 25, We successfully refinanced with an inaugural bond issuance and the implementation of the first EU taxonomy-aligned green loan by a European cloud player. So we have a solid debt profile with a net debt of just over 1 billion, available liquidity of 242 million, and a controlled leverage ratio of 2.7 times in line with the group's debt policies. At the end of August 24, 25, all the group's debt is edged, and we have an average interest rate of 4.3% over the year 25. So the refinancing was also marked by a diversification of our funding sources. So as you can see on the right, we have no major debt repayment before our fiscal year 2030. And our main sources of financing are now First, 500 million in senior unsecured bonds at a fixed rate of 4.75%, maturing in fiscal year 31. And this inaugural bond has refinanced part of the group's existing debt. It's been rating BB- by S&P and BA3 by Moodle. Second, we have a 450 million green bank loan maturing in fiscal year 2030. a multi-purpose drivable credit facility for 200 million. It's not drawn at all as of today. And it will mature in fiscal year 2030 with an extension option of 1 plus 1. Finally, we have a loan of 200 million euros from the European Investment Bank. I will now hand over to Oktar to talk about our outlook. Thank you.

Perfect. Thank you very much. And just to tease you a little bit for the investor day, we would like to talk about the different topics. The first is go-to-market. You just had the overview of the free go-to-market, but I think there was a topic to go deeper and to really understand how it works, how different it is, what it's experience, and customers, the countries, the products. So we will go deeper inside. The second is public cloud and AI, of course. Public cloud because this is what we started to invest in 2021. Now we have all products. There was also a lot of questions about AI, what we can do, what we do for our customers, where is going the market, investment, not investment, CapEx, a lot of questions I would like to

to highlight the answers from OVH perspective.

Of course, sovereignty market. A lot of opportunities in the, there was a lot to tell you about this market. This if you need specific part of the market. Of course, because we make the CAPEX, we would like to highlight data centers, where we are, what we are doing, how we use our assets and how it will work in the next years. Roche as the results of the investment and also because we will be free cash flow positive for the next five years, what we have in mind for the capital allocation.

So this is the six

topics that we would like to highlight in early 26, and I hope that you will come and spend a little time with us to have a deeper understanding of what we are doing. And now, Stephanie, we can answer your questions if you have some.

If you wish to ask a question, please dial pound key 5 on your telephone keypad or press the blue hand icon. The next question comes from George Webb from Morgan Stanley. Please go ahead.

Hi, morning, Octave and Stephanie. Thanks for taking my questions. I've got a few if I can. Firstly, Octave, maybe two for you. As you kind of step into the CEO seat as well, I presume you're looking at areas around, as you mentioned, products and maybe customer experience that you can sharpen within the business, you can improve the execution around. How confident are you that those things can be achieved without needing incremental margin investment and that you can continue to take up adjusted EBITDA margins? Secondly, you talked about maybe having a public cloud journey in the US. I guess over there, you don't have the data sovereignty differentiation. So when you think about the potential value proposition you could go to customers with in the U.S. around public cloud, would that be predominantly price versus performance, or are there other differentiators you can use to perhaps win against competitors? Maybe one for you, Stephanie, with regards to the outlook for 2026, the 5% to 7% organic growth range is below where you execute for at 7.5%. So there is a further slowdown implied. Can you give any color around how you think about the shape of the organic growth we should be thinking about for the year ahead? Actually, maybe one very last one, maybe back to you, Octave, and not to preempt the investor day in the 2030 potential targets, but you've talked about restoring growth momentum. So I think it's fair to say that your expectation would be that 2026 can represent a trough in growth. Maybe you can start to accelerate you know back towards double digits at least at some point thereafter.

Thank you Thank you, so maybe I will start on the last question of course we are not happy about our guidelines and this is why also in part of the Why I'm here today to lead to execute the strategy what we see is that for the last years, the company was really focusing on the digital scalers and corporate, and didn't invest in the right moment, didn't innovate in the right way in all this digital starters segment of customers. Of course, we have seen that, and you see that because we didn't deliver across the different years. The guidelines, we had the different profit warnings, and this was the part of this story that you didn't have a why, now you have a why because it's clearly explained that. And this is also why talking with Borg, we decided to appoint me because of this experience also across the first years of OVH that we created OVH from nothing. until 400, 500 million, and all this experience of the digital, all the experience about the acquisition, all the experience about growing the customers, premium strategies, going to different geographies, going to different countries, and trying to help the customers starting the journey in OVH. And this part is really important. Even if they are very small customers, this part of the journey is really key because a lot of customers, even corporate, even the very big digital scalers, actually they start testing us being digital starters in the beginning. So all this experience, all this journey in the beginning, this three months, six months, maybe sometimes it's more than one year before they started to trust us and to starting to grow with OVH. It's all part of this digital starter market. So this is where we will focus in this. There was few products that we identified that we are focusing right now because we just have seen that the Revenue about this product just decrease Significantly over the last five years and there my one of my goal is we go back to this Few products and to trying to restore the growth that we should have today But it's not the only part. I think the market changed those new opportunities It was AI is a new way also to be consumed by the customers, the productive customers, but also the way that we can use AI to go faster on the market. And this is also part of the story. We take the time during the investor day to have the highlights about what we are doing internally also with AI. So maybe answering your question about public cloud in US, Yes, the playbook, the way our customers will use us in the US is not the same as Europe. You're right. But we have a lot of customers that actually love to mix the products between them. They love to use private cloud in the same time as public cloud. And being smart between using something that is, for example, bar metal, for the deployment of the platforms, but having at the same time public cloud, so they can make it cheaper, some parts of the cheaper, or they can go faster with the additional 40 products of a managed product. So this is the flexibilities, the range of products that they are looking for. So the playbook is, as usually, it's not just sovereign, it's not just Public Loads is the mix of what we have on the table and what we offer to our customers. It's all the portfolio that our customers love to consume for what is the best for them. Bar Metal, Hosted Private Loads and Public Loads, what is the best for which part is the best. And Mixing DAC, this is where the value is using OVH versus cloud providers where they have just public cloud, okay? And if you want a bare metal, a lot of bare metal, a cheaper one with a lot of bandwidth, with the deploying specific software, actually you cannot do that in the hyperscale. So this is where we see a playbook for OVH in the U.S. It means that, of course, we have two data centers. We see what does it mean having the public cloud in the US, specifically when we think about all this playbook of public cloud. What is this? It's a region. A region on three data centers. This is what we have in Paris. And we will start in the next weeks in Milano, in Italy. And then we have plans to go to Germany, but also in the US, to offer this setup for public cloud. And both will make this growth that we hope that we'll have in the next year in the U.S. specifically. Maybe did I answer all the questions?

It's under seasonality, so thank you, Georges, for your question. What we do see right now is a tougher H1. We expect also to benefit in our H2 from a more favorable, comparable basis. plus all the work that is currently launched on the private cloud, for example, to answer to the concerns of our small customers, the digital starters that are facing the high price increases by Broadcom and generally speaking on all the different product branches. So as of today, a tougher H1 and improving in H2.

That's great. Thank you both. Maybe another question?

The next question comes from Hugo Paternoster from Kepler Shoebrew. Please go ahead.

Yes, good morning. I hope you are hearing well. Great, great. Thank you for the presentation. I will have three questions and the first one is on the public cloud momentum that you had. I guess part of the growth is fueled by AI-related workload. I just wonder if you could perhaps provide us more color on the AI demand that you may see.

how it participates to your growth and how are you seeing it going forward as well in terms of product development internally?

It will be my first question.

Okay, on the public also, we have more than 28% average growth over the last three years. On the AI specifically, we are very opportunistic, okay, on the AI. And let's give me two minutes just to overlay that. you have training and you have inference to another part. We are not part of this training market. It was blessed that 10 potential customers, a lot of capex, there was a lot of uncertainties about the return of the capital. We decided not to play in this part. Okay, so this on the training, this is my answer. On the inference, this is where we play. and I will explain you on the different levels.

First, on the GPU levels.

We don't use just NVIDIA, we use some other GPUs. Why? Because we found that the answer for the speed, for the price, for the performance, for the different agilities, we found some other solutions that it's not NVIDIA. Of course, we have NVIDIA. But on our side, internally, and for our customers in the product that we call Endpoint AI, we don't use NVIDIA only. So this is one part of this inference. Another part of this inference is just offering to our customers the GPUs. And we have seen the evaluation of the demands. where, let's say, one year ago, it was the smaller GPUs, and now we have the bigger systems. And this is where we play, offering to our customers not just the cards, not just GPUs, but the systems, or few systems. For example, the 8 GPUs, 16 GPUs, et cetera. And it is specifically for this market of the sovereign market that we're The customers, they want something that is totally isolated, totally secured, because their data is very sensitive. And we play in the three different locations in this area. We play in our data centers, of course, mainstream offers. You can list the GPUs. We will start very soon on the second new cloud. So it's a sovereign part of the market where we will offer this product early in 2026. We are just finishing the platform for the second new cloud of the public cloud. And then there was a totally new market that we have now the customers, the product and the customer that they're asking us to deploy that on-prem. thanks to our OPCP offers on the cloud platform, that we are able now to bring the software on the hardware and to deploy that wherever the customer wants. So we're playing on this three different go-to markets, let's say, on the AI. The growth on the percentage, there was a part, there was 1%, 1.5% of growth, but Keep in mind that we are not pushing on the CapEx for GPUs in the AI market. Why? Because we don't have the proof that we can make money on that on the long term. The lifecycle of the GPUs, it's so fast. The prices, they are so low that we are not sure that allocating the capex on the high level, lot of money on that, will bring us ROCE, will bring us IR, will probably grow revenue, will grow EBITDA, but it will not be generating cash. So we investing in that because our customers, they ask us to do, and this is where we have the growth, because they want to work specifically with us. But it's really opportunistically done not to lose the customers, but it's not used to win the customers, to have the new customers. We are very defensive on that because of the cycle of life and the capex intensity with the roadster generation. It's something that doesn't work in my mind, so I prefer to go in the very opportunistic in this market.

Okay, very clear. Thank you.

I will have a follow-up question on the guidance. I've understood that you expect a tough H1. Could you be more specific, I would say, in terms of segmentation? What should we expect for H1 and potentially for the full year in terms of private, public, and web cloud?

Well, we We don't disclose in detail the guidance by segments. What we can say is that for public cloud, we have a strong growth, like you can see, and we expect this to keep the momentum. On the private cloud, what was particularly strong this year was the impact of the Broadcom increase. We benefited from it until May 25. Right now, what we do see is the impact on the digital starter customers. We have a decrease of volume and we're working on it with the launch of a dedicated offer. And then on the web cloud, you know that we have also a different kind of market and dynamic on this web cloud. We have benefited from a strong dynamic in domain names so far, so it will probably be a bit tougher in H1 again, and it should also improve at the end of the year with the work that is also done on the portfolio and support, etc.

Okay, got it, got it.

And the last one on the capex allocation, just wonder if you could provide us a little bit of color on the data center strategy. How do you think about the average size of your data center going forward? Will you potentially put more phases on regional data centers, more data centers, and potentially staying? Will you stay on a proprietary basis for those data centers? How should we think about it?

Yeah, okay. It's a very good question because in 2016, I raised $250 million to invest in the data center. It was 10 years ago. And 10 years ago, we invested just for $250 million. We created a large asset of data centers. This is what we are talking about, the visionary decisions and the strategy. It's having the step ahead. into trying to imagine what would be the future and what are the good move today. So we invested in all these assets and we have them. If we would like to have this investment today, probably we would pay 10 times more for exactly the same asset just 10 years after. And this is the reality of the market because AI, power consumption, and all the complexities that you have today to build a data center. So we have these data centers, and it was a very good move from OVH perspective to invest this money in the data center. Now, once I said that, we can use these data centers, existing data centers, in a better way. This is what we started to do last year, FY25. If you see our numbers, you can see that our capex of hardware the percentage of this investment is much higher than the investment in the infrastructure. So the amount of money that we invested in the infrastructure was really lower if you compare to the hardware investment. Why? Because we started to optimize our data center resources and we find a way to host more servers without spending more capex in the infrastructure. So this is also the way that we started to change our investment in not just hardware, but also in the data center. But it's not the end of story. There will be a few additional moves that we will deliver next year to continue to improve our ratio between capex in the hardware and the capex in the infrastructure. Okay, so this will be a part, there will be a lot of improvement in this part, in this specific ratio. Now, if we talk about the future, where is our future in the data center? So, today we have what we call a campus, campuses. That means that we have somewhere, a few buildings, and that they work together, okay? And this is what we call the campuses. We have Graveline, we have Limbo, we have VHS in Montreal, we have the US, etc. We see that it's a very cheap way to deliver the data center resources for our customers because you have a big amount of servers. You can reduce your OPEX and the CAPEX per server. because of the volume, because of the fixed investment, fixed OPEX, and it was a very good way to really start into having the business for over 25 years. Now, we see that the next move for us will be regions, region in three data centers. So we started with Paris, and because we don't know yet what should be the scale of these data centers, should it be small? I would say one megawatt. Middle range, 10 megawatts, or it should be bigger, 40, 50 megawatts each. We don't have answers for that because our journey in public cloud just started, okay? We just, 18 months ago, finished all the range of products and we started just to sell that in Paris and now in Milano. So this is why we leased the data centers in Paris. This is why in Italy we just bought one data center and at least two of them. This is why in Germany, in Berlin, it will be probably the same strategy. It will be the same probably in Ashburn, where we already have one data center and probably at least two others. The goal is really to understand what should be our scale. And then once we know what is the right scale, because we want to spend, we want to allocate the capex in the right way and having the good return, then we will be able to decide what is this move. Should we continue to lease? Should we invest? What size of the data center if you want to invest? So we don't have answers, all answers for that. This is why for the moment we are going in this direction and I hope in the 18 months, two years, will have the clear overview of you about what is the step about the data centers if we talk about this region. And waiting this moment, and also we started to deploy what we call local zone. It's really the first right in the leasing to start a new market and to see if there was any potential growth that we can have in Madrid, in Oslo. in Denver, in Vietnam, etc. So we wanted also to invest very little money to different markets and to see where we can have very easy and very smooth growth without any risk about the capex allocation in the hardware or the network in the different locations.

Okay, thank you very much. Very clear. Thanks a lot for your answer.

Another question?

The next question comes from Daniel Chaffee from Citigroup. Please go ahead.

Hi, good morning. I just wanted to come back to the Broadcom issue. You mentioned that customers are kind of leaving due to the higher Broadcom price increase that happened earlier. What kind of churn do you see there in the moment and was that significantly higher than you initially expected? And now that you're kind of offering those OPCP offerings, what success do you see there? And are these significantly more margin dilutive, the business itself? And then maybe also then on the second note to the guidance for 26, 5% to 7% growth, just wondering, what is your expectation within that on European growth specifically? And then maybe also just to follow up on US accelerating. So this part of the business is continuing to perform very well. Just wondering, how are you seeing the competitor landscape progressing there, like the likes of Akamai or Rackspace. Do you fare well in comparison to them, or are they trying to be also more competitive? And also, do you see a risk of other players entering the space as well? Thank you very much.

Thank you. I will take the first Broadcom OPCP, Stefan, if you can, about the Outlook. So on the Broadcom, you have a way that we see Broadcom is that we have, let's say, two ranges of product. The first is entry level, and then you have the corporate customers. On the entry level, it was really the beginning of story 15 years ago with VMware. We found new customers for VMware because we had this large market of BarMetal. And we offer our customers, instead of managing BarMetal, offering them very good prices about VMware and products that they have fully automated. What is the issue from the last year? The last year, the licensing conditions changed. that the minimum course is very, very technical stuff, but it's the way that the VMware sells the licensing or server, that you need a minimum number of the CPU cores in the CPU that they will bill you. If you have very small servers, in fact, you have a minimum to pay that is quite more than the customers they had today. And so, The increase of the price, it was unjustified for the customer's perspective because I just give you the if they had eight, even they had to pay 16, okay? Why 16? Because it's a minimum of contract that they said this is the minimum that you have to pay. So it's very technical, but it's impacted all this entry level of customers. What was our answer? Our answer was, we can build a product that the customer, they don't have dedicated servers with VMware anymore, but they will be shared across five, 10, 20 different customers. So this is what we call public VCF. And our other product that we started to work, it was about 12 months ago, and now it went live, and we have an additional version, and it's coming on the January. But my point on that is that this is what I call a visionary strategy in execution alignment. We knew two years ago that it would happen. We could have developed all these products two years ago and be ahead on the market. And this is what I want to bring to OVH. Having this anticipation of the issues, having this step ahead about the opportunities and to work today about what the customers will want tomorrow because we started to see the things and the small weak signals that bring us additional information. I will give you another example. This is what you asked about OPCP. OPCP, we didn't talk about OPCP to you guys. We haven't talked at all. I think we didn't announce anything. We started to work on that five years ago. And specifically, two years ago, we started to work on this product with 30 guys. Now we have 100 guys working on that. And we are still going free cash flow positive, and it's totally financed with all our cash that we generate. But what we bring on the market? We bring on the market a new product that allows us to deploy a public cloud wherever the customer they want us to deploy. So we have a new range totally of customers because of this liberation day and also the VMware Broadcom shock wave that you have on-prem a lot of issues. Customers that we didn't have before, they have a lot of issues because there was no more VMware available for them. And they see that having the dependencies on the technology is a risk for them. This is why they started to talk, a lot of them, they started to talk with us about, hey, I want this product in my data center. And it's just right now. It started a few months ago, two months ago, three months ago, five. But we started to work on that two years ago. Okay, and this is what I call step ahead. This is what I call vision, strategy, and execution alignment. And to be really on the market before the market is creating. We want to create the market and not just to be part of the existing market. Okay, so this is one example of the OPCP. And I feel that OPCP will be multi-hundred million revenue in the next years. But because we see that this is where our customers, the potential few discussions that we have, we are talking about 10 million, 20 million per contract. And we have a lot of discussions. When it will come, even a little different quarters, because it's very big contracts, a lot of contracts. Every telco in Europe, they had the issue. Every telco in Europe talks about it. And it's just one example. So this big issue, now we have products to address. We need to transform that to the signature contract, deployment, relationship, and the revenue. And it will come, just not in the next quarter. We see that it will come in the next few quarters. And it will be massive. And this is one part, one example of a product that we just started to develop. develop in the right moment. And this is what I want more for OVH, and this is why I'm here. Stephanie, maybe on the...

Yes, so on the balance between Europe and U.S., U.S., we expect a higher growth from U.S. compared to Europe, and there are different explanations, mostly related to the mix. I mean, in Europe, most of our web cloud business, and you know that it's... single-digit growing market. Most of our web cloud business is in Europe, so clearly you have a mixed effect here. In the U.S., we have almost only private cloud, but still, this being said, you have almost half of our U.S. business that is done with what we call the digital scalers. You have a huge market in the U.S. of this kind of customers. We have a very good relationship with this profile of customers. And here you have also a market that is growing at a higher pace than in Europe. So in the US on top, we intend to expand the portfolio of products. Octave mentioned it. I mean, we are still at the beginning of the story for public cloud, so that will be one of the focus that we will have for the US development. We want also to expand in top rates. That will be another level for growth in the US. So you have different growth drivers plus a very strong local dynamic that will support the growth in the US for FY26 and after.

Perfect. There was 11. Any short question?

other question perfect so maybe the next question comes from enos Mao from BNP PXN please go ahead they can hear me yes perfect I don't know if it's a new thanks for this update I had a few follow-up questions I understand has been a substantial change in It's a paradigm in H2 for private cloud growth, but what has changed exactly in the environment? And that I assume you expect this to persist next year, given the soft indication for H1 2026. Then I have another question on the go-to-market strategy from here. I understand that the core focus on revitalizing a momentum with digital starters, customers, But from a strategic perspective, why are you focusing on them? I know they're quite big currently, but why not on enterprises, given that we're assuming this is the largest output, the largest contract as well? Are you seeing like some kind of bottleneck in terms of competition in the enterprise market? Is it easier for you to actually revitalize digital starters instead? And just one last question. There's a big focus on increasing leveraged free cash flow next year. What's a satisfactory level in your eyes? I know, Octave, you mentioned not a low single digit, but what could be a blue sky scenario typically? Thank you.

Okay, so on the private cloud declaration, I mean, the digital starters represent around 40% of our private cloud revenue. And we mentioned it, we have suffered from the workload optimization in the last quarters. We do not benefit anymore from the positive impact of the pricing trees from Broadcom. But on the flip side, like we said, I mean, we do see the decrease in the volume. So what we are going to do entirely, that's also one of our focus, to reboot the growth. First, we will reposition our entry-range products. In web cloud, in private cloud, in public cloud, but clearly in private cloud as well, we have the refresh of the usual tailored offerings and the work that we've done with the public VCF to offer attractive products after this impact from VMware repricing. We remain very focused to our price performance and here all the work that we are doing on the cost will help us in offering very attractive pricing for this kind of product. We will obviously improve also the support with the AI powered solutions and improve the customer experience on that topic as well. And finally, on the digital experience, globally speaking, from the onboarding of the customer to the billing, we have a plan to improve, again, the journey of the customers, which, again, for digital represents 40% of our private cloud revenue. And your second question, Inès, on private cloud, do you mind repeating?

We didn't get it. Yes, sure. So the second was more, why are you actually targeting digital starters from a strategic perspective? Is it because I mean, I assume enterprises, that's where there is the largest ARPU website, the largest contracts as well. Is it because the competition is kind of like, too high? And then the third question was about levered free cash flow, the blue sky horizon scenario for next year? What's a satisfactory level in your eyes?

On the leverage, it's really positive. It will not be one million plus, just one million. It will be higher. We don't want to give you yet the numbers because it's really a lot of uncertainties on the execution. It really depends on the growth. Today, we see this growth between five and seven. But if we fix early and faster the issues on the digital starters, of course, we'll grow faster. And because of this, let's say, OPEX really focusing on OPEX, we should deliver more. But I don't want yet to give you any numbers. It's too early. And we still have 10 months to execute.

And on the digital starters go-to-market, I think it's not new. We did address it and saw it through the product. Clearly, I mean, again, it's 90% of our customers from the web cloud, which is structurally growing at a lower pace than the rest of the other products. So it impacted. We had this impact from Broadcom. private cloud and the reason why we are looking at it a bit differently that we decided also to address this segment from the customer perspective and to tackle with the product offering but also with the whole customer experience and trying to rework the website, have dedicated actions on the support and at the end of the day, we want to be even closer to those customers that have been the historical customer profile of OVHcloud.

Okay, thank you.

Maybe last question?

The next question comes from Derek Markin from Bernstein. Please go ahead.

Yep, good morning, Octave and Stephanie. I have one question for you because I take your point on product roadmap and the fact that it would be great for your company to come in the market with the right product at the right time. However, I'm really wondering or trying to understand what couldn't be corrected in time by the two previous CEOs because product roadmap is a long, let's say, length, time frame. And so Michel Polin and Benjamin Reclovlich were there. What couldn't they correct at the right time on that topic. Was it too difficult to adapt the product roadmap on time? I'm really struggling to understand if everything was known before, like you said, why didn't you adapt yourself to this market reality? That's the first point or the first question. And the second question is about the CapEx. So I see on slide 18 the reduction in capex linked to infrastructure and network. But when you look to the capex linked to hardware, you see a significant increase in fiscal year 2025 compared to fiscal year 2024, plus 40%, I think, something like that. You said that you were accelerating the production of hardware or the assembly of hardware in fiscal year 2025 to prepare 2026. Look, you don't adapt CAPEX to the current reality of your top line growth. So I'm really here challenging you about your motto, i.e., doing more with what you have already built. We don't see that in the figures, at least for, let's say, for 2026 outlook. So why isn't it more visible? Thank you.

Thank you for the question. A very good question about why we act so late about digital starters. Let's answer directly because we are a public company, okay, and we have the governance. And it's governance, it's not like a private company. You cannot decide the things just because you feel these things. You have to make the decisions based on the real figures. And this is what I also to call, to name a vision. I executed the strategies with vision. That means that because you feel the things, because you have the weak signals, you can make the decisions. Once you're a public company, you have the governance, you have board, you have all the discussions, you have the the quarters, you have the half years, you have the fiscal years, and you have to work with that. And this is what has slowed us down to make the right decisions about this issue that we have seen, we've seen, we felt, I felt a few years ago. And at some point, I want to take my responsibilities and to deliver because I feel that we can fix all these issues and not to have the excuses that's why we didn't do that.

And this is also why I'm here. I'm not here to tell you what, why, when happened.

I want to tell you about how we will fix the issues. This is my goal. This is where we are. I cannot change the past. What I can change, I can change the future. And that's what I will do. On the CAPEX, you write that the CAPEX should be lower this year because of this growth. You're totally right. But what if tomorrow we find what is the issue and we have double digit growth? When we will be able to deliver this growth? because of the servers, supply chain, tensions in the supply chains, because of the AI, et cetera. We are ready to deliver the Groupe, more Groupe that we have today in the guidelines. It depends just on our capacity to fix the issue, the small issues, the issues that we have. If we think that very quickly, will deliver more growth because we have every stuff, everything that customers will want. And we are looking, we are fixing why they don't want it anymore. Let's keep that a very simple way, okay? So we have servers in the data centers that we are ready to sell. You can see that there was no complaints about from coming from the customers this year, from September, about the time that we need to deliver the barometer, for example. And if you compare 12 months ago, we had a lot. So September, October, November, December, until I fixed the issues with the chips. And from January, no more issues about the delivering. We deliver everything. And then something happened in April. We are looking what it is, why, what changed, why the perception, the different things, and we want to fix that. Once we fix that, we will go back to you with the different guidelines. Today, I don't want to deliver you the dreams that we are not able to deliver. I'm here to deliver the expectations that we give you, and this is what we see today. If we see something different, if we see more, and we can see more, we will deliver more because we have all the servers. And this is why we decided to invest this CapEx that will fuel revenue FY26 in FY25. This is what you have seen, that the CapEx on the FY25, they are a little bit higher. Why? This is the root cause, because we have decided to invest in the servers, in our data centers, to be ready to deliver the growth. And now we are working how to make this growth, where was the issue, where it's failed, what was broken, and what is broken, and how to fix that, and then go back to this 10% more growth. And in the board, we thought that I'm probably the better guy to find this as fast as possible and to deliver the growth. So you are right, CapEx, but we are here to grow. It's not just to deliver these guidelines that we have today. But I don't want to overpress.

Understood. Thanks for your transparency, Alfred.

Thank you very much. So maybe we will close this Q&A just with the takeaways. So three things. We deliver F125, all the guidelines, 1 billion revenue, significant milestones with the different areas, and major problematic financial. Unified leadership, CEO chairman. We've been talking about that. And the last one, guidelines. For the moment, we see 5%, 7%. adjusted EBITDA will be more than FY25, at least. CAPEX, maximum 30, 30 is true. And of course, leveraged free cash flow positive. Thank you very much for your time. We appreciate all your questions. And we are here to deliver.

Thank you.

Thank you very much.