Good afternoon. Welcome to Deutsche Telekom's full- year 2025 conference call. As you can see, with me today is our CEO, Tim Höttges, and our CFO, Christian Illek, as usual. They will be accompanied today by our Board Member for Europe, Dominique Leroy; our Board Member for Germany, Rodrigo Diehl; and our Board Member and CEO of T-Systems, Ferri Abolhassan. As usual, Tim will first go through some highlights. He will be followed by Rodrigo, who will dive deeper into the strategy and financials for Germany. After that, Dominique will dive deeper into the European segment before Ferri will update on the outlook for T-Systems. Christian will then complete the picture before some closing remarks from Tim. Then we have time for Q&A.
Before I hand over to Tim, please pay attention to our usual disclaimer, which you will find in the presentation, and please also note that this conference will be recorded and uploaded to the Internet. Now it's my pleasure to hand over to Tim.
Yeah. Thank you, Hannes, welcome, everybody, to our call today. You'll find us in great mood here today because we had a great Q4 , and we have a great 2025 results and a good outlook for 2026. It's gonna be an optimistic call. 2025 was a successful year, and we concluded it with a strong quarter. We wanna give you, as a team, jointly today, an insight into the operations, but even with regard to our Capital Markets Day targets, because it's half-term, and it's time to even, you know, assess how we are looking with regard to the capital markets targets forward. Srini, he has already giving you an update two weeks ago, the team confirmed the Capital Markets Day's ambitions.
They even raised some of the targets, and I'm highly appreciating the reaction from the market on this one. Despite the competition, and all this kind of polycrisis which are around us, we are delivering strong results as promised. We are on track for our CMD targets, not only in the U.S., but also in DT ex US and for the group as a whole. In the last quarter, and I like to reiterate that, our most important financial and commercial KPIs did not only grow year-on-year, but they even accelerated quarter-on-quarter. In every regard, the Q4 , and I'm very happy, and we're gonna talk about that Germany, was, you know, accelerating as well. You know, we had great developments in the Q4 . I like the momentum.
Going forward, I think, the overview shows, you know, the highlights. We delivered 3.8% organic service revenue growth, 4.7% EBITDA growth, EUR 19.5 billion free cash flow, and we grew our adjusted earnings to EUR 2 per share. We had market-leading customer growth in all segments. Our networks remain leading, and we are investing to extend our lead. By the way, in some markets, we have achieved the highest service revenue market share ever, and even in Germany, very close to 40%. We are aggressively driving AI throughout the value chain. We launched Europe's first Industrial AI Cloud. We are the first DAX company and the first multinational telco to reach zero CO2 emissions, and I'm very proud of that.
We increased our stake in T-Mobile to now 52.8%. We returned EUR 6.4 billion to shareholders last year. We announced a record dividend of EUR 1 and another EUR 2 billion share buyback. Last but not least, I'm happy to say that the supervisory board yesterday decided to extend the contracts of our board members, Birgit Bohle, HR; Ferri Abolhassan, T-Systems; and Thorsten Langheim, Group Development and the U.S. within the board. I should say, you know, I'm happy that they extended their commitment to our company and not the other way around, because these are great fellows, now our team is complete. We have the young kids on the block, we have experienced people around us. In this kind of instability of the world, the management of Deutsche Telekom is stable.
As you can see on the next page, all our segments contributed to our 2025 EBITDA growth. After another strong quarter, T-Systems leads the 2025 league table. Congratulations, Ferri! It was Dominique last year. This time it's you, with a growth of 14.4 organic EBITDA growth. With another strong performance from our Europe segment, our ex-U.S. EBITDA grew by 3.4% in 2025. On networks, we keep our market-leading position, which is our foundation of success. In the last 12 months, we passed 3.7 million additional European homes with FTTH. We now reach nearly 24 million homes, of which 12.6 million in Germany alone. In Germany, and Rodrigo will talk about that, we had a record quarter for fiber homes passed and homes connected, and by the way, utilization went up as well.
In the U.S., we now have almost 1 million fiber customers. At their Capital Markets Day, T-Mobile outlined a target of 3 million-4 million fiber customers by 2030. Our mobile networks remain leading across the footprint, we will not stop. In Germany, we decisively won all network tests, our European networks are also leading, I can tell you our competition is very sensitive on that one. You know, this is very funny, you know, to see how they're trying to legally, you know, fight against us when we are claiming that we have the best networks, they seems to be nervous about that. T-Mobile talked about this network leadership in the U.S. as well two weeks ago, we were particularly pleased that T-Mobile, for the first time, topped the J.D.
Power Wireless Network Quality Survey with wins in five out of six regions. This is the core DNA of Deutsche Telekom: always the best mobile network. AI and digital. On page seven, you know, you can see our update on AI and digitization. By the way, this chart's getting busy and busier every day. Today, we have 500 projects running in the company, and to be honest, you know, we cannot put that on one slide. We're gonna have an AI Investors Day coming soon, where we give you the opportunity to really understand not only the projects, but all details and the impact this AI project is gonna have. More to come. Two weeks ago, you already heard about the remarkable progress at T-Mobile.
Over 100 million downloads of their T-Life app and already three-quarters of customer upgrades done digitally. We continue to make excellent progress on this side of the Atlantic as well, and what you can see on this page is really only the tip of the iceberg. Overall, we feel very comfortable, and to be honest, I believe we have to increase the target on this one, but more to come. With regard to our CMD 2024 cost savings target of EUR 800 million from AI and Digital for DT ex US by 2027. I will not go through this slide because Dominique, Rodrigo, and Ferri will cover some of the topics later on as well. Let me just say that we will be hosting on AI, an event.
By the way, for me, the most important thing is that we decided that recently that all employees of Deutsche Telekom are gonna be enabled to use OpenAI ChatGPT Enterprise very soon. There are no excuses anymore for our employees. Everybody is now involved into the AI endeavor within the company. Customer growth. Our market data and customer growth continued on both sides of the Atlantic. Our mobile customer growth remained very strong, with another record quarter in the U.S. In broadband, the European segment delivered strong and steady growth. While full-year broadband subs were negative in Germany, we were able to stabilize our customer base in the Q4 , and I know that most of you have not expected this for the Q4 . Moving on to the capital allocation.
We are delivering steady growth in shareholder returns, including a proposed dividend increase to a record of EUR 1 for 2025, and a third consecutive EUR 2 billion share buyback program. At the Capital Markets Day, we outlined EUR 50 billion of expected surplus by 2027. This surplus is fundamentally intact. It has been partially allocated already. Christian, I will take the money. Is that okay? Good. Our 2026 surplus is effectively committed to be ongoing DT share buybacks and further TMUS share increases, sorry, stake increases, after our decision not to sell TMUS shares alongside their ongoing buyback. Let's review our 2025 guidance achievement. Bottom line is that we delivered on our guidance for 2025 on both sides of the Atlantic. We are slightly outperforming our latest guidance update that we provided at the Q3 stage.
Let me now move to 2026 and the new guidance. Another strong year ahead. As always, our group guidance is based on last year's average dollar exchange ratio. It's a dollar exchange rate of 1.13 in 2025, and is the sum of the guidance for DT ex US and for T-Mobile U.S., adjusted by the expected U.S. GAAP IFRS bridge. T-Mobile U.S. provided its 2026 guidance on February 11th. Outside of the U.S., we guide for an EBITDA of around EUR 15.4 billion, or about 2.5%-3% growth, and a free cash flow of around EUR 3.7 billion, or about 3% growth. For the group, we guide an EBITDA of around EUR 47.4 billion, or about 6% growth, and an acceleration compared to 2026.
Free cash flow of around EUR 19.8 billion, or about EUR 3 billion foreign exchange-adjusted growth, and an adjusted EPS of around EUR 2.20, or about 10% foreign exchange-adjusted growth. As usual, we have a page and appendix in which we compare our guidance with the consensus using a foreign exchange rate of $1.17. On the next page, we provide a midterm assessment of where we stand against the key ambitions of 2023 through 2027, that we stated at the 2024 Capital Markets Day. Our guidance, as shown here, is organic. As you can see, we are on good track for almost all targets. There are two targets where we are lagging: DT ex US service revenues, due to a slower growth in Germany, Rodrigo, and we are also currently behind on our KPI indirect cost as a percentage of service revenues.
This is largely due to the mentioned shortfall in service revenues. We are doubling down on efficiencies to mitigate this. Despite these shortfalls, we remain confident to deliver our Capital Markets Day 2024 guidance for DT ex US adjusted EBITDA, and the growth and the free cashflow which we have laid out. With this positive perspective into the future, let me now hand it over to Rodrigo for a deeper dive into our exciting Germany segment.
Thank you, Tim. Thank you, Tim, and I got the message. Welcome, everybody. I will start with an overview of how we have executed against our strategic targets before I go deeper into the financials, and I will round up later with an overview of our strategic priorities for 2026. We made good progress along our strategic targets. 2025 was an important year for our fiber build. We now reached 12.6 million homes passed in Germany and added twice as much, I repeat, twice as much, many fiber customers as in 2023. We achieved substantial efficiencies in our fiber build, helped by AI. As I mentioned already, at the end of last year, we reconfigured our fiber strategy. More on this later. When it comes to our digital and AI-driven transformation, we see very encouraging early results.
You can see a few examples on this page. Our customer chatbot, now fully LLM-based, achieves a 55% solution rate. 3.4 million contacts were solved by chat or voice bots. We have reached 40% zero-touch automatic call identification. This saves around 30 seconds per call for our agents. We have rolled out AI-based automatic call documentation, which saves another 30-60 seconds at the end of the call for our agents. All this goes to the bottom line with a 10% reduction in call volumes, each call has a cost, that we achieved in 2025. We also made progress with our differentiation and market leadership, a key element of our strategy. We won all, I want to repeat, all survey center, shop, and mobile network tests.
We established Magenta Moments in the market with 5 million active users. Magenta Moments also goes to the bottom line because these users have 3.5 x higher NPS and 20% less churn. Our first-time resolution rate reached 77%. These two drives higher customer satisfaction. Complaints, a number I personally look at every day, were down by 50% in the last two years alone, and our brand recognition has reached the highest ever level. Moving now on this page to our financials. I'm pleased to say that after a weaker Q3 , our headline financials are back on track. Organic revenues were up a strong 2.8% in the Q4 . Two factors contributed to it in roughly equal parts. First, service revenue growth, both in fixed and in mobile.
Total service revenues grew by 1.5% year-on-year. Second, strong revenues with our fiber joint ventures, mainly due to different phasing in both years. These are near zero margin. Year-on-year growth in EBITDA at 2.5% was back at a normal run rate after an unusually weak Q3 . As mentioned, both fixed and mobile service revenues improved sequentially this quarter. In fixed, this quarter, the improvement was mainly the result of easier comps after a Q3 with very challenging comps. Headlines last year were generally impacted by weaker-than-expected broadband and B2B fixed-line revenues. In mobile, the 2.5% year-on-year service growth is a continuation of our ongoing strong performance that we have seen consistently throughout the years. Here we can go a little bit deeper into mobile.
You can see consistent strength in our mobile subscriber growth. I'm particularly happy that the 282,000 net adds we achieved in Q4 2025 are higher than the 261 that we had achieved a year earlier. Our propositions continue to resonate well with our customers, both business and residential. Our network leadership remains undisputed, and we are extending it with our ongoing network modernization program. Our NEMO program, as we call it internally. We have successfully addressed growing data demand with well-designed unlimited propositions and investments. Our unlimited propositions, we have recently updated in a more for more move that we did in February 2026. Our customer segmentation is effective. We're growing in B2B, we're growing in B2C with our main brand, and also congstar is contributing at a steady pace.
We remain comfortable with our CMD guidance of 2%-2.5% mobile service revenue growth in the 2023-2027 period, having delivered growth at the top end of this range in the first two years. If we go now into fixed-line, we are pleased that our broadband net add stabilized last quarter. As Tim already mentioned, we are back to a slightly positive number there. This stabilization is a result of multiple measures taken during last year, including an improved market segmentation, a relentless focus on operations, both on acquisition and customer retention, plus a steady acceleration in fiber connections. It was achieved despite a front book price increase that we did in October, and the reduction of promos, of promotional months, from six to three, that we did earlier in the year in April.
Before I go into fiber, let's look at our good progress on TV. We added over 100,000 new Triple Play TV customers in 2025. This comes after a little over 300,000 in 2024, the year when we had the rights for the European Championship. With that, we added almost 400,000 new Triple Play TV customers since 2023. In addition to this, we added around 200,000 new OTT TV customers in 2025. Almost half of it, 80,000, in the Q4 . With that, we added around half a million OTT TV customers since 2023, on top of the 400,000 Triple Play TV customers I mentioned earlier. This is nearly 1 million new TV customers in two years. Here we are the attacker.
We are already the fastest-growing paid TV provider in the market, our TV attach rate to broadband shows that there's additional potential for growth. If we move to the next page, we talk about broadband. Or I stay on the previous one? It doesn't move, so it's okay. This quarter, we added 164,000 fiber customers. Our best ever quarter, I want to repeat it, our best ever quarter, our fiber penetration reached 16.4%, that's an increase of 11% on our fiber penetration. I can assure you this remains a key area of focus, maybe even our one area of focus, as shown on the following page.
For all of us, it is all hands on deck when it comes to scaling our fiber customers and monetizing the fiber network. In 2025, as mentioned, we increased the fiber footprint by 2.5 million. We had promised that. We remain well on track for our stated 2027 target of around 17.5 million homes passed. As I said, more important, our fiber customer numbers increased by almost 600,000 as we work towards the run rate of 1 million in 2027. As we announced with our Q3 results, to support this goal, we will further step up our fiber investments funded by efficiencies, AI is helping us there, budget reallocations, and the federal tax relief. We are increasing our focus on single dwelling units, on SDUs. We are going more into less dense areas, into rural areas.
That's where the action is. That's where we're going. Within MDUs, we are increasing our focus on connections, and therefore, as the name of the game is, buildings prepared. Not only preparing the building, connecting the entire building, we call it Vollausbau in German or full build-out in English. Doing the Homes Connect and doing the homes activated. Homes activated, paying broadband customers is the KPI in Germany. On top of these infra measures, we are also stepping up our fiber distribution in our channels, and we are seeing positive early results. Finally, on fixed line, let's look at our access revenues. While broadband net adds improved last quarter, our access revenue trends remain impacted by last year's volume losses. Retail broadband revenue growth slowed to 1.6% last quarter. This was driven by volume losses, while ARPA momentum remained positive, especially in B2C.
We see last quarter's growth rate as the low point. B2C, ARPA was up 3.4% year-over-year, driven by speed upselling and disciplined pricing. We reduced, as I said before, the promotional period from six to three months in April. In October, we raised our broadband front book prices by EUR 1 a month, and now, a few weeks ago, we started to raise the back book prices for some of our older cohorts in broadband for EUR 2 a month. In my view, these price increases are long due, as broadband was one of the few consumer categories in Germany to see prices stable or even go down over the past few years of very high inflation. More pricing discipline is needed to fund the massive investments required to build the German digital infrastructure of the future.
While our price increases may impact volumes, overall, we expect various measures to stabilize and improve broadband revenue trends again this year. In sum, we remain committed to our CMD guidance growth of 3%-4%, but of course, from today perspective, this looks challenging. Wholesale revenues improve sequentially. Here, ARPA momentum continues to offset the ongoing volume losses. ARPA momentum will continue to benefit from speed up selling in our growing fiber footprint. However, and please keep in mind, that from the Q2 , 2026, the annual price increases for lower-end DSL tariffs we agreed with our partners in 2021 will roll over. Overall, the 2023-2027 guidance period, we expect to deliver the stable wholesale access revenues that we promised at the 2024 CMD.
Before we get to the financial outlook, let's spend a little bit more time on our priorities for 2026. I already spoke about our industry-leading fiber build, where we are stepping up further. Clearly, here the priority is to keep accelerating our fiber customer growth towards the 1 million target in 2027. I am confident we will deliver this. This will help us both, of course, on the volume and on the value side. Another key focus is to take our digital and AI transformation to the next level. When it comes to leveraging AI for transformation, I believe that there is more potential for efficiency gains than previously thought. I gave some examples, but 10% compounded core reduction per year has a massive efficiency impact, for example.
Especially in the current competitive German market and environment, I see cost and efficiencies as an absolute must and as a hedge against top-line headwinds. As a management team, we are committed to capture them. In 2026, we are working on the full implementation of the message introduced last year, while we have a strong pipeline of further use cases. On the topic of differentiation and market leadership, which is the core of our strength in Germany, we are constantly extending our mobile network leadership through our market-leading network modernization program. We are translating this through our best network campaign into an even stronger brand leadership. Another priority is to evolve our MeinMagenta app into the central operating system for customer interactions, along the lines of what you are seeing from T-Mobile US. Here we copy with pride from our colleagues in the U.S.
As part of this, we aim for 70% of mobile prolongations and additional SIMs to be done through the app. In B2B, we want to extend our differentiation with secure networks, market-leading cloud, IoT, cyber, and AI propositions, the Big Five, as my friend, Ferri, calls them. Talking about Ferri, I'm deeply impressed about the resources and capabilities of T-Systems here in Germany, and I think there is significant potential to capture even more value from increasing our collaboration, and that is exactly what we will do with Ferri. Now, coming to the last page in our financial outlook. On the service revenue side, we are currently tracking below our ambitious CMD target. While mobile is tracking in line, weaker fixed service revenues are waiting, sorry.
The weaker-than-expected fixed service revenues result from 2025 broadband customer losses, as I said, and a weaker-than-expected B2B performance. 2026, we are expecting overall similar total service revenue growth as in 2025. On the EBITDA side, we expect a better trend in 2026 than in 2025, and a return to a more normal year-on-year increase. This improvement will be supported by the rollover of elevated 2025 energy and wage cost headwinds and by the AI-enabled digitization and efficiencies that I mentioned earlier. We continue to strive for our CMD 2024 EBITDA CAGR of 2.5%-3%, given the weaker growth in 2025, we now expect to be at the lower end of that range.
With that, I hand over to someone with whom I've worked closely in Europe over the last couple of years, Dominique, to talk about the European segment.
Thank you. Thank you very much, Rodrigo. you know, moving to our European segment, I'm really happy to confirm that we continued our success story once again in 2025. I will start with sharing a few highlights along our main strategic pillars: growth, transformation, and scale, and what we call win their hearts. On growth, we achieved strong service revenue growth of 3.9% in 2025, which was driven both by B2C and B2B. This success is based on our network leadership, further progress with fixed mobile convergence, and strong B2B growth with mainly a lot of ICT services. We have added 1.1 million fiber homes in 2025, bringing our FTTH number to 11.3 million homes, with an average utilization rate of 36%.
In 5G, our 5G coverage has reached 92% by the end of 2025. Looking at transformation and scale, we are proud of our 73% app penetration and growing chat share. Our transformation towards a more digital sales and service is progressing well. We are also driving AI for network automation, energy savings, and improved customer experience. We have made good progress with scaling platforms across all the countries, developing center of excellence in the B2B area, and deploying a common network operating model. On the last category, win their hearts, we are top-rated in TRI*M customer satisfaction in almost all our markets, both in B2C and B2B. We have 9 million of our customer who signed up to Magenta Moments, and about 45% are actively using it every month.
Next to our customer focus, we are proud to have very high employee satisfaction across all our countries. This success translates into a strong financial performance. In Q4, we deliver another excellent quarter, and this is 32 consecutive quarter of organic EBITDA growth. Eight years in a row where we have been growing every single quarter. I think we can be very proud of that. Organic Q4 revenue growth was 3.5%. Service revenue growth was 4.6%, and this was held by strong B2B IT service revenue, mainly strong growth in Greece. This brought full- year service revenue growth to 3.9%. Our growth was strong in B2C and B2B, and it remains underpinned by continued strong growth in customer numbers.
On EBITDA growth slightly slowed in the last quarter, but remained strong at 3.8% year-on-year, and we ended the year with a strong 5.4% EBITDA growth. The slight sequential slowdown reflects the phase out of previous price increases and the phasing of various one-timers. Our European commercial performance remains consistently strong, and it accelerated in all relevant product categories last quarter. You can see here the last quarter on all the products in a very high Magenta bar. All NatC os are contributing, and I would like to single out Poland, who is doing particularly well in 2025. To continue our success story in 2026, you know, there are several priorities that I would now like to quickly highlight. On growth, we will maintain our strong pace in building the best network.
We will add more than 1 million fiber homes in 2026, while keeping utilization high. We will push to reach 95% 5G coverage. To continue our growth in our core business, we will double down on delivering the best home experience to our customers and further enhance it with smart features around home control and security. At the same time, we will accelerate growth in new business area beyond core. We will leverage Magenta Moments in B2C by launching new propositions like gifting, travel, and dining. This will reinforce customer engagement towards the T-Brand, you can already see the gifting platform live in Czech since last month. We will double down in B2B on monetizing AI and also leveraging digital sovereignty.
On transformation and scale, we will push for even more transactions in our digital channels, aiming for more than 30% e-sales shares and up to 50% of all mobile prolongations and tariff changes on digital. We will step up the AI adoption in sales and services, including in call center and shops, through AI-assisted transactions. We will leverage AI to offer more hyper-personalized and contextualized experience and products to our customers. We will continue to further simplify and retire our legacy system for more efficiency gains. All of this will not only result in better experience for our customer, but also help us to further reduce our IDC-to-service revenue ratio. Win their hearts. For us, it's win the heart of our customers and win the heart of our employees. Going forward, we will build towards a next-gen customer experience, which will be centered around the MeinMagenta app.
We will further drive customer engagement with Magenta Moments and aim to reach 10 million registered numbers already by year-end 2026, and we will continuously double down on improving our customer experience across all domains. Next to our customer focus, we will keep investing into our employees. We will push towards becoming top five employer in telco and ICT, as we want to attract the best talent in the market, with a strong focus on future skills like AI. All of these actions will contribute to securing our number one position in TRI*M. When we now come to our last slides and how do we track towards CMD targets, I think we can say that we are very well on track for our stated CMD targets. Our service revenue and EBITDA growth in the first two years were each well ahead of our CMD targets.
In 2026, we expect further growth in service revenues and an increase in EBITDA to EUR 4.8 billion. We expect underlying EBITDA growth north of the 3% range, close to what we deliver in 2025, adjusted for the tailwind from the end of the Hungarian telco tax. You see, we are still very optimistic to continue on our strong growth trajectory for Europe in the years to come. I will now hand over to Ferri, who will take you through the story for T-Systems. Thank you.
Thank you, Dominique. Compliments to great results. We as T-Systems are learning from you. Let's have a look. T-Systems had another good year. Our order inflow was up 5%. Our customer satisfaction was on an all-time high of 99% TRI*M. Our 3% organic revenue growth was broad-based and ahead of market. We delivered 14.4% EBITDA growth, Tim mentioned that, supported by revenue growth and ongoing efficiency gains. As promised, we achieved a positive and growing cash contribution. We expect healthy order flow and revenue growth to continue, and the first two months confirm that. On the EBITDA front, we expect more modest growth in 2026, reflecting investment into future growth, including in Sovereign Cloud and artificial intelligence. I will come to that. That said, across the board, we remain very well on track for our stated CMD targets.
In 2026, we want to further build on our achievements. First, we want to accelerate our growth with AI. We already have more than 1,500 AI experts, and almost 1,500 AI and data projects are underway or completed. Our AI-related revenues are ahead of plan. We now aim for EUR 200 million in 2026. Together with our brothers and sisters of TDG and of Europe, we are well on our plan for the EUR 700+ that we committed at CMD. Second, we want to leverage our fully Sovereign Cloud. We are since 20 years in the cloud business with more than 7,000 enterprise customers. Analysts rate us as European number-one provider of private cloud and Sovereign Cloud infrastructure services well. Our Sovereign Cloud has functionality and pricing that is competitive with the hyperscalers.
In 2026, we want to increase our T- Cloud Public revenues by about 20% to over EUR 200 million, having built one aligned go-to-market with Rodrigo's team in Germany in one T Cloud tribe. Another priority for 2026 is to step up the cross-sell of cloud into digital and vice versas, so into our base. Third point, we want to further strengthen our focus on verticals, especially public and health, and our newly established defense vertical. Here we aim for more than EUR 100 million of revenues. THRONE will be a major topic of that. Finally, we want to further strengthen our productivity throughout an AI-first strategy and drive further productivity gains in cloud and digital.
We also want to deepen our collaboration with the other B2B segments in the group and strengthen our position as their production machine. We did just sign this week an alignment with T-Mobile US. Next page, please. One key enabler, which by the way shows the power of T and the power of one team, is our Industrial AI Cloud, Europe's first and largest. It single-handedly increases Germany's current AI compute power by 15%. We built this cloud together with NVIDIA and other partners, from idea to launch in less than six months. With this factory, we can provide the complete AI stack for our customers, from connectivity through data center and operations, cloud and compute, security, platform, to AI apps, everything that an industrial or public customer needs.
You can see some of the use cases on this page, and we are seeing strong initial demand from signature clients, and specifically, they use it for applications like digital twins, research, predictive maintenance, or LLMs, such as the European LLM Sophie, that we did together with the Leibniz University and Fraunhofer. Additional demand from signature clients is excellent, and we will expand if we see the market requests it. This project is a good example how T can impact the digital market as one team, and we are proud to be part of that. With that, I hand over to Christian, who will complete today's presentation.
Thanks, Ferri, last in line and short in time. For the sake of completeness, let me go quickly to the T-Mobile U.S. Q4 results, and you know there was a comprehensive report on their results on February 11. For the first time, we had UScellular in there for the full quarter. That led to a service revenue growth of 10.5% and a postpaid revenue growth of almost 14%, and everything was driven by both volume and ARPA. The core EBITDA, according to U.S. GAAP, grew by 6.8%. If you go on to the next page, you see despite a very competitive environment, the Q4 growth was again very strong and was even higher than the previous year.
The net additions on postpaid were 2.4 million, and the best ever, which we had, and postpaid phone net adds were the best Q4 since the Sprint merger, despite the effect that we had an increased churn. For the full- year, T-Mobile delivered almost 8 million postpaid net adds, of which 3.3 were postpaid phone net adds. That will conclude my operational review because everything has been said on February 11. Let's move over to the financials for the full- year. What you see is we have an impact on several dimensions. One is obviously the weaker U.S. dollars, which is obviously hitting revenues, EBITDA, but also net profit. We have T-Mobile's M&A activities.
The largest one is obviously UScellular, don't forget, there were four other transactions which happened in the last year, which was Blis, Vistar, and the two fiber joint ventures, as always, we always have some phasing. Bottom line, what you see in 25, we deliver what we promised in the last guidance, and we slightly exceeded this. On a year-on-year development on this chart, I just want to highlight one thing on net profit. You see there's a negative development of 14%. This is very much driven by the impairment reversals, which we have faced in the year 2024 related to U.S. spectrum, which was more than $2.5 billion, GD Towers and the German Fiber JVs, GlasfaserPlus. On the next page, you see the free cash flow development for the full- year.
All the quarterly results are part of the appendix. There's nothing surprising on the free cash flow bridges. You see there's obviously a contribution from cash flow from operations. We had an increase, which was planned, on the CapEx by EUR 900 million, of which roughly 70% can be attributed to the U.S. Obviously, the overall free cash flow development was impacted by a weaker dollar. On the net net profit, you see that the EBITDA grew by EUR 1.1 billion. If you adjust for the Forex relative to the previous year, that number would have been higher by EUR 1.4 billion. We had a weaker financial result that was driven by two effects.
One is obviously higher interest expenses, and the other one was losses related to the fiber joint ventures in the U.S. Let's go over to the net debt bridge. You see that we had a slight decrease in the net debt bridge, despite the fact that we had M&A activities of close to EUR 9.5 billion, and a very juicy shareholder remuneration, which you see were comprising dividends, share buybacks in the U.S., but also share buyback on our side. Bottom line, what you see is that on the leverage, I'm pretty happy with the developments since 2023, so that we're moving down leverage on a consistent basis. Obviously, the dollar helped. Moving to a review, how do we stand against our ESG targets?
Apologies for the busy chart, you may remember what I said at the Sustainability Day in 2022. We wanna treat the ESG KPIs the same manner as we're treating our financial KPIs. Therefore, you see a lot of numbers. If I'm flashing out two things where we're basically tracking behind target, one is the green PPAs. You know, our target is 50% overall. The other one is the share of female executives. Let me dwell on the PPAs. Actually, we met the 50% target here in Germany, beginning of January of this year. In Europe, we're trailing behind this. This is simply for the sake of we don't find any attractive green PPA, which makes sense from a financial point of view.
I'm not chasing a target if it doesn't make sense from a financial point of view and from an ESG point of view. That basically completes my review for the full -year of 2025, and I hand it over to Tim.
Thank you, Christian. I think before we are going to this summary here slide, I would say let's have the Q&A with Hannes first.
Okay, Tim. We start straight on with the Q&A. Just the usual instructions. I'd like. If you'd like to ask a question via Webex, please press the Raise Hand function. If you require to cancel your question, please press the Raise Hand function again. If you are calling on your telephone, please press star three and unmute by press star six. If you want to cancel your question, please press star three once more. I'll announce your name when it's your turn, and as usual, we would be grateful if you could restrict yourself to two questions and don't forget to mute and unmute yourselves. Let's start. I think the first question is from Carl at Citi. Carl?
That's great. Thank you very much, Hannes. It's one question, but in kind of a few parts about German service revenue trends going forwards. Rodrigo, just picking up on your comment that I think you said in your slides, that you expect similar total service revenue growth in 2026 to 2025. You've just done 1.5% in Q4 and 1.1% for the full- year, but that 1.1% is partially dragged down by Q3 due to the difficult comps that you had in that period. Looking forward, without IT business phasing drags going forwards or that Q3 year-on-year comparator drag, why wouldn't we see a modest improvement? I think consensus is going for 1.3%.
The only thing I can think of is the wholesale price increase annualization in Q2, how material is that? On the positive side, how much of the broadband price increases are you expecting to keep on a net basis? Are you seeing any evidence of the public sector revenues actually coming back in the IT business phase? Thank you.
Sure. Let me go through each of the points. In terms of the service revenue for 2026, let's say we see a similar level to what we saw for the year in 2025. On the broadband side, look, we just did the price increase and the communication to the base a couple of weeks ago. I think it's too early to tell on what is the impact of these moves. I would say that the early results and what we're seeing is positive, is better than what we had planned, but of course, you know, I wanna be cautious on this one and wait a little bit more to see the outcome.
If it continues to play out like we have seen, in the last couple of weeks, we might have a little bit of an upside there. Because this is ultimately a trade-off between volume and value. The value is guaranteed, and on the volume, there's a little bit of a question mark on what would be the impact. What we have seen so far in the last couple of weeks is encouraging. On wholesale, as I mentioned, we are very confident to deliver on the stable guidance we gave for the CMD time period. Please don't forget that the price increases that we negotiated with our wholesale partners back in 2021 will roll over from Q2 2026.
In that sense, we expect to be stable for the time period of the CMD commitment, but probably it was, let's say, more pulled forward from the profile. I hope I answered your question with that one. I don't know if I left anything.
Yeah, there was a question about the public sector.
Yes.
Right.
Thank you. On public sector, well, finally, there is a budget approved that came in at the end of last year. We are now starting to see also increased momentum on order entry. As you know, in public sector, there is a time lag between the order entry and when the revenues materialize. For 2026, I'm increasingly confident on seeing the positive momentum on the order entry. I think that positive momentum to convert into revenues, we're probably thinking more of Q4, and then 2027.
Want to add something to comment?
First of all, the thing that I can add is that again, like in cloud, TDG, and we are working closely together. We have one team, and that has shown good effect. We see in the moment that in elements like the Bundesministerium for digital and so on, so forth, we are making good progress. On the enterprise side, in public, that's good, where we now step in the next step, Rodrigo and I is going also in the mid-market of the public sector. That's next step to come.
Excellent. With that, we move on to the next question from, Polo at UBS. Polo?
Hi. I've just got two questions. The first question is for Rodrigo in terms of German fiber. You've built a fiber footprint of 12.6 million homes, and of that, roughly 8 million covers MDUs. Fiber take-up in MDU areas seems to be less than 10%. Can you talk about why the take-up in MDU areas has been so low, and how do you expect this to evolve going forward? Can you maybe comment on pending changes to legislation that could potentially make it easier for DT to access MDUs and deploy fiber? How much of a game changer would that be? The second question is really just at the group level in terms of use of cash.
Can you talk through why you're not participating in the TMUS buyback, and how did you weigh up buying back more stock at the DT level versus increasing your stake in TMUS? I'm just asking the question because group leverage, including or excluding leases, sorry, is 2.2x . Do you potentially think that this is on the low side? Thanks.
Sure. I start with the fiber question. Indeed, you got your numbers right. From the 12.6 million homes that we have, a big part of it is in covers MDUs. We announced at the end of last year, a shift in our strategy, which has a couple of elements. On the one hand, we will start building more in rural and suburban areas, so our share of SDUs will increase. In our SDU footprint, we are already above 30% penetration. This is a shift that we already started. We are well into implementation, and we're starting to see the results of that.
When it comes to MDUs, we are going for what we call the full build out or Vollausbau in Germany, which means, in the past, when we used to get into a building, we would enter the building, and we would connect the customer that made an order. If a second or third order came later, we would come back and make those installations at a later point in time. What Vollausbau means is the first time we enter the building, we will connect the entire building. We are starting to scale up the capabilities to do this, and this gives us, of course, a significant way then to scale and to sell to that base because every customer can then connect to fiber in a couple of seconds.
We expect to get momentum from that. On top of the Vollausbau, we are scaling sales channels. Most of our shops are in urban areas, we are building a sales force out of the shops that will go into the streets. By the end of the year, it will be roughly 250 employees that will go into the streets that know the neighborhoods, that know the customers, and that will help us also ramp up the commercialization. We already started in several of our shops with very encouraging results. We are also scaling channels, sales channels in our field services. Every day, we have 6,000-7,000 technicians who go to the homes of our customers.
What we have started now is these technicians create a lead. The customer calls immediately, a call center where we do a sale. Just to give you an order of magnitude, we are making 4,000 of these leads every week now. It is scaling, and the conversion rate is 50%, so you can do the math of what this means in terms of sales, channel power. Finally, I think you were spot on. There is, anyway, a structural issue in Germany, which is to access the MDUs. It is not easy to access the MDUs. The cable operators try to protect the, in many cases, that.
We saw last year coming from the ministry, a proposal that I believe goes in the right direction in giving us increased rights to access the MDUs. Now we are waiting for it to turn into reality. It hasn't come out yet. The proposal, we're waiting for it. I believe it is very important that we get that we make it easier for fiber builders to get access to the buildings. In any case, I want to emphasize this, we are seeing really nice take-up in our fiber momentum. Q4 2025 was our best quarter ever in terms of fiber connections. We had a very good start into 2026.
It's all about buildings prepared, full house build, Homes Connect, and homes activated. We look at these KPIs daily. I can tell you, we see them scaling, and so I'm confident that, we will continue scaling, and we will get to the 1 million ambition we put ourselves for 2027.
On the usage of the surplus, Polo, I think I'm repeating myself a bit. Look, we've always looked on two dimensions. One is: what is strengthening our strategic position? The second one is: where do you have the higher financial accretion? I think we have no doubt that buying back DT shares still has a higher accretion relative to the U.S., but it has come down, especially as the share price was at EUR 185. Secondly, what we said is, on the surface, we felt EUR 185 looks cheap, to be honest.
This is why we also said in case this stays the same, we will take the flexibility and actually increase from our side because we believe the underlying value of T-Mobile U.S. is higher than $185 U.S. dollars, and you've seen that the share price has developed quite nicely. Thirdly, increasing the U.S. share gives you more flexibility in then buying back DT shares because that money is out of the door. In case we need some money, we can also kind of decrease the shareholding in the US. From this perspective, I think I'm feeling completely fine with the direction of travel which we're having right now. I think the share price has stabilized. We still have the flexibility to go along in the U.S.
We wanna go up with our shareholding because this is the best market we're operating in by far.
Great. Thank you, Christian. With that, let's move on to Josh at BNP Paribas, please.
Thanks, guys. Two questions from my side. The first is going back to the slide, I think it's slide 14 from Rodrigo, around AI usage. A lot of the AI uses you're talking about here are about deflecting calls and solving negative customer service issues. I wanted to understand whether you're also seeing any upside from targeted advertising, curating tariffs and upselling customers that way. I guess the catalyst for asking the question is, we saw this morning from Freenet that one of your MNO partners appears to be pulling back on third-party sales channels, which typically did that in the past. This is saving that party probably quite a bit of money, and I wondered if this might be something that Deutsche Telekom is also looking at doing more in-house and through third parties as well going forwards.
Then the second question was just around satellite technology and where you see the risks and the opportunities. We've seen a lot of discussion about the Starlink risk in the U.S., perhaps a bit lesser in Europe. When you're thinking about the potential impact of Starlink, both from a fixed broadband and also a direct-to-mobile service, how do you think that will impact in some of the European markets? Specifically, how does it factor into your thinking around the rural fiber to the home rollout, which you articulated during the call? Thank you.
Sure. I can start. I mean, if you look at our cost base, we have the two biggest buckets in sales and service and in, let's say, what we call the technique, which is all the operations around building and operating the networks. We are applying AI in both areas with a lot of success. I focus quite a lot on the sales and service area, but I could spend the same time explaining the use cases that we are applying with a lot of success on the technical side. By the way, these are the savings and the efficiencies that are allowing us to increase our fiber investments into Germany.
By the end of the year, we expect to reach almost 100% of, for example, site supervision of fiber builds via AI. That is generating for us a lot of efficiencies. On the use of AI, Dominique will complement because this is something that, of course, we're driving in Germany, but that we are driving as a group. We call it consumer AI and how to monetize AI, including advertising. The one thing I can tell you is I'm personally getting more and more excited about the role we as telcos have to play in the AI space when it comes to consumers.
We started this journey, a couple of years ago, by bringing AI into our phones, into the T Phone, by bringing AI propositions to our customers. Recently, Srini Gopalan in the U.S. announced how we are now bringing AI capabilities into the network. Why am I optimistic about this one is because we're bringing AI closer and closer to the core of what our business is, which is the network, not only AI, but also security. Dominique, do you want to complement? This is a group thing we're doing.
Yeah. No, no, I'm happy to complement. You know, we use AI a lot. You know, I think on the network side, as Rodrigo said, it was, you know, a lot to make sure that we reduce, you know, the energy consumptions, that we have also a customer-driven network rollout. This is, you know, probably the first place where we used AI in the network. Next to that, we use it a lot in our sales and services. You know, we have one bot, which is, you know, a bot answering customer. We go to more and more one voice element, where we have also an AI voice answering calls of customers. Your question is also very valid.
You know, we have been working really hard over the last one to two years to bring more and more our customer data accessible to really use AI and to drive targeted and personalized, not so much advertising, but very much, you know, experiences, products, upsell, cross-sell activities to our customer base. We have now, you know, more than 100 use case that we have done in all countries, you know, all the European countries and Germany, and we really see a much higher conversion rate when we use these personalized and targeted instruments to reach out to customers. We see lower churn, we see more customer value management, so I think this is also a very promising part where we use AI.
Next to that, you know, we also try to use it to democratize AI towards the customer, and that's more, you know, when we talk about this Magenta AI, where, you know, we did a deal with Perplexity, and through Magenta Moments, we have brought, you know, Perplexity to all our customer. That was also for us a way, not so much to use it, for us to reduce our cost base or to improve customer service on seven days a week, 24 hours, but also to bring it and to democratize it to customers so that we can be seen as a brand that is really innovative and further increase customer engagement. On B2B, we use it a lot. We are trying to monetize AI. We have a lot of activities.
You know, Ferri can talk about it for Germany. We have very good examples in Hungary, in Czech, where we have already developed a very sophisticated AI tool for call center of banks, for, fraud detections, for our customers. I think there as well, we are on a good path to start to monetize AI in the B2B area. Perhaps, Ferri, you want to add anything on that one?
No, I think, I picked it up first and in my presentation altogether, on the B2B side, we aim towards the CMD target of almost EUR 1 billion. We are here ahead of plan. That's good, everybody contributing. More or less, you can say we help customers automating their business throughout AI, you mentioned some of the use cases. I could go on and on and on, I think, it's important to see we attack AI on the consumer and also on the B2B side, concretely with customers.
Maybe just to finish with your question on MVNO partner. Indeed, if you look, if you look at the German market and the pricing we see in some of these online indirect channels, you know, I've been, I think, vocal about it. It doesn't look rational behavior. However, you know, it's always something that you need to manage careful because if other players are investing into these channels, you need to be careful on giving them space there. However, and I want to emphasize this, if you look at our pricing in these indirect channels over the last 12 months, you will see very clearly that we are going in a similar direction.
I think some of the pricings we had for broadband, for example, in indirect channels 12 months ago, were not rational, and that situation has been improving quite significantly. On the satellite question, I think we can answer that one together. For, for Germany, I think recently, the Bundesnetzagentur published the numbers. I think Starlink has slightly north of 100,000 customers in Germany. That number, I think, was around 80,000 one year ago. There seems to be a play, but it seems to be more of a, of a niche play in the case of Germany because we have a very, very good mobile coverage of north of 99% for 5G.
We have strong fixed networks, and so far we see in Germany the play for satellite rather as a, as a niche, let's say, in the, in the few white spots that we have here.
You know, I can perhaps complement on satellite. I think, you know, satellite for us is complementary and very often niche. You know, we use it. We have some satellite agreements, for instance, in the B2B area, to do some backups, which is, you know, for business continuity and emergency cases, I think a good way where we partner with them and we use it. I think also for IoT services, we have some good examples where we have been able to use satellites to promote IoT services, you know, across countries. I would say specifically, you know, if we see more currently where we have some competition from satellite is typically for in country like Greece or Croatia, where there are quite a lot of islands.
There, you know, to be honest, we have upped our game with, you know, a lot more 5G development or rollout, where we are now able to offer fixed wireless access in a lot of those places. Where we have developed fixed wireless access, people really prefer today the fixed wireless access service than the satellite service. We have answers where indeed, you know, it takes some time to bring fiber to remote areas in all the islands. You know, we cannot do that in one or two years, but we have there as well, you know, a very strong mobile network, and we bring C-band and fixed wireless access, which so far has been revealed to be a very strong alternative to satellite growth in those areas.
Thanks, Dominique.
Look, to summarize the satellite service. The first one, in the fixed broadband side, Starlink is the only player who is really relevant in this play, they are playing their game alone. In these areas, we do not see big cannibalization. We have our own product, which we are using, which is high-speed internet in the U.S., or which is our fixed mobile substitution services, which we offer here. That is working. This is an adjacent offer in the market, but we have a good kind of terrestrial service here as well. The second one is D2D, so the direct-to-device. I think it provides complementary coverage in underserved areas. That's a good adjacency. I always talked about network of networks.
You know, I can imagine that we are including this service into our offerings going forward. This is a good kind of comparative service for Europe. It's a small market niche because, you know, the coverage of 5G is very high already. Germany, 99%, just, you know, to give you an impression here. Nevertheless, you know, it might be something which we include in our service. Third one, IoT service, definitely yes. Look, there we have different providers. We will not put our basket into one offer. We will have different satellite service, which can provide us on the IoT side.
The third one is, you know, on the spectrum side, in the U.S., you know, I do not know where these guys are getting spectrum from. They bought own spectrum. Here in Europe, it's a different case. Therefore, you know, we'll see, you know, how they're doing this. I do not, let's say, see us in a role that we are offering and being all our, you know, giving spectrum satellite operators. You know, our policy in the past, and we have been very disciplined in this regard. Why should we do this? In Europe, the situation is that the terrestrial spectrum is protected, so therefore, this spectrum is, by the way, being re-auctioned or extended in 2027 when it comes to the satellite services. This is not in our, you know, decision.
I would expect that the European Commission is extending this for another three to five years when it comes to spectrum. To be honest, that is something which is a speculation. You have to ask the political bodies in this regard. Our terrestrial spectrum is protected, so this is only for terrestrial service, so we do not expect that satellite operators will enter into this space.
Okay. Thanks, Tim, and thanks, Josh, for the question. Next on, I think, is Ulrich at Bernstein. Ulrich?
Yeah, thanks very much. My first question would be on German mobile. You had a strong customer intake there. Could you talk a little bit about the intake ARPU? The reason is that at least two market participants have talked about the lower front book pricing working through the base. What is the Deutsche Telekom perspective on this issue? If I may, a second question would be a bit more conceptual on AI use for the improvement of internal processes and lower cost. I'm not talking as much about revenue here. The question I have is: AI is a technology that is available to all operators, it's not invented by DT, so how would DT gain a competitive advantage from AI? More generally, maybe will AI create winners and losers among European telcos?
If yes, by what mechanism would that happen? Thank you.
Yes. On the first question on mobile intake ARPU, we're very pleased and very comfortable with the ARPUs at which we are bringing our mobile customers. You can see that in our mobile service revenue growth. We not only had more net adds in Q4 2025 versus 2024, but also our mobile service revenue growth was higher in Q4 2025 than it was in 2024. Having said that, if I look now at the market as a whole, I do believe that indeed, the German market is quite aggressive at the moment. We've seen a lot of pricing aggressiveness in the market in 2025.
We have recently, a couple of weeks ago, increased the price in our additional SIM cards, so on the second additional SIM card from EUR 10-EUR 15. We did that in a more for more logic, but nevertheless, it's a quite value accretive move. I do believe that the German market needs to become more rational. On the mobile side, you said it, we are pretty much the only player in this market that is growing revenues and EBITDA. As a market, I don't think that's a good place to be.
I think we had some early positive signals over the last couple of weeks and months that start to go in the right direction, but I think we need to see a little bit more of that into 2026. Our goal is to have a balanced, healthy strategy and growth out of volume and value. What that means for us is that on the volume side, to me, the red line is at least to get our fair share of net adds in the market. On top of that, we need to grow our ARPU because this is a market that doesn't have almost intrinsic growth at the moment. In that sense, we understand our role as market leader and as an incumbent.
Let's see what happens in 2026, but I think at least some of the early signals go in the right direction.
Good. Let me take the question on AI, which is a bit broader. First of all, I would like to differentiate between AI usage for consumer and AI usage for B2B. If you look to the consumer side, we as Telekom have been one of the first that was, by the way, using AI long before the LLM models come up, because we understood that using that towards the customer in chatbots, we did more than 10 years ago. If you take today, Frag Magenta, that was based even on very simple AI tools that we further elaborate on. What does it mean? It means at the end of the day, yes, there is software and there are foundation models that not necessarily come from Europe.
On the other hand side, we as Telekom pick those up as a tool very quickly, and we elaborate that on the usage for our customers. Therefore, we have today, on the chatbot level, I would say, a very advanced, educated system that, by the way, we can also export to others. Also, Rodrigo mentioned that, we picked up the opportunity with AI to further drive automation. When we talk about AI in Technik, that's exactly that. You could say we could do it also with classical tools such as ServiceNow, et cetera. Yes, with AI, you can do it better. Therefore, on the one hand side, AI first, AI and everything is important. That's our philosophy. Also we try to adapt it, and we try to adapt it for our use and the customer.
On the B2B side, it's another animal, I would say. There, you need to combine it with the pain points of customers. Where are they really using automation, and how can you pick AI tools up to make them being faster, more efficient in doing that? In that regard, we have the advantage that we are probably one of the providers that deals with the most Fortune customers, from Shell to Volkswagen and so on and so forth. We work with customers on the automotive side, and we do learn what is their need on automation to build a modern fabric through digital twins. That said, brings us to, for instance, also our chemical pharma customers, where we learn what are they doing for their new pipeline, for their research, et cetera. You can drive it on and on and on.
In total, we are using more than 25 foundation models across the group. We are using a lot of applications, and we brought it, by the way, also back to that stack I was mentioning in Munich, where we combine the complete stack with our connectivity, security, foundation models, and applications. It's on the one hand side, picking it up quickly that we do. It's on the other hand side, adapting to customer needs, that what we do internally and externally. Tim, you might complement.
Yeah. Look, Ulrich, in principle, you're right. The technology is indeed available for all operators and for all telecom companies, you know. By the way, the same is true when it comes to building a network, being it mobile or fixed line. Everybody can build the components. It's all the supplier bases. In principle, it's the same true for car industries because it's an OEM model and you can buy all the components, and everybody's differentiating in the way how he brings the stuff together. Now, going into the specifics and how I see that as a differentiator going forward, I think the industry-specific adaptation is the one which is creating a differentiator. Now, think about, let's say, the, telco-specific AI models.
Who has biggest data and then could use the biggest data in the best way? I think companies with scale have an advantage towards the small companies when it comes to using the tools and using their leverage of AI within the company is bigger for us because, you know, we have a bigger cost base, we have a bigger, you know, industrial base. The moment we have productivity insights, the benefit for us is by scale, by definition bigger. That's the third one. The fourth one is the element of trust. You know, the brand who has the biggest trust, has the biggest, you know, legitimation to bring AI to the customers.
That is what we're doing since years to build this T, including the data and the AI as a kind of trustworthy brand so that people trust in us. The success of Perplexity in Germany, by the way, the second biggest market of Perplexity in the world, just gives you that our customer is confident in this regard. Therefore, I think, you know, if you have a great trust base, you can easily resell the AI models into your client base as well. The next one is the partnership ecosystem. I can tell you, when it comes to the big players of AI, you know, they like to work with the big players because they can scale their products easier than going to all the small players.
That is, by the way, even one of the challenges which I see going forward, that the big players have a big advantage out of AI, while the small companies in all industries, you know, will suffer from late adaptation of it, AI. I see a lot of differentiation possibilities coming from ethics regulation, coming from the partnership ecosystem, coming from the propriety data leverage, coming from the telco-specific knowledge and the AI models, which we can use as a leader in this industry, and we will do everything knowing that the components might be available for all players at the same time.
to be honest, Ulrich, I see us, compared to all the other players, and you know, I'm very keen on this one, I see us leading in this category, in the European landscape at least.
Great, thanks. With that, we move on to Andrew at Goldman's, please.
Thank you. Good afternoon, everyone. I just wanted to push on two of the areas that have been discussed already at reasonable length. I'm very conscious of that. Just on satellite risk and also the AI cost-saving side of things. Just on satellites, the answer's really helpful. I just wanted to look forward a bit more, given there's a lot of your very grand ambitions in terms of pumping satellites into the sky, and just ask: what do you think would have to happen for satellites to shift from being a partner to becoming more of a threat to your business? Tim, you mentioned MVNO contracts.
If they were let in the door, is that letting the wolf in the door, if they get an MVNO contract and start to be able to control the customer? Spectrum allocation, that was also mentioned, is that something that could change, that would enable them to become bigger threats? What are the things that you worry about from this perspective, given that, you know, all of these satellite operators are talking about basically a 10-fold increase in satellites in the sky over the next decade, and that could lead to over a 100-fold increase in capacity? Just a follow-up on that one. You touched on it, but are we correct in thinking that at the margin, the risk is higher to the U.S. than Europe? Because, as you said, the network geographic coverage is not strong there.
just second, and much quicker question, just on AI. We've heard about AI cost savings for a number of years now. Obviously, it's taken a bit of a step up, but in previous years, those AI cost savings that telco operators have talked about didn't really drop through to the bottom line because they just got competed away. What makes you confident that they don't get competed away this time? Tim, I think again, you mentioned that companies with scale have an advantage, and maybe that's the differentiation that means you keep hold of it. What's really shifting? What are the key factors that make you confident that you can actually hold on to AI cost savings here? Thank you.
Look, very clear on, I do not want to make that satellite topic so big here, and I think I was very precise on laying out the picture between Europe and the U.S. with regard to the potential threat there. In this regard, to be honest, due to the coverage in Europe, I see the satellite business model more attractive for the U.S. market than I see that for the European market going forward, including, by the way, spectrum availability.
Second, look, if you know, Starlink would be a global service with X and having local MVNOs, you know, and he can play a terrestrial game because he has to play a terrestrial game, he will never, you know, be able to digest all the traffic via satellites, even if he pulls out millions of satellites there. He needs the terrestrial service. You know, he might have a different model, but then he's in our business model, and then he's in our economics. You know, I don't foresee that right now, that this is gonna happen, at least, you know, from a Deutsche Telekom perspective. The spectrum availability is something where Europe and the U.S. is different as well.
I do not want to repeat myself, because the more we talk about that one, we might be big and make it, and to be honest, we look at it. I see a potential for this satellite market, but I do not see that as a substitution for our industries, just by physics and just by capacity. That is, let's say, the way I'm looking at it. On the AI side, you are spot on. I'm happy, by the way, with the, with our development in the group. I'm happy with, let's say, how the attempts are with regard to the AI projects, five- front project in the organization. I see even the benefits on how quickly things, you know, get resolved in the company due to AI.
Look, I'm a little bit tired about the laziness of my people to reduce their IDCs, especially their headcounts. Therefore, you know, you're spot on. We have to push more on that one. I think the EUR 800 million is committed, but I think there should be more, and I'm looking here to the right side of my colleagues, because they are running the organizations, and they have to come up with something. To be honest, we're working on this one intensively, but I do not want to go today. I promise you, in our AI Day, we will do this in an intense way. We will not only show you one showcase after the next, we will show you how our organization is operating practically. Not in the future, but today.
We will show you the productivity gains which we make in each of the divisions. Which is the impact on the IDC today and going forward, and what is, let's say, the time to market and the other gains we had or have out of AI, challenge us on this one in this specific session.
Maybe just as an additional piece of information, of course, you heard T-Mobile at the 11th of February guiding for their digital and AI-related savings. They guided for $2.7 billion of incremental savings between 2025 and 2027 from such initiatives. Thanks, Andrew, I think we move on with Mathieu. Mathieu, Barclays.
Hello, can you hear me?
Mathieu, we can't hear you.
Hello, can you hear me?
Yes, now we can hear you.
Okay. Thank you. Hi, everyone, thank you for the presentation. I had two questions. The first one on T-Mobile. The guidance, the message for 2026 points to a shift towards a more value focus rather than volume compared to previous years. It seems very sensible. I wanted to understand what opportunities you have there in terms of the value. For instance, does the backbook versus the frontbook pricing enables for more backbook price increases without a risk of spin down or churn? On the frontbook, are you still well positioned versus your peers and hence have a bit of pricing power there? That's the first question. The second was on the exciting topic of ROCE. You're getting ROCE down for 2026.
That comes after a decline in 2025. I think last year you were hoping for a rebound in 2026. you now expect that rebound in 2027. Now, I fully understand that there are non-recurring elements that impact the way you compute ROCE, as you explained in your annual report. I was wondering, besides the one-off, if there are other elements that are driving ROCE down in any of the main geographies, and what makes you confident it can increase in 2027? Thank you.
Mathieu, thank you for the first question. Look, I'm very happy about the developments in the U.S. For us, pricing power is usually associated with the best network and the best customer experience, and you have seen that five out of six, you know, categories we own in the U.S., and I do not foresee on the short term that we will lose them. In addition, you know, we are always leading in value. Therefore, at the recent, you know, update, they highlighted that their backbook is lower than Verizon's on ARPA and lower than AT&T's on ARPU, both by double digits. Yes, there is always potential for us, you know, to do something, and we have not ruled out further rate plan optimizations. I think Srini was very clear on this one.
He said: "You know, we are driving the company by value." Therefore, you know, I think he was very crystal clear. To be honest, due to the fact that Srini has laid out a very detailed plan and a very committed plan, which was highly respected and appreciated from the longs, you know, I would not add something on what he has said. This is now our commitment going forward. We have a great position on this one, and everything, you know, we will do considering what is the best for customers and the best for the company.
I would basically add one point, where Srini was very adamant about, is he said 60% of the customer intake is on premium plans, whereas the installed base is only on 30% on the premium plans.
Right.
That shows you, I think, a very distinct growth platform for ARPA growth, in the U.S.
The second question, Mathieu?
Good question. I owe you the question on the reversal. Look, first of all, in last year, we had quite a bit of one-off effects, which were the impairment reversals, which obviously hit NOPAT, where we came in with a very high ROCE. We don't see this this year. This year we have some, I would say, negative impacts from the integration, which will also hit NOPAT in a negative way. These are the two things which are popping into my mind, I think we can follow up on that question, Mathieu, give you a little bit more detailed feedback.
Yeah, I think you were spot on. The main factor is the restructuring charges that T-Mobile incurs in relation with UScellular integration and also the headcount and network optimization items that they have outlined.
For me, the most important thing is that Deutsche Telekom is, in all businesses, earning their capital costs. That is the most important thing. We are creating added value in this company, and by the way, with an increasing momentum here. I was the inventor of the economic value added, and everybody laughed about me when I did that in 2004 or 2005? You know, a very long time ago. We're still pursuing this logic within the company, and I'm very happy where we are today, that we earn our capital costs, that we pay out our dividend out of our earned money, that we have this kind of sound financial system. That is one of the principles, the narratives on how we are driving this business. Yeah.
Thank you very much.
Thank you, Mathieu, for the question. Next, we move to Paul Sidney from Berenberg, please.
Just building on a couple of questions we've already had. In the German market, I was wondering, did the positive net add performance you saw over Q4 2025 give you the confidence to raise prices in Germany in Q1? Or was the decision to raise prices more a directed strategic move? Just really trying to get a feel for whether the price moves were independent of the performance or whether one's leading the other. Secondly, just building on a previous question on capital allocation, and I know Christian, apologies, you get this question every quarter.
it'd be great to get an update on your capital allocation priorities, and maybe if you could give us a feel for how much unallocated capital there is to allocate over the next few years, in terms of a sense of quantum, and maybe is there a potential to accelerate the fiber build? Again, I know you've had those questions many times before, but it would be great to get an update on your thoughts. Thank you.
Sure. On the first question, I can say it now because now we did it, we started working on this one more than nine months ago. The movements that we were doing on the front book were, for us, a preparation for the back book move. To me, it was pretty obvious from the beginning that this market needs to get a healthier value development on the broadband side. As I said, we first took out promotions in April of last year. We then did a front book increase of 1 EUR in October, which allowed us to test the price elasticity. Something that we did that probably didn't caught your attention is, we did some pilots last year, also on a back book to understand elasticities.
This has been prepared for a while. I want to emphasize that we didn't apply the increases to the entire base. We focused on a portion of the base, particularly on those with older tariffs. We are talking about tariffs that were paying prices before there was a war in Ukraine or before we had a global pandemic. Still, those loyal customers who were on old tariffs are still paying below what are our front book prices. You know, this shows you on how we were very careful also on taking care of these old customers. We have now a couple of weeks of experience.
We're tracking, of course, the churn and the impact of these moves on a daily basis. As I said before, what we're seeing so far in churn is well below what we had in the business case, and we keep tracking that very carefully. I hope I answered your question.
Okay. On the capital allocation, on principle, what do we say at the Capital Markets Day? We said, either share buyback on the DT side or increase shareholding, or we may have some strategic flexibility for other purposes. Bear in mind, we just established in 2025, for example, the AI Gigafactory, which wasn't planned, but we basically used the potential which we have on the balance sheet in order to fund this. Look, and on your question, why now T-Mobile U.S.? I think I answered that question because on the surface, it looked cheap as we made that announcement. For 2026, there's not a lot of wiggle room because we have the $2 billion on the DT side, we have 50% of the share buyback in the U.S., which is roughly $5 billion U.S. dollars.
We're tracking now towards 53%. On the fiber build, whether we accelerate this, I think we announced it in Q3, that we accelerate the fiber investments by EUR 200 million in Germany by reallocating budgets, but also trying to get better efficiency. The third one is by using the surplus, which we're getting from the tax relief here in Germany, in order to refinance or, and reinvest into the fiber build-out. So far, I think I'm fine, and I want to see that the strategy is holding water. What Rodrigo just said, and especially on the MDU space, that we get a higher utilization if we go for a full build-out, because we don't want to build out if there's no demand.
I think this needs to be proven, and then we continue with that going forward.
Great.
Okay. Thank you.
Great. Next up is James at New Street. James?
Yes, thank you, everybody. I had two questions, please. The first one was about line losses for Rodrigo, but not your retail line losses, but your wholesale line losses. Those have been around 400,000 a year for the last three years or so. How do you see that trend going forward? On 1 hand, do you see that some of your wholesale players could move more lines to other Altnets, or because Altnet build is slowing, actually, the rate of those losses might actually slow from here?
Then the second question, may be one for Tim, but as the kind of part of the leading consortium on the AI Gigafactories in Europe, in the largest country in Europe, I presume you're a kind of front runner to be successful in the bid for an AI Gigafactory. I think I asked about this three months ago, but given this is moving at speed, I was just wondering if you could give an update on any further developments on that, you know, particular size of maybe your equity contribution, if you were to be successful, and what new visibility you have, if any, on the revenue models that could underpin the AI Gigafactories. Thank you.
Hi, James. I'm gonna answer your question on the wholesale side. Still, I'm gonna make a comment on retail, which is, You know, we see. I think the last time we talked, I told you my estimation is that this market as a whole, the German market as a whole, is making some 200,000 to 250 ,000 net adds per year, and that the Altnets are taking some 500. If I look into the future, I think those 200 will probably become rather 100, and on the Altnets, we see those 500 stable maybe coming down, but not well below 400. There you have your math.
On the retail side, in the mid and long term, we feel confident to, as I said before, keep our fair share of retail net adds. On the wholesale, it's a little bit of a different story. When you correctly say we've been seeing a similar trend over the last couple of years. Nevertheless, you need to double-click on that one and the average, because when you look into our wholesale losses, there is an increasing portion of that which is related to our JVs.
In our JVs, which are off-balance sheet vehicles, you see the retail net adds that we make on those JVs in our P&L and the numbers that we present, but you do not see the wholesale net adds that those JVs make, on which we 50%, let's say, of the economics we keep, because those wholesale net adds are kept in the JVs. And of course, as we are scaling our JVs, and by the way, we are scaling them successfully, we are on track to our expectations on the JVs, this weighs on our wholesale numbers. If you take this out, I would say that the trend that we see on wholesale broadband losses is stable.
By stable, I mean, along the lines of what we have seen in the last couple of years. The reason for that is because a lot of the overbuild we had and a lot of the exposure we have on the wholesale side is exactly in those rural areas and in that long tail where we have been overbilled and where we have not yet deployed fiber infrastructure. Looking into the future, there is of course, a path to, let's say, a rebound and increase our infrastructure share, including wholesale, because we're seeing increasing signals of market consolidation coming up in the market. We also have 50+ cooperations and growing.
I believe that, over the coming years, you know, we will be part of that consolidation that is gonna be in the market, and that is gonna help us regain the infrastructure share that we are at the moment losing on the wholesale side.
Yeah, look, James, I hope that once you get out of your home office, whenever you see, I see you under your roof, and there's no paintings, nothing. I hope you can still escape there. Anyway, Just was joking. Let me go for your question with regard to the AI Gigafactory, and look, we remain actively involved into the consortium discussions, and we're doing a lot of preparatory work here. However, the project is still subject to a lot of, let's say, alignment, funding, clarity, and industrial governance and structure, which is laying in front of us. Let me maybe frame the first thing.
On the industry side, we are very close and aligned with the Schwarz IT Group, so you will not see, you know, a very competitive landscape fighting for, let's say, different RFQs here from the German angle. The big guys are aligned, and we are joining forces in this regard. The second thing is, with regard to locations, we have alternative locations where we can build it under reasonable terms. That's a discussion which is ongoing, so that's not an obstacle and a problem. The problem starts first with the unknown conditions of this subsidization come from Europe.
By the way, this is affecting all European applications, because it's very unclear what the GPU price per hour is gonna be, under which the European Union is willing to contribute volumes into this Gigafactories. We have heard numbers which are under the price for the GPUs. To be honest, then it will never work because nobody would consider that. Then we heard as well that, you know, these guys might only commit for two years, and I can tell you, nobody will get a financing for an investment like this if you only have a commitment for two years on the volumes which are coming from the governmental services.
There has to be a 35% utilization at least, which was the original commitment from Europe and the member states. I do not see at that point in time how they're getting their act together with regard to the volumes. The last topic, which is a very specific German issue, is the energy price. You are aware that the energy prices for Germany are, you know, not that favorable compared to France or to the northern countries. Therefore, even there is a discussion which is taking place here in Germany. You know our principle here: we are not making bad deals. To be honest, I don't need to build a Gigafactory.
We are willing to build a Gigafactory if it makes commercially sense and we have a good utilization for that one. You know, My statement in the press this morning was: Deutsche Telekom doesn't need it, Germany needs it. If Germany wants to have a kind of sovereignty with regard to their data use and their data handling, they have to get their acts together, and we are the facilitator. We are the enabler for this factory here in the landscape of Germany. That is, let's say, the status quo. I do not know whether we will apply. I do not know the framework for this RFP at that point in time.
Therefore, we cannot say what we're doing, but I can promise you one thing: we will only do that if we have a kind of a clear financing in place, with a business case which is paying off. With regard to the Industrial AI Cloud, you know, these are the 10,000 GPUs. We made a step forward independent from this Gigafactory, and I'd like to hand over to the mother of this factory, which is Ferri, and he give you a very short insight about the utilization, about where we are, and the promising outlook which we see here.
Yeah, there are also a lot of fathers, including yourself. I think you put it very clear under which circumstances we would go for the AI Gigafactory, but we did say we don't wait for that. There is a need for companies like Siemens and others, and they have a demand, and that's why we created in Munich, this first open, secure, and sovereign AI fabric. If you would ask me now, what's at the moment demand, we can say we are ahead of the business plan. We see that almost utilization around 42%, even more so, is given. That shows there is demand, and alongside this demand, we can grow alongside with our customers. We have in Munich the chance to double.
We can even in another location, Munich, triple, that brings us to a capacity around 30-35 GPUs, which shows that's almost a third of what the Gigafactory, by the EU definition, is bringing to the table. We do that alongside the demand that we have in hands, and that connects to what Tim says: we only do what makes commercially sense with us. In Munich, we work with the disadvantage of the energy price, that's what we bring forward to the politicians. We have other things in Munich that makes it a bit more favorable. We have a river close by that helps us cooling. We have a good deal with NVIDIA that allows us to use the newest technology to a very attractive price that we can scale.
Therefore, Munich is, for us, testing the water. Munich is, for us, something that makes commercially sense. Munich is, for us, also the chance to scale, and thereby we see how far through Munich we can come to something that the EU would call Gigafactory and if the three conditions that Tim was making, bringing across for the Gigafactory are coming more clearer, bring then commercially sense to us, then, yes, we can scale Munich into something bigger, like a Gigafactory.
Guys, one last sentence on that one. We are running almost 300 MW of capacity of data centers already today. We have under DTCP, two entities, which we own, which is GreenScale and maincubes. We have Biere, our big, you know, data center, which we own ourselves, which we are utilizing. We have a GPU factory now in Munich. I can tell you, tell me one other telco, maybe the Chinese ones, but tell me one other telco who has this kind of competence of IT systems in the back with this infrastructure competence, which is, you know, deploying AI and data centers at that point in time. We are at a lucky position because our capacity is almost sold out. Yeah?
We are expanding this, you know, within the footprint today. We are thinking about whether we should organize our data center, you know, capabilities in another way in this company going forward. That is more fantasy than, you know, already a business case, but at least it's an opportunities others don't have.
Thank you, guys. I'm sure this was music to Emmet's heart. Next up we got David from Bank of America, then Emmet. Given that we still have four more people who want to ask questions, can you maybe restrict yourself to one question, and we promise to be to answer swiftly. Next up, David.
Yeah, I hope you can hear me, and yeah, guys.
Yes
Thank you for giving us a couple of hours of your time today. I'll get on with this. I did want to challenge this concept of German fiber. You are pivoting towards rural fiber, but you're adding, you know, EUR 200 million from the tax rebate. That really isn't much, quite frankly, against a German, you know, CapEx budget of EUR 4.7 billion-EUR 4.8 billion. Given the difference in cost between rural build, which is probably EUR 1,500 or so, versus urban build of EUR 1,000, I don't see how that touches the sides. What additional CapEx are you bringing from other projects? Is that not a bad thing? I guess my ultimate question is, you mentioned, Christian, the demand curve, but it's so much more than that.
It's about leading, it's about being the incumbent, accelerating copper switch off. Why would you not want to accelerate fiber, even more, given your current opportunity to really dominate the landscape before maybe some of your competitors get refinanced or get their game together? You know, would it really be so bad to take some of that additional capital you clearly have on the balance sheet and reallocate it here? I don't really see why you guys are still hesitating, apart from obviously you have a 2027 guidance target, but is the network incumbency of the network not the absolute priority? Thank you.
On the fiber question, first of all, the number on increased investment that we communicate is net. On top of that, we are reprioritizing within our CapEx envelope. What in the end goes into fiber is quite more than what you see net. This we are doing through a combination of efficiencies and reprioritization within our own envelope. At the moment, we are investing more into fiber than all of our competitors combined, and we are getting our fair share of infrastructure build. Having said that, it is up to me and the team in Germany to now also show and prove that we can fill these networks, that we can drive up the utilization and monetize those networks.
I'm sure that, if we do that, I will be able to go to Tim and Christian and ask for more money.
I like the answer.
Good. Okay.
We change the strategy, and we'll figure out, whether it's working out. I think, David, just to add two things, I think even if we would increase the envelope significantly more, that will not prevent municipalities from building out. I think they will build out in any case. What we're seeing right now, the build-out of private equity-backed Altnets has come to a standstill, almost to a standstill. From this perspective, I think, there are two sorts of competitors: ones you can stop, the other ones you can't stop. I think whenever we are spending more on a certain direction, I wanna see the return.
I think we have had a very good discussion on the strategy shift towards SDUs and MDUs full build-out, but I think we also wanna see whether this yields results.
David, let me just say very clearly, we are absolutely focused, each and every one of our employees in Germany, on filling this network, on scaling the sales channels, on getting the customers on the network, and hopefully, we will earn our right to build out even more. As I said, I see an opportunity to participate in the consolidation that I think inevitably we will see in the German market.
Great.
Thank you.
Okay. With that, we move on to Emmet and his question. Emmet?
Yeah. Hi, everybody. Hope you're all doing well. Thank you for taking my question. I'm just gonna follow up on James' question, please, unsurprisingly, on the AI Gigafactory, Tucherpark in Munich. Tim, you've been very clear in the past that you have the data center exposure through a couple of partners, and you've also said, obviously, rolling out the data centers is extremely capital intensive. We've seen the hyperscalers form kind of joint ventures to roll out some of these data centers. If you look at the agreement you have in the Tucherpark, it looks like a kind of a combination of the biggest German corporate heavyweights. Looks like the Muhammad Ali, the George Foreman and Joe Frazier of German corporates in there.
With this in mind, if the conditions are right, are JVs a way that you can roll out more of this capacity going forward and keep this CapEx off balance sheet?
I think that goes now hand in hand with you, Christian, when it comes to the CapEx, et cetera. Let me say so. In Munich, as I said, we did a good deal, both on, we don't own the data center. We rent it on good terms with a good hosting partner that we have, and we have a good deal also on the hardware side. All the rest comes from us. We can really drive that with a good margin, and we can ramp up the capacity. From that regard, I'm not seeing that we have to engage us into questions like, is that more CapEx? Is that more data center, et cetera?
If we go along, and Tim made the point, let's not forget that we own already a lot of data center, which we also can use. For instance, Munich is a good example. If we have the inference need, we can put another data center, which is close by us, on top of that. Therefore, we go along when we see there is demand. We go along when we see we do margin. We go along with either capacity in terms of data center that we have or through data center capacity that DTCP has and their partner, or data centers where we get a good, attractive price. With this combination, we can scale, and we can adapt alongside customer demand. Does that answer your point?
Yep. All is clear.
Emmet, the answer is very clear, yes, and we are looking into the opportunities here. By the way, within DTCP, GreenScale and maincubes is expanding. They are on the road to expand 4 new data centers. By the way, tenants are the hyperscalers. It's not that we are always competing against them, you know? They are working with us on the utilization as well. Therefore, there is a collaboration taking place as well for European footprint. The demand is high, and yes, we are looking in the opportunities here, how and when we are doing that.
Tim, if I can just follow up, just very, very quickly. Are you finding that German corporates are coming, looking for sovereign solutions at the moment, like pure European German hosting capabilities?
Look, I'm hesitating a bit because there is a kind of sovereignty wokeness here. You know, everybody's talking about sovereignty, but look, beyond, let's say, stating it, you know, I wanna see the clear commitments. We see that in the defense area, there's no question about it. There's a lot of money going in. We are, by the way, with the systems and with DTCP participating on that one as well. Let's, on the cloud side, you know, before you break the learned paradigms of the big corporates, you know, in the way how they're collaborating with hyperscalers, I think it will need a little bit more time before they really shift their data. On the classified side, there's definitely a willingness, but look, most of the data doesn't sit in the, in the clouds already.
That is another, you know, question, how they migrate them perspectively into a Sovereign Cloud environment.
Yeah.
Okay-
on top of that, what we see, like when I left the T-Systems like 10 years ago, the direction was clearly we shift everything of our real-time data into hyperscalers, et cetera. You know, companies like Shell, Volkswagen, they all went that way. That has stopped, and it's on a reverse term. Now, Tim is right, this is not everywhere already on that reverse term, and it's not fully on that reverse term, but it's, at the moment, a hot topic on the market, and it goes. For instance, if health is one of our biggest sectors where we see that at the moment. The questions: are data really leaving the European environment, is now getting hot. No hyperscaler can, architectural-wise, sign that. That brings us into the game, and that's, at the moment, for us, a good USP. But I'm with Tim.
This train has not fully geared up, but it will.
Great. Thank you very much. Thank you.
Great. Okay, two more, one from Robert Grindle on the on the mail. He we didn't manage to connect him. It's a question for Dominique. It's not on data centers. It is about you're running a portfolio of countries, Dominique. What do you think are the main cross-border synergies, if any?
You know, I think cross-border synergies, you know, we see them more and more because the more you go to digital, the more you go to software-defined network, the more you can, you know, bring the network locally and run the software and the intelligence part of it in a central way. This is true for networks. It's also true for TV platforms. We have a common TV platforms across all our country. We have been rolling out, you know, RDK for the routers. We have same routers, so we take, you know, all the routers, set-top box, what we call the CPEs, you know, in one go, and we have the same across all the countries.
That's providing us, of course, synergies and also synergy from procurements, but not only, also synergies in configurations and in managing, you know, the home as I highlighted, as one of the key future development for home experience. These are a few example. The one app or the MeinMagenta app has, you know, the same architecture across the different countries. We are developing currently also one portal on the B2B side. Even in B2B, we see synergies. We have competence centers where we have, you know, center of excellences that are supplying product, experience, people to the different countries so that we do not need to duplicate talents in all the countries. One portal is the same.
We're rolling it out now across all the country where customer can log on to our B2B portal. It will be the same across the different countries. You know, we also use Salesforce across the country. A few examples just to show you that, you know, there are many opportunities for synergy. It is not always easy to implement them because, you know, we are still, you know, companies that have quite some legacy. When we go into new development, into new product, into new platforms, we really try to do it together in a common way.
Currently, also from a cultural perspective, I think people see the advantage of it, trust each other, and it has become much easier to implement and to get the benefits of those cross-border synergies.
Thanks, Dominique. I have a question from Akil at JP Morgan. The question is now on German broadband again, and what he would like to know is where does the improvement come from? Are we taking share from Vodafone, or are we doing also better against other players like the Altnets in this quarter? How is that going to develop going forward?
I think the improvements are coming from really a 360 plan. We have focused a lot on the nuts and bolts of operations. Throughout last year and this year, we, as I mentioned, we are scaling new channels on the acquisition side. We are getting better on the acquisition side. We are also getting better on the retention side. We are running proactive retention campaigns. We have established and upgraded our service desk cooperation, we are starting to also see our churn come down. It's really, you know, the one-on-one of how to pull all the levers of operations.
Where it's coming from, I think it's as I said, as I answered in the question to James, this market is probably making, at the moment, 100,000-200,000 net adds per year. You can see from the Altnets 400-500 that they take per year. Then there is the rest, and I think you can do the math on, you know, how these volumes are moving. In our case, I wanna emphasize this: we have not done this through price aggressiveness. On the contrary, I mentioned all the moves we did on the front book, the moves we are doing on the back book. This is based on hardcore operations, and that's what give me confidence that the worst is behind us.
In 2025, of course, there is now an element of uncertainty, as I mentioned, because of the back book move that we did. The early indicators are very positive, but of course, we prefer to be paranoid about this one and have all hands on deck on operations, and that was what we're doing every day.
Thanks, Rodrigo. Now we have, Tim for closing remarks.
Thanks, folks, for taking the time. It was a pleasure to have you for more than two hours here together and on the questions. It was for me, very relaxing because, you know, I had three colleagues who were answering the question, and I could play with AI. This is, by the way, the picture of which is prompted about your answers. That is about that team. At least, you know, I used the time very productively. Guys, summarize it up. Despite the competition, we are delivering the strong growth we had promised. We are overall on track of our capital market signs.
I want to reiterate that not only in the U.S., but also in DT ex US, and as a group as a whole. You know, half-term, great achievements. You can rely on us when it matters. This afternoon, you heard from my management colleagues about some of the achievement and strategic priorities going forward. A few weeks ago, you heard from T-Mobile, from Srini and his team, the same. Building on that, I'd like to highlight the six super priorities, which are the ones, you know, which are relevant for this year. The first one is doubling down on connectivity as our core differentiator and extending our lead here. I think this is no question. Nobody else can afford it. We discussed, you know, CapEx and the CapEx envelope here.
That is our commitment to you guys. You heard Rodrigo, you heard Dominique on this one. At the same time, we are keen to explore new high-value business opportunities in the data center space. Second, we want to strengthen our B2B portfolio. We haven't talked so much about B2B today. You guys, this is a 15 billion business, so therefore, it's worth looking at that one. We want to do that through an integrated portfolio and pursuing the sovereign AI and cloud opportunities, very laid out. Ferri and the German team and Dominique's B2B troops are aligned on this one to pursue on these opportunities. Thirdly, AI will redefine everything here.
This is something where we will consequently invest, and we will act accordingly to our commitments here, and we will, by the way, lead the industry. We will show you where we stand in our especially AI deep dive workshop offering to you during due course. We are yet to unlock the full power of our transatlantic scale. This is another priority and the collaboration on B2B with the U.S. on the 6G development, on the AI development, on the digitization, on the procurement elements. All of this is something where we have a unique opportunity as Deutsche Telekom, we are making good progress with our collaboration here. We will maintain our successful capital allocation.
I think the moves we made around the T-Mobile Capital Markets Day from both angles of the world were well appreciated. Another proof that we understand what the market is looking for, and hopefully, you appreciate that. This is a lot of work which is going into this one to continually drive superior value and sustainable shareholder return. Yesterday, you know, we even, you know, proved in our board the EUR 1, which is another way going forward to the EUR 1.50, which we have clearly in sight with regard to the dividend payouts going forward. The last thing is, you know, all of this is only power possible with the culture. We always underestimate that, especially in the capital markets.
You guys think everything is a spreadsheet. I can tell you 50% is emotion and heart, therefore, you know, we are investing a lot into our brand, by the way, the most valuable brand of Europe, still, I'm not happy where we are, as well into the best team. You have seen parts of our best team already today. There's a lot of stability in this team. This is good going forward. We have a structural and cultural transformation started here. Not only the rejuvenation of our leaders here, as well in the way how we develop culture into three directions: collaboration, performance orientation, and becoming uncorporate. Less bureaucracy in the way how we operating. More to come.
If you are interested, you know, I can make a special T-Style workshop because, you know, we call our cultural program and the culture in which we are acting, T-Style. With this kind of T-Style, I want to say goodbye. Thank you for this long session today. Thank you, my colleagues, for supporting us today, and thank you all for your trust into T. Bye-bye.
Yeah. Ladies and gentlemen.
Thanks. Bye.
Okay, bye-bye. The conference is about to end. We'd like to thank you for participating at this call and, of course, also thanks to the management for taking the time and your patience. Should you still have further questions, please address them to the Investor Relations department, and speak to you all soon. Goodbye.