Good morning, ladies and gentlemen. I'm happy to welcome you on behalf of the chairs of 1&1 AG, Ralph Dommermuth and Mr. Huhn. The board, in the context of this telephone conference, will present you the results of the first half of the business year 2022, and then give you the forecast for the second half of the year. After the presentation, the board will be available to answer your questions. Thank you, and I'll hand over the stage to Mr. Dommermuth.
Thank you, Mr. Keil. Thank you, ladies and gentlemen, for joining us here at our half-year conference. Mr. Huhn and I split up our parts, that I give you the corporate development 2022, the first half of the year, and inform you of where we are with the build-up of the mobile network so far. Then Mr.
Huhn will present the financial figures for the first half and give you the forecast for the second half of the year. I would like to start with the corporate development in the first semester. Many of you have seen this chart before. It shows where we are active. We have two areas, broadband connections, DSL and FTTH, and mobile contracts. In the broadband connections, we are the largest alternative DSL provider in Germany. We offer VDSL, FTTH, complete packages, including voice and IPTV, from our sister 1&1 Versatel. The last mile, the Regelnetz is provided by Deutsche Telekom or local city carriers. We have a very good net quality. We regularly win the Connect landline test, or we are on the second place. You see the placings in the past years always taking turns with Deutsche Telekom.
In the mobile sector, we have 11.3 million mobile service contracts. We are a mobile virtual network operator. We have a broad market coverage with different brands, most of all one and one, eins und eins, but also co-brands, GMX and Web.de, our sister companies. We have a number of discount brands from our cooperation with Drillisch. Our customers are highly satisfied. This is a logo from Focus Money. Focus Money, together with ServiceValue, did a survey where they looked at how customers recommend brands and one and one made the first place in that survey. You may say it's easy to say, easy to do because there's not so many telecom, telecommunications companies. That's not quite true because over 1,300 companies took part in different industries.
If you look at all of them, 1,355, altogether, 1&1 took second place. The first place was taken by our sister GMX. I think there's no better way to do this. We have a high quota of recommendation. Our brand is very popular. We are currently working on the 5G network. I'll come back to that later. We will stay an MVNO until that network, with our services from Telefónica and Vodafone. Our customer contracts were increased by 120,000- 15.55 million contracts in the first half of the year. 4.71 of these are broadband lines. That means a drop of 70,000 contracts. In the mobile internet range, we have 11.3 million contracts, which is a growth of 190,000 contracts.
I think it's important to mention that in this year, we have the so-called TKG effect, the amendment. Since 2021, we have a new Telecommunications Act in place, and this allows customers to cancel their contracts more quickly as soon as they have left the minimum contract time, which is usually 24 months. Before, customers had a notice of three months to the end of the contract year. Today, they can cancel the contract on a monthly basis. This is why we differentiate the number of cancellations we have operative and how many came about because they are premature, so to say, so that we can compare the years. Here we see that TKG effect. The earlier effectiveness of the cancellations has hit us with 60,000 in the broadband lines and mobile network with 110,000.
Apart from that, I think the performance is quite beautiful. The 10,000 broadband contracts are, of course, 10,000 too many. However, about a year ago, we lost 30,000. We see an improvement in the situation and a mobile internet with the 300,000 contracts that we have operative, we see a beautiful growth. The TKG effect will fade away over the course of the year, but in the first semester, as you see here, it kicked us with 170,000 contracts, where the effectiveness of the cancellation was earlier than in the year before. If we look at the revenue, we were able to increase that by 1.1%. That is because of the other revenues.
For example, smartphone sales with a contract maturity of 24 months in our contracts, that they will be paid back over that course of time. That is volatile depending on the devices of the producers and how aggressive we price them, but there's little margin on that. The margin comes from the service revenue. Here we increase that by 2.6%. As I just said, we lost some DSL customers, and this loss of customers led to the broadband business to have a minus of 0.9%. The mobile business. For the mobile business, this means a growth of 4.7%. The service turnover in mobile services has increased by 4.7%. Looking at the EBITDA, we look at two different segments. One is the access segment. That's our traditional segment that we have always been doing.
In this segment, we were able to increase the revenue, the EBITDA by 9.5% to EUR 333 million, EUR 380 million. In the mobile networks, we are also increasing it. We had startup cost of EUR 16.1 million earlier in the year to further buildup of the network. That was a bit more than the year before. The increase 9.5% nicely show that we are well able to make the contract profitable that we have and keep and develop customers that we have developed once. Now, we can move on to the mobile network of United Internet. We've started a first friendly user test, and the focus is on the first test here. We are right at the beginning.
However, I would like to show you some results of that test already. The first thing is the question of the latency times. How quick do the servers respond? In the test, we have latency times of 3 milliseconds. If we have applications in our far edge computing centers. Here you see a multiplayer game, Minecraft, and here we got 3 milliseconds of latency times. That is close to what is physically possible. Parallel to that, we looked at the bandwidth. Here we have bandwidth a bit above 1 Gb. I'm probably gonna say others have done that as well, but you have to see what frequency you have available. Compared to Telefónica, Vodafone or Deutsche Telekom, we have quite low frequency bands. Despite of that, we allow for gigabit bandwidth.
That was an important goal for us because I think in marketing later on, that is a speed we'll have to deliver and that we cannot stay behind. That worked in the friendly user test as well. It's also important to look at the stability. Are we able to keep our nets so stable that the data transfer and telephone calls work consistently? Here we are doing quite well as well. Within 24 hours, we could download over eight terabytes to the customers. That is all quite nicely looking of what we had theoretically planned at one point when we decided to take that Open RAN architecture. It's proven now that this work in Karlsruhe, Mainz and Frankfurt. Quite clearly, we have to develop this, scale it up, build it up.
What we can say as of now, that the technology works quite well so far and fully fulfills our expectations. Moving on to the next steps. In the next quarter, we are going to see an extension of the Internet connections. We are looking at 36 additional pairings. In the first quarter, we want to offer the first product, fixed wireless access. That's what we have tested. For that, we're going to have 140 regional edge data centers and 24 decentralized edge data centers and 2 core data centers. Next year, we will look at the connection of the network with other telephone networks. First, for telecom, we're gonna start with national networks first, and the second quarter, we're gonna connect to international telecom networks, and that's also gonna see the start of national roaming.
As a test phase with Telefónica, we agreed that in the third quarter we're gonna provide national roaming and commission it. We want to start testing it in the second quarter. In the third quarter, together with the availability of national roamings, we want to start on a federal level. That's the point when we can offer mobile products. The fixed wireless access is not mobile without national roaming. The customers who leave Drillisch will not be able to provide it anymore. That is why it's very important for us to get national roaming and get it in time as well, so that in quarter three we can start in with that. At the same time, we're gonna start with the sales of our MVNO products. Discontinue these, and then we're gonna start to migrate the existing customers.
It's 11 million, and we took a time span of two years to migrate these customers to our network. That's quite a number of steps that we are facing, but I'm confident that we will be able to manage. We are proceeding quite well. Of course, there is a day or two where you think, well, could have gone better, but in general, we are proceeding well, and I'm optimistic that next year in August, we will be able to commission the network. Thank you for your attention so far. Now I'd like to hand over to Mr. Huhn for the financial figures.
Well, thank you very much, Mr. Dommermuth. Hello again. I'd like to start with slide 15, the earnings for the first half of the year 2022. Dommermuth already spoke about the revenues. The cost of sales increased only slightly in the first half of this year compared to the first half of the last year to EUR 1.317 billion. The gross profits from turnover went from EUR 616- EUR 634. What's important is the out-of-period revenue that we generated in 2021.
If you eliminate those and compare the adjusted result, we are at EUR 577 million compared to the EUR 634 million of the first year of 2022, and that would be a growth of 10%. The distribution costs due to higher expenses for marketing of EUR 235 million increased to EUR 248 million. Administrative costs increased slightly from EUR 59 million- EUR 60.2 million. This includes the expenses for 5G, which increased by EUR 1.2 million. If you eliminate those, then the administrative costs actually decrease slightly compared to last year. What significantly increased are the other operating income and expenses from EUR 8.7 million- EUR 11 million from debt collection costs.
We have a position of depreciation impairments on receivables and contract assets, which increased from EUR 34 million- EUR 49.2 million. This increase of EUR 14.8 million results, on the one hand, from higher income from reminder and debt collection fees and also the effect of the TKG, the Telecommunications Act. There is a higher blocking limit which means that we disconnect our customers later. That leads to higher cumulative debts that has a total effect of EUR 5.4 million and has an impact on revenues, of course, and another EUR 5 million results from higher return rates and higher debt collection rates.
During the pandemic, we saw that the debt collection was much better, or payment by customers was better than prior to Corona, but we're slipping back to the same level we were pre-Corona. In the first half of the year, we saw an increase compared to last year. The financial result in the first half of the year was at EUR -3.2 million. EUR 3.2 million resulting from investments in white spots. We have committed to EUR 50 million worth of investments. By contrast, we don't need to pay the frequency charges for 12 years. Without this, we would have a balanced financial result. Profit before taxes was EUR 284.5 million. Last year was EUR 296.1 million.
That includes, of course, the EUR 39.4 million out of period revenue. If you eliminate that, we're at EUR 257.7 million compared to the EUR 284.5, which is a growth of 10%. Tax expense decreased slightly from EUR 88.9 million- EUR 85.6 million. The consolidated result was at EUR 99 million compared to EUR 207.2 million last year. Let's now move on to the balance sheet. Balance sheet sum increased by EUR 344.5 million- EUR 7.063 billion. The short-term liabilities increased to EUR 7.063 billion. EUR 45 million result from the increase of the active write-offs.
Another 30 million result from higher demands from reserves. Long-term assets increased due to the fact that contingent down payments. On the liabilities, the short-term liabilities increased by EUR 166 million due to liabilities against consolidated companies and liabilities from services. Long-term liabilities slightly decreased, and equity increased by EUR 199 million in the course of the first half of the year. On the next slide, you can see the development of the cash flow in the first half of the year. Net inflow of funds from operating activities was at EUR 238 million, composed of EUR 273.2 million cash flow from operating activities, minus EUR 1 million from change contract assets and contract liabilities.
What had a positive effect with EUR 171.9 million were changes in receivables from and liabilities to related companies. EUR 31.1 million positive effect came from an increase in income tax. A negative impact with EUR 61.6 million was the change in trade and other receivables and trade and other payables. Further negative impacts at EUR 157.1 million were change in accrued expenses. That is the down payments with the contracts with 1&1 Versatel and Deutsche Telekom. The change of inventories had a negative impact of EUR 18.3 million. Cash flow from investment activities was at EUR 220 million, EUR 51.3 million resulting from CapEx and EUR 169 million from investment of free cash.
Cash flow from financing activities was slightly above last year. We're at EUR 16.1 million. EUR 7.2 million were repayment of lease liabilities and eight point eight million resulting from dividend payments. The free cash flow was at EUR 186.8 million after EUR 177 million last year. On the next slide, we have shown the bridge for the first half year from EBITDA to free cash flow. You can see the EUR 368.2 million EBITDA in the first half of the year of 2022 against the negative effect of accrued expenses of EUR 157.1 million. Then also minus EUR 37.8 million were due to trade and other payables and other liabilities.
A positive effect, EUR 171.9 million from liabilities against related companies. The balance from other working capital had a negative impact of EUR 44.7 million. Taxes with minus EUR 62.4 million, CapEx minus EUR 51.3 million. We arrive at a free cash flow of EUR 186.8 million. Now let's move on to the forecast for the business year of 2022. We confirm the forecast so far with a service revenue of about EUR 3.2 billion after EUR 3.1 billion last year. The EBITDA we expect to be at previous year's level at EUR 672 million. What's important here is that the forecast includes EUR 70 million of costs for building the mobile network.
This was at EUR 38 million last year. We expect for this year 450,000 new contracts. The operating growth with contracts we expect to be 650,000 contracts compared to 600,000 last year, minus the 200,000 contracts shifts due to the recent Telecommunications Act amendment. CapEx, we still expect EUR 400 million for this year after EUR 37 million last year. The CapEx increase is mainly due to our investments into the mobile network for antennas, computers and software. Maybe one more.
Point before we speak about or start with the Q&A. Mr. Dommermuth had mentioned already the buildup of the mobile network. We would like to give more detail on this in Q4. Now, I'd like to hand back to the operator for the Q&A session.
Thank you. The Q&A session will be done electronically. If you want to ask a question, please press the asterisk button and key one on your phone. Please check whether you are muted or not so that you can be heard. The questions will be answered in the sequence that they are signaled. We are going to answer as many questions as the time allows. If you find that your question was already answered, you can retract your question by pressing the asterisk button plus the two on your keypad.
Asterisk one to ask a question, asterisk two to retract the question. We're going to wait for a moment to give opportunity to everybody to signal their question. We start with Martin Hammerschmidt, Citibank.
Yes. I have two questions. One is concerning the drop of customers, 170 in the first half of the year due to TKG, with a minus of 200,000 for the overall year. Sounds a bit optimistic. How can you help us to understand what is going to change in the second half of the year to see that improvement, so that we shouldn't take that 400,000 minus? Last year, Telefónica, we heard that there's gonna be price increases during the course of the year.
If other players will do the same, would you follow these steps, get the more for more price increases, and if not, why not? Thank you.
Yes. Please let me start with the TKG effect. You're quite right. If we look at the 170,000 in the first quarter and then the forecast that we gave initially and that we have renewed today with 200,000 for the second half, then it looks as if it's a bit asymmetric. Actually, we do expect the second half to see less of TKG effects. Whether we will really end up with 200,000, I can't say, but the net growth of 400,000 will be delivered. That's a fixed goal, and we do assume that that's gonna work. You asked for the price increases that Telefónica did with more for more.
Difficult to comment because I don't know what's gonna happen when, what dimensions that's gonna have, whether that's O2 or Blau or whether it's in the service provider business with mobilcom-debitel or whatever. That is difficult to comment at this point. We usually follow the market for our navigation, and if we see opportunity to implement higher prices, we will take opportunity of that. We have no concrete plan at the moment, but we are observing the market conditions. Of course, we have to see how the inflation is gonna develop over the next months.
Thank you. If I may come back to Mr. Huhn's statement of the second quarter, the CMD, could you just give us a little flavor of what is to expect, what the goals are?
Well, we are going to give more information on the architecture and the partner landscape and the operations model that we have established to build up that network in order to give a better overview of the technologies, the architecture, and how we're gonna cooperate with the partners. What financials are concerned, we will give an overview of the investments and cash that we plan for the next coming years. Of course, we are going to look at the KPIs that may be of relevance for the 1&1 mobile network and give an overview of what that will mean for the next years and how that'll affect P&L.
Great. Thank you.
We can now take our next question from Yemi Falana, Goldman Sachs. Please go ahead.
The next question from Yemi Falana, Goldman Sachs.
Afternoon, everyone. Yes, thanks for the intro. Yes, it's Yemi from Goldman. Two questions from me. Firstly, on network rollout, maybe a multi-part question. Clearly, you've reaffirmed the EUR 400 million CapEx guide for this year, but some of your European peers have seen rising investment costs as they roll out network. Dish CEO Charlie Ergen has been flagging that their network build-out in the U.S. is going over budget to some extent. As you get deeper into that network build-out process, are you seeing higher cost prices relative to when you initially made your plans? That's the first part. Secondly, do you continue to expect to pay for that entire rollout out of organic free cash flow? Maybe finally on the network.
Rollout side beyond your existing agreements with telcos. Could you provide some color on how actively you plan on leveraging the various telcos that now exist in the German market? Maybe if I could sneak one more in on, just on the margin profile, because one aspect of the network rollout rationale was ultimately to achieve higher margins as a network owner. Do you still expect rising margins from the end of 2023 onwards as you begin that customer migration period? That'd be great to know. Cheers.
With respect to the first question, as far as our assumptions for the business case are concerned, CapEx and OpEx, we are still on the level that we had expected before. We have had a large number of contracts that exclude the inflation risk that we have made agreements on the price increases for the next years. These EUR 400 million that we have given us guidance for this year haven't increased due to these effects, but they are moving in the framework of what we have calculated in the initial case. With respect to the funding, the plan is that over the next years, we will look at the financial needs for the investments and the build-up of the network to fund that from our cash flow and the credit or the positive that we have at the moment.
If we have clearance on the low band frequencies and the funding needs resulting from that, we will decide on how to gear up and fund ourselves medium and long term. As far as the margins are concerned, we are expecting levels of EBIT margins that are going to have a strain because we will make the investments in the network, which is not under full capacity at the beginning. Our expectation, however, is that if we have built up the 50% and the customers are migrated and we generate the traffic coverage that we have calculated, that we'll have an EBIT margin situation which is quite comparable to today's margin.
That's very clear for the most part. Could I just make one clarification just on the spectrum frequency point? I think you mentioned that the plan is to fund via organic free cash flow, but dependent on some factors around potential spectrum, you'd have to reconsider funding arrangements. Could you maybe clarify what you meant there?
We can't comment on that at this point. That is a bit of speculation, and we cannot do this at this point of time to give a figure for the value.
Okay. Thank you very much.
We'll now take our next question from Joshua Mills of BNP Paribas.
BNP Paribas for Joshua Mills, BNP Paribas.
Hi, guys. Thanks for taking the question. I have two, please. It wasn't quite clear from the translation, but I think you were talking about the fact that bad debt provisions had increased this quarter. Could you just clarify whether that's the case and also what's driving it? I'd be very interested to know whether you're seeing any impact on the brand mix, maybe people shifting towards the lower ends of your tariff range as a result of the macro backdrop. Any commentary you can give there would be great. Secondly, coming back to the net add number, it sounds like you're very committed to hitting the guidance.
If I look at the mobile and broadband trends for the first half and even adjust with 170,000 that you attribute to the German telco law, it does look like the run rate in mobile is lower than in the past. Then in fixed line, you're continuing to lose subscribers. Is there a brand or tariff reset that you're anticipating which could help to re-accelerate those subscriber growth trends? Or are you pinning all of the hope on the fact that once the regulatory effects wash out, you'll have much better underlying growth? Be very interested to hear on that. Thank you.
With respect to the first question, as far as the drop in credits is concerned, the effects from these are not a shift to the contract between the different brands. The major reason for this is an increase of the dropout quota in the post-corona phase. During corona, the dropout curves have declined with all brands. The default quota are at the level of prior corona at the moment. The other effect is due to the TKG.
Regulations. I have pointed out that the limit value is increased. It was at EUR 75 before, and it is over EUR 100 now. This is why the non-paid amounts increase before we can block customers and hand that over to Inkasso offices. That hasn't got to do with more discount contracts now real compared to premium brands. The same applies for the TKG effect, the 170,000 that we had in effect of the in the first half of the year. This effect is mainly in the first half of the year because not all customers had 12 months maturity, but maybe six to eight months only, and then they could have left the company. That effect is stronger in the first half of the year than it will be in the second.
As Mr. Dommermuth had said, for the second half, we expect a lower effect of about 30,000 contracts. The split out in the brand world corresponds to what we had seen or expected in the beginning of the year.
Okay, great. I mean, maybe to ask it another way. We've seen your broadband net add trends get progressively worse for several quarters now. I know that the message that you give is: We're happy to see or we're relaxed about seeing some of these customers leave. They're lower value. We're waiting for new wholesale contracts and better terms to upsell people to faster speeds. Could you just give us a timeline on when we might be able to expect to see that broadband net add number stabilize? Or should we assume that the negative 30,000, 40,000, 50,000 run rate each quarter will continue in the second half as well? Thanks.
We've seen, as we have explained before in our discussions also over the last few months, that due to the extension of FTTH coverage, but particularly by Deutsche Telekom, we see that a new momentum will build up in the market there, thus much higher opportunities in marketing. Now, how fast this can be leveraged, this potential, is difficult to say at this point. Of course, we started marketing this, but at the same time, the areas where we can market are only emerging at this stage. We believe that we will have to wait for developments over the next year to see how coverage expands and what sales opportunities result from this. Stabilization and improvement is something that we expect for next year. We can see it in the operative figures of this year.
In the first half of the year, we had the -10,000 mentioned by Mr. Dommermuth for the first half year. Before that, we had -30,000. We believe that next year through the fiber-optic changes, we will see positive trends going forward.
We can now take our next question from Polo Tang of UBS. Please go ahead.
Thanks for taking the questions. My first question is, can you maybe just talk about what you're currently seeing in terms of your mobile and broadband net adds for Q3 so far? Can you talk about the competitive dynamics in the market? My second question is just really about your coverage obligations and the status of your network build out. Can you clarify if you're gonna be able to activate 1,000 cell sites before the end of 2022? If you're launching a new service in Q3 2023, what level of population coverage do you expect to have achieved by that point? My third question is really just about Huawei. We obviously have reports in Handelsblatt suggesting that the government could push for greater restrictions on the use of Huawei equipment. What's your view on this?
Does it impact your mobile network build out in any way? Thanks.
Yeah. Maybe I'll start with the last question. In our network, we're not planning to install any Chinese hardware or any software from China, and therefore, this discussion about Huawei doesn't affect us at all. Concerning the marketing situation and growth in Q3, we can say that right now we are stable. The sales figures are developing seasonally as expected. We are optimistic that in the context, we're in the framework of or in the range of the forecast that we gave. I don't want to give you any details on where we'll wind up at the end of Q3, but the stable trend, the continuation of what we saw in Q3 is what we expect for Q3 as well. Now, concerning the question on the coverage obligations.
On the one hand, we have the obligation to build 1,000 antennas by the end of 2022. Which we still believe is a realistic goal, and the next goal would be then 2025 to have 25% pop coverage. The extension here is mostly planned for urban areas, obviously. As we had communicated in our last call, our objective is, of course, to roll out the build-up much faster than the obligations require us to.
Can I just clarify where you're gonna be with your population coverage at Q3, in 2023 when you intend to launch?
We can't say right now. It depends on how the ramp-up of customers and the migration beginning 2023 will pan out. Traffic coverage in Q3 and Q4 of 2023 will be still rather limited because we're only beginning migrating and bringing customers to our own network, and it also depends on how many antennas are live already at that time. That's very difficult to predict at this stage, but it will certainly be a very low level initially.
Okay, thanks.
We can now take our next question from James Ratzer of New Street Research. Please go ahead.
Yes, good afternoon. I'd two questions, please. The first one, just regarding your service revenue trends. I mean, for up to Q1, those for the past three or four quarters have been running at kind of 3%-4% per annum. It was 3.5% in Q1, and then in Q2, that slowed to 1.7% in terms of the service revenue growth trend. I was wondering if you could just talk a little bit about what's changed between Q1 and Q2, on the service revenue trends specifically, please. Secondly, would just love to get an update from you on your discussions with the regulator and what you think is likely with regard to the 800 MHz auction renewal that's been scheduled for 2025.
I know some of your competitors are lobbying for that to be delayed until at least 2030. I was wondering if you could give us an update, please, on what you're expecting for that spectrum renewal process? Thank you.
Yeah. Well, we don't know today how the regulators will decide today and what form of frequency allocation will take. You're right. Our competitors are lobbying hard to extend licenses. This is, of course, runs counter to our interest because the frequencies belong to the Federal Republic of Germany, and the utilization rights end as per 31st of December 2025 with the 800 MHz frequencies. After that they have to be redistributed in a fair and equitable way. Again, that's what we read, what we hear. There is a process ongoing where the Federal Network Agency with involving the stakeholders is determining what can be done, and you can speak your mind there and let them know what you think.
We did, and what we hear from the Federal Network Agency and from the Federal Ministry for Digital and Transport is that allocation will be discrimination free. We think that we will have opportunities to acquire frequency. In principle, it doesn't necessarily have to be by way of auction. We are also open to an industry solution. France is often quoted, where the New Deal for Mobile was made, where frequencies were basically cut in four quarters. If we find that low-band frequency, i.e., 700, 800, 900, that's the entire range. If they are broken down into four quarters, then that is perfectly acceptable to us. For any other negotiation solution, we're open to it. Should the Federal Network Agency send out an invitation for discussions, then we'll be happy to participate.
We'd be willing to look for an industry solution should that not be successful, then the industry auction will be the tried and tested procedure that has been used for a long time in Germany and other countries in order to allocate a scarce resource in a fair and equitable way. Concerning the question of development of service revenues in the first half of the year, that is purely a quantity effect, a volume effect, due to the fact that 170,000 additional contracts were lost due to the Telekom effect. The other effects, such as ARPU, is still developing very stable across all areas. In every segment, we have a uniform, sometimes slightly increasing ARPU.
This low service revenue is due only to the loss of 170,000 contracts.
Those 170,000. Thank you. Are they coming off at similar levels of ARPU to the existing base?
The ARPUs are still quite stable in fixed landlines. In the premium segment, it's actually slightly increasing and stable in the discount range.
Great. Thank you.
Yeah.
We can now take a-
Well, the ARPU. Just one addition to the last question. The ARPU of those customers who were lost is of course comparable to the ARPU of those customers that we are winning.
Thank you.
We can now take our next question from Usman Ghazi of Berenberg. Please go ahead.
Into the network that you've done on obviously statistics. But I'm just wondering, like, what you need to be-
Usman, sorry that I step in. Excuse me.
Yeah.
Excuse me, sir, that I'm stepping in. The line is pretty bad, so we absolutely could not hear what you're asking. Pardon me. Can you repeat kindly, please?
Sure. Is it better now or still bad?
Now it's better. Thank you.
Okay. Yeah. I just wanted to dig into the fixed wireless access, you know, trial that you've started. Obviously the performance looks quite impressive, but I wanted to understand what needs to be done to scale the service up. You know, and whether this trial was, you know, like an antenna with direct line of sight to a household or, you know. I'm just trying to understand what happens when obviously there are more people in a cell. I mean, will the performance still hold, or will the performance degrade? Or, you know, just any color on how you see this service scaling up would be helpful.
Well, we expect that what we achieve as a friendly user test in three cities can be rolled out to 390 cities. That is, of course, what we are targeting, and it's a shared medium of course. The more users we have in a cell, the more bandwidth we use per user that we can allocate with any other mobile phone network, as well.
Next question is from operator.
Thank you. As a reminder, it is star one if you do wish to ask a question. Our next question now comes from Adam Fox-Rumley of HSBC. Please go ahead.
Thank you very much. I wondered if I could ask two, please. Firstly, could I just take advantage to ask you to reflect on the working relationship with Rakuten? Now, there's been a bit of time with some live sites up and running. I guess related to the live trial that was just asked about, and making another reference to Dish. I think in a sense they were surprised by the time it took them to optimize performance on cell sites as they got them up and running. I think you might have said it took you 24 hours. Just like a little bit of detail around that would be helpful. And then my second question was on CapEx and just, can you really spend EUR 400 million this year, bearing in mind that the money spent in the first half to date?
Thank you very much.
Maybe I can start with our cooperation with Rakuten, which is going very well. I said it yesterday in a supervisory board meeting, if it continued as is, that would be perfect. We really have the best possible partner there. I and the whole team are enthusiastic about it. Of course, it has to remain as is now over the years, but what we've seen so far is really very pleasing. Now concerning performance optimization, we didn't take a lot of time there. The test began in July, the friendly user test did, and I'd say we're three weeks into it now. At the beginning, of course, we didn't have the values we have now, but within two to three weeks, the desired values did show and hasn't been a problem for us so far.
Concerning CapEx, what happens is that, as I said earlier, CapEx is created by servers in the computation centers, but also through software that has to be adjusted. In the first half of the year, a lot of work has been performed. More than is reflected in the EUR 51 million CapEx, but the process looks at these software packages and created intangible assets being accepted first, and then they will be invoiced and paid for. There is a bit of a delay, and some shift from the first to the second half of the year, but it is not the case that there hadn't been anything done so far, as it were. A large part of the performance that is being delivered now will be invoiced in the current quarter.
Very helpful. Thank you.
As we don't seem to have any further questions, I would like to close the question and answer session and pass the word back to Oliver Keil for a closing remark.
Thank you, operator. Thank you to the audience. As always, we will be available for further discussions later on, and we wish you a good meeting later on with our mother, United Internet AG. Stay healthy, and hopefully we'll see you soon in person as well, maybe.