Ladies and gentlemen, welcome to this Ströer SE & Co. KGaA H1 Q2 Figures 2025 Conference Call. I am Valentina the Chorus C all operator. I would like to remind you that all participants will be in listen-only mode and the conference is being recorded. The presentation will be followed by a Q&A session. You can register for questions at any time by pressing star and one on your telephone. For operator assistance, please press star and zero. The conference must not be recorded for publication or broadcast. At this time, it's my pleasure to hand over to Christian Schmalzl. Please go ahead.
Dear ladies and gentlemen, dear annuals, welcome to our Q2 H1 2025 call. As always, I will begin the presentation with a brief overview of the key figures. Our strategic focus topic in this quarter's presentation, besides our core out-of-home business, will be a deeper dive into Statista, as we want to share with you some interesting developments and potentials we see going forward. Henning will then comment on the developments and effects of our Q2 figures in more detail. This will be followed by remarks on what we expect for the third quarter and the remainder of the year. As always, we are looking forward to your questions after our presentation. With that, let us start the call with a short overview of our H1 2025 developments. Before we look at the main KPIs for the first half of the year, a brief introductory remark.
After a very solid first quarter with revenue growth of around 5% at group level and revenue growth of over 15% in our core out-of-home business, we are working in the second quarter of 2025 against very strong prior year figures due to the UEFA European Championship 2024 Germany. As a reminder, consolidated revenue increased by 12% and out-of-home revenue by as much as 21% in prior year's Q2. The developments in the first half of the year should be interpreted against those outstanding quotes. Overall, revenue rose by around 2% to EUR 980 million in the first six months. Organic growth for the period was at 0.5%. Adjusted EBITDA rose only slightly by around 1% from EUR 263 million to EUR 266 million due to an increase in the cost base.
Compared with the previous year's figure, EBITDA adjusted declined by around 4% due to a minor increase in DNA compared with the prior year period. Accordingly, net income adjusted declined by 5% to EUR 52 million. H1 2024 was EUR 55 million. Free cash flow adjusted for the reporting period was -EUR 1.6 million. Six months 2024 was EUR 21.7 million. In line with our communicated CapEx strategy, which includes a more focused expansion of our digital out-of-home network and to further optimize the utilization of our digital outdoor advertising.
Ladies and gentlemen, please hold the line. Sir Christian, you are back in the conference.
In line with our communicated CapEx strategy, which includes a more focused expansion of our digital out-of-home network and to further optimize the utilization of digital outdoor advertising media, CapEx was slightly below the figure for the same period last year at EUR 39.6 million and roughly 40% below where we've been mid of 2023, just two years ago. Let's have a look at the Nielsen numbers in the middle of this chart first. As always, please keep in mind that those show gross rate card developments and the net revenue, including all discounts, is on average six to seven points lower. TV shows minus 5% on a gross level. ProSieben and RTL were reporting double-digit decline for linear TV and dock on a net basis. On a like-for-like and gross basis, the out-of-home category is outperforming all other media.
As we've already shown in the numerous quarters, the category typically outperformed the German advertising market by 5%- 10%. The second quarter of 2025 is a prime example of this. Even though the overall market in all categories is weaker than previous quarters, the outdoor advertising market, where we account for around 65%, has once again demonstrated its strength and resilience. Overall, the German advertising market declined by around 2% compared to the same quarter last year. Looking at the individual advertising categories, it is clear that the declining trends we saw in previous quarters have accelerated. Desktop mobile declined by 1%, radio by 4%, and television, the largest advertising category in the German ad market, by as much as 5%. In contrast, print grew by 3% and out-of-home in particular was the strongest category with a growth of 5%.
With that, outdoor advertising exceeded the important 10% share of Nielsen advertising market for the first time. This is an important milestone for us, clearly demonstrating the growing importance of outdoor advertising in the media mix, especially in the upper marketing funnel. Looking at our developments, it is also clear why out-of-home is growing as an advertising category. Digital outdoor advertising grew by 9% in the past quarter. Programmatic digital out-of-home performed even better, growing by 18%, which clearly shows where the trend is heading. With these developments, programmatic digital out-of-home is on par with the average growth rates of global platforms like Google, YouTube, or Meta. A similar picture emerges when looking at the cumulative period from January to June. During this period, the German advertising market remained flat with zero growth on a gross level.
The largest category, television, also declined by 3% during this period, on a net basis probably by around -8%. Out-of-home advertising, on the other hand, grew by around 11%, on a net basis probably by around 5%, driven in particular by the strong performance in the first quarter. With net growth of 8% in out-of-home, Ströer made a decisive contribution to this positive development. Out-of-home wins share in the Nielsen ad market and on top, we win share within the out-of-home sector. Our strong digital and tech positioning are the key levers. Our digital out-of-home grew by 17% and programmatic digital out-of-home by as much as 26%, significantly outperforming international players in the market. This positive and resilient business is the result of the targeted and consistent development and expansion of our digital outdoor advertising business with digital premium assets.
Given the many discussions around retail media and its growth potential in general, we show you an additional view on our portfolio. Ströer thus has around 76% of all two square meter or larger digital outdoor advertising media and public spaces in its sales portfolio and ownership. The only relevant competitor is JCDecaux, focusing primarily on street furniture. The picture is similar for other digital advertising media, such as retail screens that are mostly smaller than two square meters. Here, around 60% of all advertising media are installed, operated, and monetized by Ströer, and the remaining 40% are widely fragmented, very different to the classic out-of-home market. In addition, Ströer is one of the major players in the ambient media sector, marketing around 37,000 advertising sites, which counts on a net basis for only roughly EUR 30 million in the market spend.
It's a nice add-on business, but clearly a smaller niche with limitations for overproportionate growth. To put that into context, excluding the services and just comparing digital out-of-home versus classic out-of-home, the share of digital out-of-home roughly doubled over the last five years. Within digital out-of-home, programmatic was clearly the growth engine and has added many new business opportunities for our sales team. In other words, over 80% of our growth is digital, over two-thirds of the digital growth is programmatic. Over the last 10 years, out-of-home sustainably outperformed every other advertising format in the German advertising market. As a result, our category has more than doubled in the period from H1 2015 to today. Even the best alternative format lags around 60 index points behind out-of-home.
Put simply, apart from out-of-home, all other forms have more or less remained static over a period of 10 years on a Nielsen basis, and out-of-home shares in the advertising market have risen steadily, reaching a new high of over 10% in Q2. We are convinced that this is not the end of the journey, but merely a milestone along the way. We expand our network on the basis of already strong and long-term protected market share, increasing targeting options and available audience data, have low to no churn with existing customers and constantly add new customers while digitization puts all other local competitors under ongoing structural pressure. The macro environment defines what's overall possible, yet the relative momentum towards digital out-of-home is rather accelerating since the end of the pandemic.
Software as well as data quality will be the key growth drivers going forward to monetize our digital out-of-home ad inventory. In the last two to three years, we clearly benefited from the simple fact that we had invested rather early back in 2014 in an own SSP setup to make our digital out-of-home eyeballs available to online and cross-channel DSPs, and therefore bringing our volume to their demand. Today and in the coming years, we see similar beneficial trends around our proprietary DMP setup, which we established already seven years ago, fueled originally by online data only. The DMP provides finely detailed location-specific information for the pre-qualification of inventories based on the advertiser's clause. It's the central technology for collecting, analyzing, and using data from various sources to specifically activate ad inventory. Four concrete cases from the last six months where we leveraged tech and data for brands.
Otto was promoting an Adidas sneaker exclusively designed for them via 3D creatives on our towers. By scanning the code on the ad, people could directly do a virtual shoe fitting via the Snapchat lens. A perfect combination of large-format digital out-of-home branding and digital interaction and engagement. Audible was expanding their online content strategy on digital out-of-home with targeted native stories and high-impact formats. Laurence Paulsen went from TikTok to Big Screen, targeting specifically Gen Z. The content of the brand's TikTok campaign was amplified via public video, including integrated storytelling across the various screens. Alpecin, from Dr. Wolff, a long-term TV believer, used a locally targeted campaign in Berlin to leverage the fact that the third league team of their headquarter city, Bielefeld, reached the German DFB Cup finals as a smart marketing hook.
Now let us switch gears and focus on the potentials for Statista deriving from the AI age. Overall, we are at a turning point in the development of a data-driven business model. Artificial intelligence is changing the rules of the game, and Statista is excellently positioned to actively shape this transformation. Our goal is clear. We are building the world's number one data engine for AI-powered business intelligence. What does that mean in concrete terms? By using AI technologies, we are set to unlock exponential usage of Statista's data. With our data, we establish Statista as a trusted infrastructure layer for our knowledge management systems of globally leading companies. Our data pool is facilitating as a decision-relevant resource in corporates.
According to Capgemini and PricewaterhouseCoopers, 88% of companies plan to increase AI budgets in the next year, driven by the rapid rise of agentic AI, which can independently research, analyze, and support decisions. By 2028, these AI agents are expected to generate EUR 450 billion in economic value. Already 61% of organizations see agentic AI as transformative, and 58% of business functions are expected to be AI-supported within three years. Notably, 70% report near-complete integration of AI agents across workflows. These figures show we are not at the start, but in the midst of a shift how we work. To stay ahead, companies must act with clear strategies, targeted investments, and openness to new technologies. A common misconception is that LLMs reduce the need for data providers like Statista. In reality, the opposite is true.
In an AI-driven world flooded with low-trust content, the value of reliable, curated, structured data surges. That is why Statista is evolving from a seat-based expert tool to a scalable embedded data infrastructure. Our structured data is now integrated into clients' internal knowledge management systems, powering enterprise search and LLM tools for tens of thousands of consultants. In short, generative AI doesn't replace verified data. It makes it indispensable. Statista's value in the AI era isn't reduced, it's multiplied. The rapid spread of AI tools, especially large language models such as ChatGPT, is changing the strategic direction of companies worldwide. Within a matter of months, these models have reached hundreds of thousands to millions of users. ChatGPT, for example, took just five days to surpass the 1 million active user mark. Today, it has over 600 million monthly active users.
The speed with which AI models are gaining users worldwide is impressive and demonstrates the enormous relevance of this technology for business and society. Current monthly active user numbers of LLMs, just like Gemini, Google, are over 450 million. ChatGPT, OpenAI, more than 600 million. Meta AI, around 1 billion, are massive and send a clear message. AI has become a part of everyday life, and the demand for reliable, structured data sources is growing rapidly. With this momentum comes a growing challenge. AI needs structured, verifiable, and trustworthy data to deliver reliable results. This is precisely where the strategic opportunity lies for companies like Statista. Here we step into the ring. While generative AI dominates headlines, its success depends on trusted, clean, and compliant data. Without it, even advanced models fail to deliver meaningful insights.
A Salesforce survey of 6,000 global knowledge workers found that over 75% view accurate, secure data as essential for AI trust. Yet many organizations still face data gaps that limit scalability. This is especially critical with the rise of agentic AI systems, making autonomous decisions with minimal human oversight. In this context, data reliability becomes paramount. A Forrester study for IBM shows companies that scale AI effectively are seven times more likely to lead their industry. The main barrier? Poor data quality and/or limited access to data. The economic impact is clear. Firms investing in data pool, governance, and curation are best positioned to unlock AI's full value, driving efficiency, innovation, and growth. Trusted data isn't just technical, it's a strategic asset.
Statista is evolving into the trusted data infrastructure for corporate knowledge systems, shifting from a data provider to a strategic enabler of AI implications, and thus lays the groundwork for deep integration into enterprise environments. This infrastructure supports AI-driven decision-making through 1 million statistics based on enriched external data, 84,000 industry reports, 300,000 in-house statistics, 3 million global consumer interviews across more than 15,000 brands, 20 million structured data points across 170 regions, and 85 million synthetic profiles with 48 demographic and behavioral traits. Statista's content is more than information. It's a strategic asset for companies seeking reliable, data-driven decisions. Let me give you a quick example of the possibilities that emerge when using our data in enterprise knowledge management systems for consumer insights.
Based on 3 million interviews worldwide, coverage of 56 countries and territories, mapping of over 3.5 billion consumers, analysis of 15,000 international brands, and insights into over 500 topics and industries, our customers can generate targeted messages for their target groups in a data-driven, precise, and scalable manner, opening up a wide range of possibilities, just like optimization of target group addressing through understanding of media use and consumer behavior, analysis of personas, needs, and attitudes for tailored communication, identification of relevant marketing touchpoints along the customer journey, trend monitoring to quantify global consumer trends, and many more. Let us now focus on market insights and our offering for deep market analysis based on proprietary indicators across major life and economic sectors.
With over 2,000 proprietary indicators from eight key areas: economy, demographics, health, telecom, finance, transport, infrastructure, and education, 40 years of historical data for long-term trend analysis, coverage of 1,000 plus markets in 170 regions, up to five years of forecasts, and over 20 million structured data points, we provide not just retrospective views, but a foundation for strategic foresight. Beyond data collection, we ensure source validation for credibility, data harmonization and preprocessing, gap correction, and consistency checks, forecasting via statistical and machine learning models, and quality assurance through expert review and scoring. This enables us to deliver key market indicators that enhance the reliability and value of our insights. In recent years, Statista has established itself as a leading global platform for accurate and trustworthy business data, and this is also reflected in our digital visibility and reach.
According to the renowned SEO analyst tool SISTRIX, Statista is one of the most visible websites worldwide when it comes to business-related content, as our trusted content is cited millions of times every day by Google. Our Google visibility is on par with leading institutions such as McKinsey & Company or the Financial Times. Statista also ranks amongst the top international domains in the Backlink ranking, which measures the number and quality of links from other websites. We are in the direct company of organizations such as NASA, the U.KM government, Wired, Time, Conva and eBay. Statista has created a structured, harmonized database that not only forms the basis for our content, but also enables integration into a wide variety of digital ecosystems. Our data is systematically structured and enriched. Harmonized data formats and multi-layered taxonomy ensure consistency and reusability. Qualitative annotations create context, meaning, and connections between data points.
People remain an integral part of the process. Our human-in-the-loop approach ensures quality and relevance. Our platform is not isolated. It's interoperable. With Statista Connect, we offer API access and integrations with leading tools and platforms, including Canva, Perplexity, and Microsoft Copilot. This makes our database directly usable in enterprise knowledge management systems and general LLMs. Statista goes beyond traditional data access. In addition to search and discovery, we also enable interactive AI-powered usage, for example, through our new MCP services. Our data is stored in AI-readable chunks, optimized for machine processing and semantic analysis. Centralized quality control ensures that our data is not only available, but also reliable and consistent. Breakthrough agreements with leading AI platforms, including Microsoft 365 Copilot, Perplexity, and Canva, allow Statista to successfully lay the foundations for positioning itself as a player of high-quality data for professional research. These partnerships are not just integrations.
They are distribution multipliers. In Microsoft 365 Copilot, users can search and pull Statista data directly into Word, Excel, and PowerPoint. With Perplexity, Statista enhances AI-generated answers with trustworthy, up-to-date, and verifiable data, reinforcing its role as a go-to source for reliable insights. In Canva, Statista powers real-time data visualizations, democratizing access to high-quality data for designers and communicators alike. The integrations create a new data journey, one that's deeper, faster, seamless, and more actionable. This means Statista will no longer be just a platform. It has the potential to become an embedded layer in the global AI productivity stack, unlocking massive scale and recurring usage. In recent months, Statista has entered into groundbreaking partnerships with leading global companies in the field of logistics, technology, and professional services, which are extensively using Statista's data via API interfaces, which adds additional access and usage of the data.
These three form strong use cases. Number one, a leading global professional services firm, where Statista is integrated into the central knowledge management system of the firm. This enables access for all professionals worldwide with direct search function for internal documents and selected external data sources. Currently, we are in a paid pilot phase on a usage-based pricing model. Secondly, a global logistics service provider with basically the same setup. Third, last but not least, a leading technology company. Here, Statista is directly connected to the internal market research tool via an MCP server, a model context protocol. The payment model we test here is consumption-based and also credit-based. Currently, we are in a paid pilot phase with access for all commercial employees. These examples illustrate our data is not only relevant, it is scalable, strategically valuable, and seamlessly integrated into business workflows.
A strong customer base is not only a sign of trust, it is a strategic foundation for sustainable growth. Statista has a globally unique cross-industry customer landscape comprising leading companies, institutions, and organizations such as consulting and professional services like McKinsey or BCG, banks like JP Morgan or Barclays, Nestlé, Unilever, Amazon from the FMCG and e-commerce sector, tech giants like Google and Meta, and many more. AI will act like a catalyst to significantly transform our platform usage, scaling, and monetization, especially in the enterprise segment. Statista is evolving from a pure research tool to an integral part of corporate knowledge systems. AI enables us to embed our content seamlessly, contextually, and deeply into decision-making processes. This not only increases relevance for all knowledge workers, but also boosts customer loyalty and usage intensity. At the same time, we are adapting our business model.
In addition to traditional per-seat pricing, we are increasingly focusing on consumption-based models such as API calls or other data consumption metrics. This flexibility allows us to tap new revenue potential and increase scalability. Statista is becoming the core data infrastructure layer for company search and knowledge management. With Statista Connect and AI-empowered tools such as conversational and agentic AI, we enable direct integration of our data into company workflows quickly, automatically, and scalably. Access to our curated, reliable data now takes seconds instead of minutes. Statista is no longer just for experts, but can be used by all employees. A clear lever for usage intensity and customer loyalty. Paid customizability, fast time to insight, and API connectivity make Statista an indispensable provider for data-driven decisions and create the basis for sustainable growth and recurring revenue. Those are my remarks, and with that, over to Henning.
Thank you, Christian, and a very good morning, everybody. Let us start the finance section, as usual, with a review of the Q2 2025 P&L. In total, and against a somewhat challenging ad market context, as well as higher prior year comparables, we are looking at a robust set of financials. Out-of-home media and in particular online ad business turned in slightly lower growth than we anticipated three months ago, and Christian already elaborated on the weaker than expected context in the second quarter. Revenue growth for the quarter was -1%. This includes around 100 basis points supported mainly from the acquisition of RBL Media. The corresponding activities have been included since November last year. Excluding this effect, organic growth came in at -2.3%. EBITDA adjusted amounted to EUR 149 million and is in line with our internal forecast for the quarter after EUR 155 million in Q2 2024.
The exceptional items for the quarter were EUR 3.7 million and thus roughly on the prior year's level. The exceptional items essentially comprise four components: EUR 1.4 million for various smaller restructuring measures, EUR 1 million for software implementation, EUR 4.4 million for transactional exchange rate effects, and expenses related to our stock option program. Accordingly, reported EBITDA was EUR 145 million after EUR 151 million last year. Depreciation and amortization increased by 5% and thus more or less by the same growth rate like in Q1 to EUR 83 million. With that, reported EBIT for the quarter came in at EUR 62 million, some EUR 10 million lower compared with Q2 2024. The financial result improved to -EUR 16 million after -EUR 18 million in the prior year period, mainly due to lower interest rates and positive exchange rate effects. Earnings before tax therefore decreased to -46 million after EUR 54 million in Q2 of the prior year.
The tax rate was basically unchanged with around 30% in the reporting period, and with that, the tax result follows the development of EBT. All in all, reported net income for the quarter came in at more than EUR 32 million after EUR 38 million in Q2 2024. Adjustments were down by EUR 1 million, mainly resulting from aforementioned exchange rate effects within the financial result. Accordingly, net income adjusted was EUR 36 million after EUR 42 million in the prior year. Let us now switch over to the cash flow. Our cash flow in the second quarter is compared against an exceptionally strong development from the prior year. On the back of strong trading around the UEFA European Championship, last year's Q2 free cash flow adjusted surged by EUR 47 million, representing a very strong improvement. Nevertheless, this year's Q2 delivered solid positive free cash flow.
The decline against the prior year mainly derives from lower EBITDA, higher cash out for interest, including the first-time fixed interest payments on the EUR 268 million note loan that we placed in June last year, and higher cash out for taxes that benefited last year from a reimbursement. These developments were to some extent compensated for by an improvement in net working capital and stable cash out for investments. Including broadly stable lease liability repayments, free cash flow adjusted amounted to EUR 34 million after EUR 46 million last year. Just as a reminder, free cash flow adjusted in Q2 of 2023 was still slightly negative. Compared to our calendarized forecast, our Q2 H1 cash flow was still somewhat better than we anticipated in our internal forecast, and we continue to expect good cash flow generation for the full year, especially in the fourth quarter of the current fiscal year.
Let me come to the net debt development. In the sequential view from the end of Q1 2025 to the end of Q2 2025, net debt was up by roughly EUR 91 million, including the adjusted free cash flow for the second quarter of +EUR 34 million and cash out for our dividend payment in Q2 of -EUR 128 million for the holding and of -EUR 5 million for minorities. The remaining difference in the reconciliation of around EUR 9 million mainly relates to a reduction of accrued interest liabilities paid in Q2, representing a cash outflow with no effect on net debt.
Net debt year over year was up by EUR 112 million to EUR 956 million, including the cumulative free cash flow over the last 12 months of EUR 135 million, cash out for the acquisition of RBL Media amounting to -EUR 106 million, and cash out for dividend payments of -EUR 128 million, and cash out of -EUR 20 million for minority dividends. With that, our leverage ratio in the second quarter amounted to 2.47x after 2.28x at the end of prior year's Q2 and 2.18x at the end of Q1 this year. Let us now take a look at the performance of the individual operating segments of the past quarter, starting with our core segment out-of-home media. Out-of-home media continued to outperform the German ad market, as we have seen at the beginning of this presentation.
Against a record and the underlying sales growth rate of more than 20% last year, growth in the second quarter this year came in at 1%. This development, in particular towards the end of the quarter, was slightly short of our own internal expectations. Nevertheless, growth in digital was with close to 10% still at a very healthy level. Since the prior year was heavily influenced by the Euro Football Cup here in Germany, it is useful to look at the two-year comparison. Compared with the same period in 2023, Q1 2025 achieved growth of more than 33%, while in Q2 growth moderated to 22%. Even though the extent of this moderation is a little bit exaggerated by the Q1 business benefiting from the 2025 federal elections, still there's a bit of sequential slowdown from Q1 into Q2 that relates to the overall market environment.
At the same time, it is also important to note that the average growth over the last two years in Q1 and Q2 was above 10%. Looking across the different revenue streams, out-of-home classic declined by around 3% in Q2, including positive effects from the acquisition of RBL Media. Excluding this effect, sales were broadly stable for the first six months. As already mentioned, digital out-of-home grew by almost 10% in Q2 and by more than 17% for the cumulative period. Within digital, programmatic digital out-of-home grew by almost 26% year to date. Q2 EBITDA adjusted was flat despite only moderate sales growth. For the first six months of 2025, EBITDA adjusted grew in line with sales at a broadly stable margin of more than 44%. For the remainder of the year, we remain constructive regarding our ability to gain further market share.
The current Q3 will, however, be somewhat more challenging than we anticipated in our initial budget. Nevertheless, we expect a clear acceleration in the fourth quarter based on low comparables and the fact that last year the advertising peak around the UEFA European Championship led to exhausted budgets towards year-end 2024 trading. Please also bear in mind that our margin in Q4 last year was impaired by a one-off accrual for a VAT settlement. Q2 revenue for digital and dialogue media amounted to EUR 210 million after EUR 215 million in the prior year period. Digital media thereby, after an increase of 2% in Q1, declined by 3%, reflecting a more challenging market environment for online ads.
Within digital media, sales from our owned assets such as programmatic digital out-of-home and owned internet content such as T-Online grew but could not fully compensate for a decline in sales in our operated third-party portfolio. Our dialogue business came in with revenues of EUR 105 million in Q2 and with that 2.5% below the prior year level. In detail, the two dialogue activities, call center and direct marketing, showed different sales dynamics in the second quarter. Our call center's activities grew by high single digit but could not fully compensate for a sales decline in the door-to-door business, where we're still facing a higher than usual churn in the field force. In total, the segment EBITDA adjusted for the first half came in at EUR 59 million after EUR 69 million in H1 2024, and this decline relates roughly half-half to digital media and door-to-door.
Finally, some comments on our data-as-a-service and e-commerce segment with Statista and Asam. Revenue developments in the second quarter were more or less in line with the first quarter. Sales declined slightly from EUR 86 million to EUR EU85 million. Statista sales grew by 1.5% in Q2 and were impacted by currency effects. Adjusted for currency, growth amounted to 3.6% for Q2 and also for the first half. Christian already elaborated on the transition at Statista, where we positioned Statista as a trusted data source in a rapidly growing AI market. At the moment, though, we are facing considerable headwinds in our inbound platform business. Here we are dealing with smaller businesses, in many cases freelance knowledge workers, with usually only limited budgets for monthly subscriptions and reduced needs for professional in-depth data access.
At the same time, in our outbound platform subscription business, where we market our product to larger industry clients and organizations, we are growing very healthily. Based on that, and as outlined before, we see enormous potential for Statista, in particular given the great contributions we can make to data integrity and reliability in the current AI revolution. However, it will take some time before we see the effects in our numbers. Sales at Asam continue to be impacted by the challenging consumer environment in Germany. Although trading was a bit better in Q2 compared to Q1, our most relevant sales channels, retail and e-commerce, had to face declining sales. For the first six months, sales in retail showed a mid-single-digit sales decline, while e-commerce was down by a low single-digit number. The China business was responsible for around one-third of the sales development in the first half.
Going forward, we expect the e-commerce business to remain challenging, while we expect to return to growth in our retail channel based on compliant planning forecasts by our retail clients. Q2 earnings for the segment remain stable despite the sales development, reflecting tighter cost control in both units. With that, let me hand you over back to Christian for the outlook and some closing remarks.
Before ending the presentation, let me just add some comments on the outlook for Q3 and the current trading momentums. What do we expect for Q3 and the remainder of the year? For our core business, out-of-home media, we expect Q3 revenue growth broadly in line with what we have seen for Q2. Q4, however, will benefit from significantly lower comp. The growth will be driven by programmatic digital out-of-home and digital out-of-home as they are established as main growth contributors. Digital and dialogue media's revenue development should be in line with H1 developments for DAOs and e-commerce. We expect an improved revenue development compared to H1. Our guidance for 2025 remains unchanged, with a higher dependence on the business development in the fourth quarter due to summer business being more subdued than originally expected.
Mid-long term, we continue to expect double-digit top-line growth on average in our core business, out-of-home media. With that, let me now close the presentation with a short look into our financial calendar and the remaining publication dates for 2025, which is Q3 on November 11. The release date for 2026 will be published on this occasion. As always, updates, reports, and roadshow presentations can be found on our iorg website. Thank you, everyone, and we are now happy to take your questions.
We will now begin the question and answer session. Anyone who wishes to ask a question may press star and one on the touchstone telephone. You will hear a tone to confirm that you have entered the queue. If you wish to remove yourself from the question queue, you may press star and two. Participants are requested to use only handsets while asking a question. Anyone who has a question may press star and one at this time. The first question comes from Annick Maas from Bernstein. Please go ahead.
Good morning. Thank you very much for the whole presentation on Statista and to show the value of it in an AI world. I guess my question is, you've shown that you're already in a paid pilot phase with two companies, and yet we haven't really seen any strong revenue acceleration in the second quarter. If all of this plans out or works out the way you imagine it to be, how do you think about revenue acceleration in the next quarters and years for Statista? My second question is, under fully guidance, you confirm it. Can you, besides comps, maybe give us a little bit more of an explanation as to the confidence that you will be able to do that? My last one is on digital and dialogue.
Comps are getting much easier, and yet you're saying that the revenue development should be in line with the first half. I get the comment on Ranger or on the churn for Ranger, but if you can tell us a bit more whether there might be something structural that call centers are less used in an AI world, or why is this Ranger churn now an issue but hasn't been in the past? Just a bit more detail there, that would be great. Thank you.
Annick, let me maybe take the first question to Statista. It is definitely too early to say that. I mean, we wanted to give her an outlook today because many people were questioning what is the future of Statista in an AI-driven environment. We are really optimistic about Statista, but we're going to change also our distribution model from selling seats to selling volume. For example, look, if you're integrated in a knowledge management system, then the AI-driven knowledge management system is going to have automatic access to our data pool depending on what a certain employee is asking. The pilot phase is paid, but this is more symbolic payment. Let's keep it like that. This is really a transformational phase. What becomes more and more clear every day is that AI, especially in the research area, is only working with trusted and verified data.
A big part of Statista's data, don't forget, that is NDA proprietary. It's all these market insights, consumer insights. This is data that we created based on interviews, et cetera, et cetera. It doesn't exist anywhere else. Even if you go to publicly available data, it shows if you don't have metadata, structured, standardized metadata on top of it, it's very difficult for the AI to deliver the right results. The feedback we're getting is very, very good, but it's definitely too early. We see also in the development of the turnover this year. This is a completely transition year. For us, it's very important that we see that the demand from the big LLMs and also from our big global corporate clients, it's very strong after a phase of unsecureness, and nobody knew really what's going to happen.
We had a big client, for example, he canceled his account whatever six months ago because they thought they can do it only with an LLM, and then they came back three months later and paid double than before. It's a real transition period, that's why we wanted to share some insights today, why we are really confident about Statista's future.
Okay, thank you.
On your question around full-year guidance, Annick, I think what we've clearly seen, and you find similar information also from other media companies, is that the whole summer was overall rather soft, including July and August. Nevertheless, afterwards, we are running against soft comps of, I think, low single digit last year. What we see at the moment is that the order book for late August and early September already looks quite promising. The general feedback from the market, including agencies and large customers confirming their annual commitments, just, I would say, makes us feel comfortable with our original plans for the full year. I think we had a completely different phasing this year than last year and probably underestimated a little bit that whole momentum that we had last year in the ad market through both the UEFA European Championship as well as the Olympics afterwards.
Definitely, the whole tariff discussions around Trump's latest moves haven't helped as well. We don't see fundamentally any issues with the feedback we get from all of our optimism agencies, especially on our advertising business. Maybe following up on the third question in that context as well, I think especially the contact center business is doing really well this year, both top line as well as bottom line-wise. With the door-to-door business, we've come to a point or a level where we need to scale the organization to the next level with far more than 3,000 sales guys in Germany. That combination of bringing in more people, managing the churn, being more demand on more complex fiber sales shops, with the bonus system we are ultimately working on, just makes the business a little bit volatile, I think, at the moment.
The underlying, I would say, demand from the market, and I would even say the growing structural importance of push channels like door-to-door and direct sales, is unchanged.
I think it's worth noting that our outlook here is based on a tariff agreement between the U.S. and the EU. The visibility in reality is quite low because we don't know what's going to happen. We see there was an agreement, now it's maybe not an agreement anymore. Overall, in the economy, you see a clear impact on this question. This is clearly something which is impacting the whole, let's say, trading momentum in all industries like never seen before from my point of view. We saw a lot of different crises and stuff, but this insecureness about what's going to happen is really impacting people. In case we are not going to see a settlement, that can be really a more difficult situation towards the end of the year, but nobody knows that. We are dependent on a unique external driven situation here.
I think everybody's hoping this comes to an end now, but nobody knows. I mean, we don't even know where we are standing right now, we as an economy and the tariff stuff. That is clearly, I think the overall economic development is really related also to this question.
Okay, thank you very much.
The next question comes from Julia Roch from Barclays. Please go ahead.
Yes, good morning, and thank you for taking my question. The first one is, can you remind us what is your full-year guidance? Because you're reiterating it, but it's always clearer when you state or remind us what it is. That's my first question. The second one is on Statista. Annick already asked, can you say it's too early to give us a guidance, but maybe to help us frame it, you explained you had two types of clients, the small one inbound platform business, smaller business, freelance knowledge worker, and large clients. The first one was doing badly, the second one was doing well. Could you give us an indication of the split of revenues between the two and what kind of growth rate they're experiencing in the first half? That is my second question.
The third one for Henning, if we could have some help on free cash flow for the full year. Indication of the level of cash exceptional, cash interest, cash taxes, working capital. I know it's hard to forecast, but some indication, CapEx and leases. Only six numbers, Henning. I'm not greedy. Thank you.
All right, thank you, Julie. Maybe I start with guidance, sort of you asking what it is. I mean, we set in our annual report 2024 that we expect organic growth in fiscal 2025 to be roughly on the organic growth level of 2024, which was at the time slightly more than 6%, I think 6.5%. We also forecasted an EBITDA adjusted, which is growing ahead of this. That is, let's say, the official guidance. At the same time, I think it's worth mentioning that what we see currently in the consensus is quite reflective of this original prognosis. On the question, what is the share of the inbound business within Statista? If I recall correctly, it is probably 25%, 30% of the sales right now are inbound.
The rest of them, the outbound platform sales to the larger clients, you can see that there must be considerable pressure, quite strong decline in the inbound accounts at the moment. At the same time, we see that the overall customer mix is shifting positively towards high valuable customers, so with a higher spend per unit, if you will. On the cash flow for the full year, as I said, we remain quite positive, in particular for the fourth quarter. Don't forget that if you look at the real net cash flow in Q4, so including M&A investments, we had a considerable cash payout for the acquisition of RBL Media. By nature, I think we will see a very strong cash flow in the second half.
I think also for the free cash flow adjusted, as it's also part of our forecasting regime, we are constructive a little bit more for Q4 than we are Q3. Roughly going through the lines, I think in terms of interest expenses, I would say we have seen now an increase in the first half. I would not expect an increase for the second half. I'd rather see this number coming down. On the cash out for taxes, we're currently forecasting, I'd say, a number between EUR 70 million and EUR 80 million, with some chances that it could be closer to the EUR 70 million than to the EUR 80 million. In terms of working capital, we had now a slight positive development in Q2. For the full year, we remain a little bit conservative and forecasting a little bit of a cash outflow. Investments have been down for the first half.
We're forecasting for the second half a slight increase, but I think also probably here is some chance that this will only reach the prior year level. I think these are roughly the relevant lines, and I hope that is useful for you.
Yeah, very useful. Just missing leases and exceptional.
Leases, good point. I think if you look at the first half, I would sort of extrapolate this trend for the lease liability repayments also for the second half. There's always a little bit of fluctuation from one to the other quarter. What you see in H1 is a trend I would extrapolate for H2. For now, exceptionals, right? Your question. Right now, for the first half, we are slightly lower in the exceptionals. For the full year, we would still forecast at least not exceptionals above the prior year level. Rather, there's a chance that we see a little bit lower exceptionals.
Okay. Coming back on Statista, is it possible to have a sense of by how much the inbound is declining in the first half? I mean, you say it was a big decline, but are we talking down 10%, 20%, 30%, 40%, 50%?
I would say it's more than 20% than 10%.
Okay, thank you very much.
The inbound stuff is small corporates and also private individuals, let's say, for entertainment purposes sometimes. Okay, this stuff goes clearly to the LLMs directly now where you don't need a super accurate answer or you can guess already what is wrong and what is right. I mean, our focus here right now is definitely not on optimizing the last percentage of growth or turnover. Our focus now is clearly to achieve exponential growth in usage because the value of Statista will not be decided if you make now 5%, 10%, or 20% more turnover this year or next year. The value of Statista is clearly decided on are we able to show and to prove that any knowledge management system delivers better results for business data-related questions with Statista than without Statista. This is clearly the focus.
I mean, that's why we want to give you an update today because we're really happy that we could strike deals with some of the biggest LLMs already where we went through extensive testing beforehand and where we saw and where the LLMs saw that they deliver better data in this field, better results in this specific field of research than without us. This was a very positive outcome for us. The same is the first, which we cannot name now, but it was the first global corporate integrating in the internal management system, which means Statista is going to evolve from a researcher tool where you... Because the problem was Statista generated more and more and more data. This is obviously positive for a data, let's say, marketplace. The problem is it became more and more difficult to use. It became really a specialist tool.
Nobody, let's say, normal employee or a CEO was not able to really work with Statista efficiently. We had to ask a specialist in the company who had to start to search for statistics. This is now going to change completely because integrated in a knowledge management system where every employee has access to, suddenly, whatever you go from 1,500 seats in a big consulting firm to 20,000 users who are using implicitly Statista without even knowing it because the internal system has access to a bigger number of different data tools. This is actually what's happening right now, why we are totally focused on achieving a higher usage by developing from a specialist tool to a data infrastructure layer.
Thank you.
The next question comes from Craig Abbott from Kepler Chevreux. Please go ahead.
Yes, good morning. Not unsurprisingly, my questions were also very much Statista related, which obviously most of them have in the meantime been answered. If I may just follow up on that, please. Nevertheless, we obviously have to try to forecast in some way, you know, revenue streams for the coming quarters and years. I mean, is there any way for us to try to gauge, you know, what magnitude over time these deals with the leading LLMs, how that could potentially translate into the revenue potential for Statista? Say if we look out over the next year, two years, three years. Thank you.
The answer is no, unfortunately, because we don't even have the final pricing system here because everybody is collecting new experiences. Let's come back to this knowledge management system for a big consulting firm. Now suddenly, 20,000 people have access to Statista, but we are now in the pilot phase. It will take another, let's say, three to six months before we know if one consultant from this company, for example, takes a Statista service one time a day, 10x a day, 100x a week, nobody knows it. We are actually right now exactly in this process. For example, there we have a pricing system where we got paid by call, but nobody knows how many calls there are going to be. It's clearly what we see.
It looks very promising, but the pricing will be actually developed with large language models and with the big corporates in the upcoming months. They're really in a transformative process right now. We are totally focused on a number of users and clearly right now less focused on optimizing the turnover right now because what I already said, valuation of Statista will be one-to-one connected to how many people are using it and how many calls we get from the large language models and from the company GPTs. That will be key for Statista valuation building in the next, let's say, difficult to say, four, six quarters. That's the process we are right now in. It's absolutely impossible, unfortunately, to say, okay, in three years we're going to double the turnover or whatever. It's too early, Craig. I'm really sorry for that.
I think we have a lot of good news here, and that's what we wanted to share with you.
Okay, thank you for that. If I may just follow up on that real quick. I realize it's still very early on this in terms of tracking this KPI as well. Can you give us a feel for how you're seeing those numbers of users in your pilot cases so far, how they're developing? I guess since you're highlighting all this, that it's quite encouraging. Is there any color you can add there? Thank you. I've got one more.
No, as I said, the point we want to share here is that we see already the fact that we signed, as one of the very first data sources, signed a contract with Perplexity, with Copilot. That is the news here. We are now technically, there was a beta phase where we saw it's generating better results. We are actually in the phase of technically integrating it. We have to develop a pricing model together. There are no, that is really in the very first beginning. Also, for example, these knowledge management systems in global corporates, they're also in the moment they are actually developing. I don't know. I mean, for example, our customers we are talking to, they are also in the process of rolling it out.
For example, this, I don't want to come back always to the same consulting firm, but only a fraction of these 20,000 consultants have now access to their own system because they're also in the process of rolling it out. It's unfortunately too early, but the pure fact that after extensive testing, we are integrated in the most advanced solutions, AI solutions, AI-based knowledge management solutions today in the market. I think this is actually the positive fact here because in the moment, 20,000 people, for example, to 1,500 are getting access. We calculated, for example, how many employees have actually Statista current customer base. This is like 28 million, but only a fraction of seats what we sold up to now because it was the specialist system, only for researchers. Now, actually, we have a completely game changer here.
AI is a game changer for Statista because everybody gets access in the moment he needs data from Statista, but connected with AI-driven LLM or company GPT.
Okay, I understand. Thank you. Now, moving over to out-of-home media, please, I just want to make sure, when you were talking earlier about the bookings that you're seeing also, Tom, at the end of August, early September, that they're starting to pull up, which obviously they will have to do if you're going to achieve your guidance. Can I just be a little more specific here? I mean, were you specifically referring particularly to out-of-home media? I may simply have misunderstood that acoustically earlier. If you could give us some color here on what you're seeing that underlines your confidence again for that pull-up. I understand the base effect, but you still need that number to be pulling up. I just would like to see if you could give us some better indication of the visibility you have here. Thank you.
Yes, the earlier comment was for out-of-home media. I think that's what we see. All the feedback from agents and agencies and clients at the moment is extremely positive. We have a lot of cases we are working on with customers that are just trying to find solutions for the kind of audience gaps they meanwhile have in their TV plans. I think what was, as Udo said, was surprising is the kind of impact short term through the volatility around tariffs. They're creating uncertainty, plus that prior year's effect of really strong summer months. In fairness, I think the Olympics ended sometime in July and August. June, July, and August last year were extremely strong. This year, softer than expected. The other way around, the general overall feedback for the full year commitments from clients and agencies has not really changed in the last couple of weeks.
That's where we are at the moment.
Okay, that's very helpful, Christian. Thank you very much.
There is really also risk. I really want to highlight what we highlight frequently. For us, it's a key point that the underlying structural growth for our core business is completely unchanged. We are outperforming the market, but we are also not able to compensate everything that's happening around. Our guidance is clearly based on the expectation that the tariff stuff is going to be finished more or less now.
Okay, that's understood. Thank you.
The next question comes from Nizla Naizer from Deutsche Bank. Please go ahead.
Great, thank you. I have two questions on Statista, one on the general ad market. On Statista, could you kind of tell us, Udo, are you working on adding more LLMs to sort of the partnerships that you have already? Could you help us understand how these partnerships work in the sense that do the LLMs pay you to get access to Statista data, or is the monetization likely to come from the end user and their usage alone? Maybe some color to how that is structured would be great. Secondly, when you think of these opportunities now, how does that change how you think of crystallizing your own portfolio? You've previously mentioned a potential sale or a spin-off of Statista. Has that changed with these opportunities now at play? My last question is on the.
Marketing HQ, in case it does become more challenging than expected, do you have measures in place to protect your margins? Some color on how you think about profitability for the rest of the year would be great. Thank you.
Yeah, let me first answer this Statista question. Overall, our target to lay off non-core businesses hasn't changed, but it's clearly a matter of timing. Based on the AI development, we go through a transitional phase. I always say it's like in waves. It goes sometimes up and down. It's never a straight line, these disruptive transitional phases. That's what we saw in the past when the internet was coming, as we say, and what we see now with AI. Our target here is clearly that we want to integrate Statista as a strategic data layer in every important large language model and every important knowledge company GPT, knowledge management system. That is the process we are right now in. From the payment side, that's a good question. Right now, this is not our decision clearly because it's a strategic decision of the large language models.
Are they buying, let's say, data sources to create a difference to other large language models? Or are they, right now, we see more the development that they try to become a marketplace where you can choose as a customer. You pay a fee to the large language model, and then you can choose actually what kind of data sources you buy on top to make sure that the large language model is delivering you the results you really, and that you really need to have. That might be different if you would use it only, let's say, for entertainment purposes or for professional purposes. Therefore, you might have a different strategy in what you want to buy. I think it is also very early days. This might also change. It's moving extremely fast, like when we saw ChatGPT 5.0 last week, for example, and also with new features again.
The whole environment is moving extremely fast. We wanted to show here that we are fully on track and that we make very good progress. All these commercial questions, I completely understand that you guys want to know more. I also would like to know more, but it's just right now happening. Every day, something is happening. That's why it's too early now to go deeper into commercialization options. Everything is focused on usage now. That's why our target is clearly to be wherever we can be and to put up the usage numbers as much as possible. Monetization also in terms of laying off non-core businesses, we follow up on that. The strategy here is completely unchanged. It would be now the completely wrong moment to do that because we have the feeling that we are on a very positive track. That's what we clearly want to follow.
Maybe, Nizla, on your question on how we are reacting to the ad market, which might potentially be weaker than we anticipated. We are not really waiting until the market is better or weaker than we anticipated. Already as a reaction to Q2 trading, we launched, I would say, a comprehensive cost freeze and cutting plan where we go through all relevant cost positions, and we drop out everything which is not absolutely critical for the sales success at this point in time. We are already in earnings protection mode to be prepared for any scenario we might face now in the second half.
Thank you very much. Very helpful.
As a reminder, if you wish to register for a question, please press star and one on your telephone. We now have a follow-up question from Julian Roch from Barclays. Please go ahead.
Yes, hi there. Me again. On door-to-door, you said it needed scale and 3,000 salespeople were not enough. Can you elaborate on what you mean by that? What's your strategy there? What's the investment on door-to-door? Another one is what percentage of programmatic advertisers are incremental? Thank you.
On the range of business, what we meant with more salespeople is that, for instance, the demand for more sales capacities for fiber is not only unchanged, it's growing. For example, Deutsche Telekom, for instance, wants to improve the yielding of their fiber infrastructure, and they ask us to ramp up our sales force. The challenge for us is in those times, if you go from 3,000 to 3,500 or 4,000, you have to adjust the organization a little bit. Especially when you hire incremental people, you have at the beginning a little bit more churn. That has an impact on the productivity of the total sales force, and that has an impact on the bonus scheme. That's a temporary, I don't know, adoption to the growth that is made by unchanged demand from our key customers. That was the underlying message on door-to-door or the range of business.
Maybe to build on and add on what Christian just said, just to remind us, the sales force we're talking about is not on the payroll of Ranger. It is rather, let's say, the network of independent sales organizations that Ranger manages. It's their field force, right? Christian, would you want to take the question on how much of the programmatic sales is, I guess, incremental to out-of-home bookings?
Yes, I would say 80% or so. Of course, you have meanwhile a smaller part of clients moving from managed service to more automated booking processes. I think that's just underlying adoption of both agencies and clients using the more automated version. Still, the biggest chunk of the incremental programmatic digital out-of-home business is coming really from new customers that predominantly find the way to our medium, either being already an online customer on our SSP and just extending the online portfolio that they already buy to digital out-of-home, or from TV clients that come from TV only or TV plus online video and expand that footprint to digital out-of-home. Very often, they go the programmatic way just to have a better audience measurement picture across all the eyeballs that they get from the various video platforms.
Thank you.
Ladies and gentlemen, that was the last question. I would now like to turn the conference back over to Christian Schmalzl for any closing remarks.
Thank you very much for spending the time with us and especially spending the time to dig a little bit deeper into the current promising momentum around Statista. Have a good remainder of the summer and hope to speak to you soon. Take care. Bye bye.
Ladies and gentlemen, the conference is now over. Thank you for choosing Ströer and thank you for participating in the conference. You may now disconnect your lines. Goodbye.