Ladies and gentlemen, welcome to the Scout24 preliminary full year 2025 results conference call. I'm Moritz, the call's call operator. I would like to remind you that all participants will be in a listen-only mode, and the conference is being recorded. The presentation will be followed by a question and answer 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 Filip Lindvall, Vice President, Group Strategy and Investor Relations.
Good afternoon, everyone. Welcome to Scout24's earnings call for the preliminary fourth quarter and full year 2025 results. My name is Filip Lindvall. I'm Vice President, Group Strategy and Investor Relations at Scout24. With me on the call today are Ralf Weitz, our Chief Executive Officer. Dirk Schmelzer, our Chief Financial Officer. Ralf will start the presentation with key business highlights. Dirk will provide a detailed overview of our financial results. As always, we will conclude the call with a Q&A session. You can find today's presentation on our website under Financial Reports and Presentations. This session will be recorded. A replay will be made available as quickly as possible after the event. Please take note of the disclaimer on page 2. Ralf, now over to you.
Thank you, Filip, and welcome, everyone. Let's turn to page 4. 2025 was another great year for Scout24. We delivered strong financial results while continuing to execute our interconnectivity strategy. Revenue grew by 15%. Ordinary operating EBITDA increased by 17%. Adjusted EPS rose by 20%. Growth was driven by our B2B and B2C subscription businesses. Our ecosystem expanded across all key metrics. We gained more customers and more content. We introduced better and more innovative products. User engagement increased. Cross-selling continued to grow. In 2025, we further strengthened our leadership in search. AI is now integrated across the entire customer journey, from semantic and elastic capabilities to agentic experiences through HeyImmo and our ImmoScout24 app in ChatGPT. Internally, we continue to embrace technology. We simplified processes and embedded AI into our workflows. Combined with healthier revenue growth, this drove higher profitability.
Our margin increased to 62.5%. Looking ahead to 2026, we expect revenue growth of 16%-18% and an ordinary operating EBITDA margin of up to 61%. Scout24 is more relevant, stronger, and better positioned than ever to power the German real estate ecosystem. Technology and AI will continue to enhance our platform and strengthen our competitive position. Let's turn to page 5 and look at customer development. Starting with our professional segment. We closed the year with 26,400 professional customers. This represents growth of 5.1% in Q4 and 5.7% for the full year. Clearly, industry-leading rates. Our responsible pricing approach is working. Churn is at a minimum, and we see virtually no net churn.
Our success in B2B is driven by a strong and integrated product portfolio, including our market-leading agent software, Propstack. We are deeply embedded in agent workflows through software, data, valuation, and marketing tools. Our Bronze, Silver, and Gold membership tiers serve both smaller and larger agencies. In 2025, we spent more time than ever with our B2B customers. This strengthens relationships and positions Scout24 increasingly as a business enabler. 2026 has started well. We are adding customers and continue to see healthy revenue growth across the base, including non-residential. Turning to the private segment. We ended the year with 503.6 thousand subscribers, representing 14% growth for the full year. Growth moderated in Q4 due to normal year-end seasonality, softer rental demand, and following the rollout of our new group-wide ERP system.
We are currently updating the private subscriptions offering, including product tiers, features, and pricing. Looking at page 6, let me walk you through the German real estate market dynamics. Our Scout Transaction Momentum Index stands at 97. The residential transaction market has stabilized. Based on preliminary data, 2025 recorded around 600,000 transactions, up 14% compared to 2024. This environment supports our residential B2B membership business. We are also seeing improving trends in commercial as we enter 2026. Our listings index reached 146 in January. Inventory continues to grow following the normal seasonal dip in December. Approachable content grew to 17 million objects. This reflects improving activity and strong trust in the ImmoScout24 platform. On the demand side, rental contact requests moderated slightly from very high levels as rents increased.
Buyer side, contact requests remain stable at elevated levels, indicating healthy demand to buy. Let me now provide a strategic update. Over the past year, we have continued to execute our strategy with discipline while embracing technology to stay ahead. In the current market environment, it is important to understand what differentiates Scout24 and why our competitive position continues to strengthen. Let me highlight five points. First, we have accelerated execution of our interconnectivity strategy. Second, we have expanded our AI-native search and extended audience reach across the entire user journey. Third, we have built the market-leading software platform for German real estate agents. Fourth, we are driving structural margin expansion through technology and AI. Fifth, we are outperforming our CMD 2024 targets and raising ambition again at CMD 2026. Let me walk you through each of these.
Turning to page 8, let me be very clear, Scout24 is not only a listing platform, we have built the digital ecosystem for German real estate transactions. Over the past years, we have invested more than EUR 400 million and executed our interconnectivity strategy with discipline to achieve this. At the core, sits our proprietary backend. We operate the leading agent software, subscription products across the buy and rent journey, and Germany's leading residential and commercial real estate database. More than 10% of residential units are registered in our property hub. We also provide valuation and banking software used by financial institutions. These systems are deeply embedded in daily workflows. They are exclusive to Scout24. They are not accessible to AI models or web crawlers. On top of this infrastructure, we connect all stakeholders. We reach 20 million users per month and serve 26,000 professional customers.
Around 25% of annual real estate transactions in Germany now involve a Scout24 B2C subscription product. As the ecosystem grows, our data becomes stronger. As our data becomes stronger, our products improve. This is digital infrastructure powering German real estate. Scout24 is embedded in how the market operates. At our Capital Markets Day in early 2024, we outlined our vision for AI-driven natural language search. Today, that vision is fully live across semantic, elastic, and agentic search. First, AI-powered filter search. User describe what they want in natural language and receive relevant results instantly, including image-based search. Second, HeyImmo, our specialized real estate LLM. It is built on 28 years of proprietary behavioral and transaction data, combined with exclusive residential and commercial inventory across our ecosystem. This enables contextual accuracy that generic LLMs cannot replicate. We are seeing strong user growth and engagement.
Third, we integrate with external LLMs as audience extension. Through our ChatGPT app integration, ImmoScout connects users directly back to our platform for engagement and transaction execution. Referral traffic from LLMs remains insignificant at 0.4% in December and below last year's peak levels. General AI models will need trusted platforms with proprietary data to deliver a strong real estate experience. That is where Scout24 plays a central role. Our backend for agents is Propstack, our fully cloud-based software platform. This is where agents run their business. They manage contacts, listings, marketing, and their pipeline. It also serves as a marketplace for Scout24 products and partner integrations, many powered by AI. AI is embedded directly into these workflows. More than 30,000 AI actions have already been executed. Voice to Listing reduces listing creation time by over 80%. These are real productivity gains.
We estimate our market share in Germany at around 30%, and it is growing. We are on track to become market leader. Propstack significantly increases our relevance by covering a large part of the agent value chain, and we are taking the next step. Our ambition is clear: to transform Propstack into agentic AI, software that not only documents workflows but actively drives them. Software that helps agents prioritize leads, draft communication, prepare listings, and move transactions forward. We will share more at our Capital Markets Day in 2026. At our Capital Markets Day in 2024, we committed to building a scalable technology platform and deploying AI internally to improve efficiency. Today, we operate a very cost-effective tech platform and one integrated builder organization serving all three customer groups. We build our product solutions centrally and deploy them across consumers, professionals, and homeowners.
This allows us to grow without increasing complexity or capacity. The impact is visible in our cost structure. Personnel, our largest cost bucket, has remained broadly stable in an organic level over the past three years, despite continued revenue growth. At the same time, we continue to invest in technology and AI to enable that leverage. Even as we invest further, IT costs grow at a disciplined mid-single-digit rate and represents only around 9% of our total cost base. The structural leverage is reflected in our margins. Ordinary operating EBITDA increased from 59.7% in 2023 to 62.5% in 2025, and we continue to see further margin expansion over the next years. Technology and AI are structurally increasing our profitability. Turning to page 12. Our CMD 2026 will define the next chapter of Scout24.
We have evolved beyond classifieds into the digital back end of German real estate. AI is strengthening our platform and our proprietary data advantage. As our products become more personalized and more automated, our ecosystem becomes more valuable, and our competitive moats deepen. We will present the next evolution of our ecosystem, the acceleration B2C expansion, and further scaling of our B2B leadership. We will also introduce updated midterm targets for 2027-2029, including higher margin ambition. Let me close with a few key takeaways on page 13. In 2025, we delivered what we set out to deliver. That is Scout24. We set clear targets, and we execute, while continuing to invest in technology and AI. Our ecosystem is increasingly central to the German real estate market. We are deeply embedded across stakeholders, workflows, and data, and that position continues to strengthen.
We have shown that disciplined investment and margin expansion are not mutually exclusive. Innovation and profitability reinforce each other. AI is already making our products more personalized and automated. It increases the value of our data and strengthens our competitive moats. With this foundation, we expect to deliver again in 2026. At our next CMD, we will outline the next chapter, including further margin ambition. Before handing over to Dirk, I would like to thank him personally. Working with you over the past 6 years has been a pleasure. You have powered our financial profile, expanded margins, improved cash generation, and built a strong finance organization. Dirk, thank you for your leadership, your partnership, and your commitment. On a personal note and on behalf of the entire Scout24 leadership team, we wish you all the very best in your next chapter. With that, I hand over to you.
Thank you, Ralf, and good afternoon, everyone. 2025 was an excellent financial year for Scout24. Revenue grew 15% to EUR 649.6 million. Ordinary operating EBITDA increased 16.5%, and adjusted EPS rose 19.6%. We delivered strong operating leverage. The margin expanded by 100 basis points to 62.5%, while continuing to invest in product, AI, and the integration of acquisitions. Cash generation remains a core strength of our model. Operating cash flow reached EUR 284.8 million, up 11%. Turning to page 16 and the professional segment. Revenue grew 14.8% in 2025 to EUR 470.5 million, driven by subscription growth of 15.4%. Average customers increased 5.7% to 26,027.
We completed over 5,000 migrations, bringing adoption of the new membership model to around 60% of the base. Growth was particularly strong in the Bronze tier, creating further up-migration opportunities ahead. Subscription momentum and new customer wins have continued into January and February 2026. ARPU increased 9.5% for the full year and accelerated further in Q4. Transaction enablement revenue grew 17.5%, supported by CRM expansion and M&A contributions. Overall lead demand remains muted. Homeowner lead products showed solid momentum, growing 9% in 2025. Ordinary operating EBITDA rose 14.5% to EUR 292.9 million. The professional ordinary operating EBITDA margin remained strong at 62.3% for full year 2025. In Q4, the margin increased to 63.4%.
Acquisition-related dilution was fully offset through operational improvements, keeping margins at a structurally high level. Turning to the private segment on page 17. The segment delivered strong performance in 2025. Revenue increased 14.5% to EUR 179 million, driven by subscription growth of 18.8% and strong PPA performance. The average customer base expanded 14% year-over-year to approximately 507,000. Q4 included a temporary adjustment related to our system migration. ARPU increased 4.2% for the full year. As expected, growth moderated in Q4 as we lapped the Schufa monetization benefit. PPA delivered a strong performance, growing 10.9% for the full year and accelerating further in Q4. Growth was supported by improving market activity, product simplification, increasing brand strengths, and targeted marketing initiatives.
Ordinary operating EBITDA increased 22.3% to EUR 112.8 million. Margin expanded by 4 percentage points to 63%, demonstrating the scalability of the subscription model and strong operating leverage. Turning to page 18. Operating expenses increased 9.6% in 2025, largely driven by the consolidation of recent acquisitions. Organically, cost growth remained disciplined. Personnel costs increased on a reported basis due to M&A, but remained broadly stable organically, despite continued revenue growth and investment in technology and AI. To build on Ralf's remarks, we continue to invest in AI and technology while keeping IT costs under control. IT expenses increased 14% for the full year and only 4% in Q4, despite the significant product launches you have seen across search and other areas. Marketing expenses were organically flat, with efficiency gains offsetting selective brand investments.
Purchasing costs grew most strongly, driven by higher valuation volumes and increased service components within B2C memberships. Overall, we maintained cost discipline while investing strategically. As a result, ordinary operating EBITDA increased 16.5% to EUR 405.7 million, and the margin expanded by 100 basis points to 62.5%, demonstrating continued operating leverage. Turning to page 19, where we show the items below ordinary operating EBITDA. Non-operating effects increased in 2025, driven by M&A-related expenses and higher share-based compensation linked to share price development. I will comment on these positions on the next page. D&A increased moderately, mainly reflecting acquisition-related PPA amortization. The financial result improved compared to last year, supported by lower subsequent measurement effects on M&A purchase price liabilities and reduced interest expenses.
This was partially offset by negative foreign exchange effects due to the depreciation of the US dollar. On taxes, the year benefited from EUR 46 million one-time deferred tax gain following the reduction of the German corporate tax rate from 2028 onwards. This led to a temporary reduction of the effective tax rate to 14.5% in 2025. Net income increased to EUR 240 million, and basic EPS rose to EUR 3.33. Adjusted net income and adjusted EPS also grew strongly, with adjusted EPS up 19.6% to EUR 3.47. Finally, the weighted average share count declined by 1.4% due to our share buyback program, further supporting EPS growth. Now let's turn to page 20 and walk through the bridge from reported to adjusted net income.
Non-operating effects, excluding share-based compensation, increased during the year, primarily driven by M&A-related items, including earn-out revaluations related to Sprengnetter and Neubaur Kompass, as well as transaction costs associated with the Spain acquisition. Share-based compensation increased year-on-year, mainly driven by higher share price levels during most of 2025, as well as strong performance factors. We expect share-based compensation to decline in 2026, which I will address in more detail on the guidance slide. The tax bar reflects the reversal of the one-off time deferred tax revaluation benefit recorded in 2025. Importantly, the majority of these non-operating effects are non-cash in nature. Speaking of cash, now turning to page 21 and cash flow. Free cash flow for the full year 2025 amounted to EUR 253.1 million, representing growth of 13% year-on-year.
This reflects our strong operating performance, solid cash conversion, and the fact that a significant portion of non-operating costs are non-cash items. Free cash flow conversion remains strong, corresponding to 101% of adjusted net income and 62% of ordinary operating EBITDA. Turning to page 22, to leverage and capital allocation. At year-end 2025, net debt stood at EUR 144.5 million, resulting in a leverage ratio of 0.36x, supported by strong cash generation and disciplined capital management. In 2025, we returned significant capital to shareholders. We allocated EUR 124.3 million to share buybacks over the full year and paid a dividend of EUR 95.4 million.
As of Monday, we have already executed EUR 47 million of the initial EUR 100 million share buyback program, demonstrating our continued commitment to disciplined capital return. Moving to the guidance on page 23. For the 2026 financial year, we expect group revenue growth in the range of 16%-18%. 6 to 7 percentage points of that growth are expected to come from Spain. This reflects approximately 10 months of contribution, as we expect to close the transaction tomorrow. In terms of profitability, we expect the group ordinary operating EBITDA margin to be up to 61%. Excluding Spain, the organic ordinary operating EBITDA margin is expected to be up to 64%, reflecting continued operating leverage in our core German business. A few phasing comments to 2026.
Our B2B membership business has started the year strongly, supported by sustained demand and continued new customer acquisition. Churn entering 2026 is lower compared to the prior year, providing a solid foundation. In B2C, we are testing new product tiering and pricing initiatives. This may impact growth rates in the first quarter. As a result, organic growth in Q1 might be a bit softer, with acceleration expected from the second quarter onwards. Overall, we expect growth to build as the year progresses, consistent with our full year guidance. Regarding share-based compensation, we expect cost for 2026 to be in the range of EUR 15 million-EUR 20 million. If the share price remains at current levels, the lower end of the range is more likely.
On the implied ordinary operating EBITDA contribution from Spain, this includes one-off transition-related costs, such as TSA arrangements, which are recorded as operating expenses and therefore impact ordinary operating EBITDA. These effects are expected to unwind from 2027 onwards. Overall, we are confident in our 2026 guidance and the momentum across the business. In particular, the up to 64% organic ordinary operating EBITDA margin exceeds the 2026 target communicated at CMD 2024. We will provide an updated, increased midterm margin ambition at our Capital Markets Day in May. Turning to page 24. Our Capital Markets Day will take place on May 12th, 2026. As Ralf mentioned earlier, the company will present its updated strategic framework and midterm financial ambition in more detail. It will provide a comprehensive view on the next stage of Scout24's development.
Before we conclude, I would like to say a few personal words. After more than six years as CFO of Scout24, this will be my final earnings call. It has been a privilege to help shape the financial and strategic development of this company during a period of strong growth and transformation. Together, we accelerated growth, strengthened profitability, improved capital allocation discipline, expanded our ecosystem, and positioned Scout24 as a scalable, resilient business. I would like to thank Ralf, Gesa, and the entire leadership team for the close and constructive partnership over the years. Our collaboration has been built on trust, strategic clarity, and disciplined execution. It has been fun working with you. I would like to thank my directs and the entire Scout24 team for their dedication and execution, and of course, our shareholders, for their trust and support.
Scout24 is in excellent shape, strategically and financially, and I'm confident the company will continue its successful trajectory. Thank you. With that, let's open for questions. Please limit to two questions per speaker. Operator, over to you.
Ladies and gentlemen, we will now begin the question and answer session. Anyone who wishes to ask a question may press star and one on their touchtone 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. In the interest of time, please limit yourself to two questions. Anyone with a question may press star and one at this time. One moment for the first question, please. The first question comes from Ed Young from Morgan Stanley. Please go ahead.
Thank you very much. My first question: could you perhaps give a bit more color on the private customer growth development in Q4? You mentioned a system migration impact, also seasonality and user demand. I wonder if you could help quantify what that might be, or build a relative impact of those three different parts, and how we should think about the timeline for new products and tiering there. Second of all, on buyback and capital returns, just wonder if there was any updated thoughts from you, given the divergence between the share price performance, the sort of overall center in the sector, compared to your continued confidence in the business. Thanks.
Hi, Ed. Thanks very much. It's Dirk. Let me start off with the second question you raised with regards to capital allocation. As you are probably aware, we announced a EUR 500 million buyback program in December last year. We started that with the first tranche of EUR 100 million early January this year, and we're now down the road, I think of EUR 50 million roundabout that we already bought back.
... Of course, you can imagine that we, as executives of this company, are convinced that the intrinsic value of the company is higher than what we see at the moment, given the recent backdrops at the capital markets. We are still discussing and continue to discuss additional share buyback programs with the board and within the management board, and we will update you as soon as we have new information. On the private growth with regards to third and fourth quarter, some of you know that we changed our ERP system last year, and the biggest chunk of change here was the order to cash module, and we changed that in the transition from the third to the fourth quarter.
In that transition, there were a few customers which we kept in the systems in the third quarter and only deleted out of the system, so to say, in their data in the fourth quarter, and that came to a slight change and a slight delay. If you look at the numbers around that, I think you should deduct around 2 percentage points in Q3 and put another 2 percentage points in Q4 on growth. That will bring you roughly there where we are. Apart from that, I think we're quite optimistic when it comes to the 2026 growth in private subscribers. Ralf will elaborate a bit more on the initiatives we're putting forward there. For this, for the second part of the first question, I hand over to Ralf.
Yeah. I think, we also reported a bit to you last year that there is a kind of saturation if it comes to the private Plus subscription product. Saturation means not that we are not able to grow in subscriptions because subscription numbers are driven by lifetime in particular, and this is actually what we're working on. The other thing we're working on is actually the tiering, the product tiering. I mean, this product now is the market standard here in Germany, if it comes to real estate search trend. The question is, how can I differentiate as a seeker? We will come up with new product tiers, and we are testing those actively.
We started last year. We continue to do this in Q1. What I can say is we see a positive positive impact here to all the KPIs. We are quite optimistic that we can, let's say, launch another rocket for our Plus subscription business. Yeah, that's what I mean. So far, we haven't had a tiering here, as you know. We follow actually the path what we did over the last years in the professional business. It's actually a well-known, let's say, playbook to us. Yeah, that's how you should read it. There's no decline in the subscription base, as Dirk said, right? We changed the ERP system.
I mean, the customers, the subscribers been there. We had a payment delay process we changed with the ERP system. That was actually the reason why we, saw the numbers declining in Q4.
Thank you.
The next question comes from Will Packer, from BNP Paribas. Please go ahead.
Hi, thanks for taking my questions, and Dirk, a big congrats on the Spinters CFO. 2 from me, please. Firstly, you've entered into a wide-ranging partnership with OpenAI across HeyImmo, Enterprise, and of course, the ChatGPT app. Whilst I wouldn't imagine you share specific commercial details, could you share some details around the app? In particular, around data sharing, for example, to what extent do they get insight into your inventory and consumer behavior, or is that in a Scout walled garden? Secondly, in terms of the guidance for Spain for 2026, the margin looks a fair bit lighter than consensus expectations. I imagine the deal timing is a factor, but could you help us through the factors that are weighing on margin in 2026 and how that will impact 2027? Thank you.
Yeah, I can start, Dirk. Maybe you can jump in then. To your first question, regarding the ChatGPT cooperation here. First of all, we don't pay for this app, right? I mean, it's we built the app. This is actually an API wrapper, we call it. You get access via API to our listing, let's say, database. We are not giving access to proprietary data we have in the system. This is guarded behind lock and vaults and everything. If you want to contact somebody on the listing and so on, you have to do it in our ecosystem. AI wrapper means it's actually you funnel the traffic from ChatGPT into our system directly.
It's not that they're taking over customer data, ever, it's just they are able to search, to execute a search on our listing database, which is publicly available anyway, more or less. If you want to do more, you execute everything on our website. You can experience that, right? You start the search, you get the results immediately. If you click on the results, you land immediately on our website, you also continue there the search. Actually, what we see so far is that the traction is okay-ish, but it's not really there's no exponential growth as people might expect, I don't know.
We see traction in our HeyImmo product. That's more a couple of 100,000 now using it. There's some exponential growth in our own product because the experience here is much better. This is, and the HeyImmo is also getting limited access to proprietary data if people are logged in, for instance. Yeah, the partnership is good. We actually took the first mover advantage here as we did with the Apple ecosystem a while ago. We are quite happy here. Second question was.
About the Spanish development where.
Uh,
Yeah, maybe I start, Ralf?
Yeah.
Maybe I start with that. Ralf will certainly add something to that. With regards to guidance, I mean, we're getting the keys for the business tomorrow. Give us some time to take a look under the into the motor of the business. As of today, we're talking about 10 months instead of 12 months that we're getting the business. We've been guiding something around EUR 60 million of revenues. If I read it correctly, last year ended on plan. January started okay-ish in Spain. Most importantly, I think what you need to take into account is our constant message that we are deploying the German playbook when we are entering into Spain with Fotocasa.
That also means that we will take some changes on the user interface and the user experience on the platform, which we did in Germany a few years ago as well. That had the effect that advertising revenues were going down. Today, advertising revenues in Spain are around 10% of revenue, and you can imagine us driving that down very fast to a level that we see in Germany, which is below 2%. That has an impact this year already of around EUR 5 million-6 million, and that's what you see reflected in our guidance with regards to Spain. I hope that clarifies.
What you also have to see, I mean, we're doing a carve out of a carve out, right?
Yeah.
There might be parallel costs for doing this because we have the TSAs we have to pay, and parallel, we have to build the systems we need in order to take over. This will be relevant for this year, not for next year. Hopefully we can finish most of the TSAs end of this year. That's actually the plan, but as Dirk mentioned, right, we just get the keys. We will get the keys tomorrow or so. We lost two months this year. Let's see. Yeah. It's on plan, and for us, everything is in our expectations, let's say, and now we look more into it and driving actively the business from beginning of all.
Thanks, thanks for the color. In terms of the 2027 margin outlook for Spain, that's something that comes with the CMD. Once you've had some more time with the business, because some of those costs will unwind, right?
If we have something to tell you in May, we can rely on, I would do it. Give us a bit time here. I mean, this is. We said in the midterm perspective, the business should be able to deliver 40% margin. That's what we said, and we still believe that. As I said, also, we have to stabilize the business here. For us, it's more important that we deliver on the key metrics, and key metrics are the customer numbers here. It's the listing numbers, it's traffic up to here is in particular, and also leveraging the synergies we have. As I said, right there's also benefit for German customers.
That's what we want to deliver on, and then hopefully revenue and margin applies in as we expect. Executing the playbook, I think what Dirk mentioned is correct, and that is actually also a challenge for us because it's not that easy to take over a business, 300 people, in such a situation. We stay optimistic here because everything is as we expected so far. Then we give you an update if we, if we can, maybe in May.
Thanks for the color.
The next question comes from Joseph Barnet-Lamb from UBS. Please go ahead.
Hi. Yes, Joe from UBS. Firstly, I'd like to add my congratulations on your tenure, Dirk, and all the best for the future. A couple of private questions, if I may. Firstly, should we read into the launch of the new tiering that you're pivoting to a more ARPU-driven growth strategy due to volume-based growth running out? Is this about elongating subscription duration by giving a variance of products to consumers? Can you sustain Teams growth in private post the product launches? The second question is on private margin. It increased significantly in 4Q. Was this at all related to the slower revenue growth in the quarter? Maybe asking the question differently, is there any reason why margin within private wouldn't be sustainable for FY 2026 or even expand further? Thank you.
I'll start with the second question and hand over to Ralf for the first part of the question. No, there's no reason why the margin in private should change. As you've probably seen, we came from a margin of 30%-40% in the early innings of the product, we now reach professional margins here. I think that's a good sign, and that is something that will continue. Joe, it's a good sign that our analysts understand the business so well that they answer their questions themselves. Nonetheless, I would hand over to Ralf to answer the remainder.
I think it's a really good question, to be honest. I mean, this, you have two key metrics you can optimize. As revenue growth is created by number of customers and the prospective subscribers, and the ARPU growth, so the pricing power you might have. On the pricing power side, we've been quite conservative in the past because we grow heavily on the customer side. And this is something we're going to change with our tiering here. There will be hopefully more pricing power in the future on this, because we just had 4.5% ARPU growth last year, blended, and we had 14% subscriber growth last year.
The year before in 2024, it was over 25. I think what we can also do on the subscriber side, on the number of customer side, and this will be iterated sometimes, is that if you extend the lifetime, today we have lifetime close to 6. If you double the lifetime, then you would double the number of subscribers, number of customers who are paying at the same moment. This is also where we have some initiatives on, but it's a bit tricky, right? You need to add also new product features to those product tiers as well to make the difference here, but also to extend at the same time, the lifetime. This is actually what we're working on, and we are testing it.
Maybe it's... To why is it so difficult? I mean, nobody really has this product, so there's no playbook we can copy or so, where there's not much experience on this, right? We are the market maker here, and therefore, we have to learn, and that takes time a bit. We started last year in Q3, end of Q3 with the testings and we continued this in Q4, and we will continue this in Q1. We are making progress, that's actually the message I would like to send. We have an interest that we create sustainable growth here, and at the same time, we can deliver the margin.
The margin, you might remember, we optimized the margin because we changed a bit the products here. From products we used from third-party vendors into products we developed on our own capabilities, and that led also to this margin expansion. This is also now part in the new product here, that we have to come up with own products, new own products, and also to maintain the margin profile we delivered already. You can see it's not easy to run such a business, but we are, as I said, positive. The first signs we see now will hopefully accelerate the subscriber growth, but also the ARPU growth in Q2.
Excellent. Thank you very much.
The next question comes from Marcus Diebel from J.P. Morgan. Please go ahead.
Hi, everyone. Yeah, I would also echo that, Dirk, thank you very much, and Tili good luck in your new endeavors. First question is again, on private, just to be really clear, so sort of like the comment on sort of like the some bad debts in there. That effect will basically stop in Q4, so we won't really see much of this in Q1 and Q2 going forward. I think that's the case, but just wanted to be really clear on this. And secondly, also for Dirk, I guess, on personal costs. I mean, clearly quite an impressive performance, 4% only increase. Could you talk a little bit more about the moving parts here?
Where is the efficiency really coming from, and how should we think about personal costs going forward, in the next quarters? Is it volume? Is it number of headcount? Is it salary? If you can just explain a bit more where the efficiency is coming from. Thank you.
Thanks, Marcus. First of all, to your first question, I can confirm that the migration to the platform has been finished in Q4. All customer numbers and everything related to non-paying customers and bad debt, so there will be no effects leaping into the first quarter, 2026. On the efficiency side, I think it's quite an interesting discussion here. I would like to remind you to the discussions we had in 2024, when there were a lot of companies standing up and saying, "Oh, from AI, we will achieve 25% savings, 30% savings on headcount costs in the next years because our efficiency will increase in engineering, our efficiency will increase in programming, product development, and everything else." As you are aware, we took a slightly different approach here.
We said at a certain point in time, that was Q3 2024, where we said, "Okay, we're sticking to the full-time equivalents we have in the company," and that is around 1,000. "We are not growing here, but we're growing revenues and we're growing profitability." Hence, what we did was we gave every one of our employees a second employee. In 2025, that second employee was called Claude, when we had the arrangement with Anthropic. This year it's called ChatGPT, 'cause we've changed the arrangement from Anthropic to OpenAI. Now, AI is structurally embedded across the organization. We have a huge amount of users. We have 1,200 AI-driven projects company-wide.
A lot of sort of what's going on in AI is really helping us to maintain that amount of 1,000 full-time equivalents. That is the approach we have been taking, that is the approach we have been communicating with you, and that is the approach we will continue in 2026 and 2027. You shouldn't expect our personal cost to increase beyond any merit increases over the years to come.
Yeah, great. Thank you.
The next question comes from Craig Abbott from Kepler Cheuvreux. Please go ahead.
Yeah. Hi, good afternoon, everyone, and also Dirk, from my side, all the best going forward. Yeah, two follow-up questions, please. Just turning over, first of all, to the traffic sourcing mix again, could you just confirm that your direct traffic share has been stable, still around 80%, I think, and that the cost of the inorganic traffic is also stable? Excuse me. Secondly, also, as you continue to roll out your AI product suite, can you also confirm that you do not expect a material increase in your product development cost? Thank you.
Yeah, I can confirm all three.
Craig, that was a short answer.
No, no, please elaborate a bit.
Yeah, I think, I mean, on the traffic side, there's what we see is that there is no change here when it comes to organic traffic. I mean, yeah, as you know, quite strong on the brand awareness, this hasn't changed. We don't expect that there will be a change in the organic traffic. Direct traffic, it's still over 80% and dominated by the app traffic in particular here, which is the main organic traffic source. On the paid side, I mean, there's no paid traffic channel in LLMs at the moment. We do, of course, some marketing here and there to integrate ourselves better, to become more visible in LLM sources. That's what we do, and that costs a bit of money.
We expect if this is, if they are models, they will replace other traffic sources like Google Search in particular. We don't expect additional paid marketing costs for our model. The last question is whether we expect more AI costs if we are going to roll out our AI suites to the market. That's of course. I mean, if you, if we have more, let's say, LLM calculations in the system, that would drive a bit cost. As I iterated in one of the other meetings is that, we had before that, actually, the way how we use LLMs is a bit different. We have most of the data in our own databases.
It's called vector databases, we have different access, even if people execute an LLM search. We have the cost side here under control and token prices, by the way, dropped by 30% in Q4 last year. Because of the competition we have in LLMs, there are also arguments why the costs will not going up dramatically here, even if the usage would go up exponential. Therefore, we are quite confident that we can manage the cost side would be for the market, right?
Yeah. Craig, maybe a short advertising break on my end, for Andrew Ross's report from Barclays on LLM traffic on classifieds, because this report mirrors exactly what we are seeing on our platform at the moment. The LLM traffic is roughly stable to slightly decreasing and constantly remaining below 1% on our platform. We believe that this will sort of not go into an exponential growth by the end of this year or next year.
External, yeah.
Yeah, that's the external traffic, Ross. corrected me here. Yes, that's the external traffic. The HeyImmo traffic that we're seeing is a positive for us, to be honest, right? It's a great customer experience, and I elaborated on the cost for HeyImmo versus OpenAI/ChatGPT, and that would cost us below EUR 1 million this year for a great customer experience. That's good.
Yeah. Okay, thank you both very much.
The next question comes from Fathima-Nizla Naizer from Deutsche Bank. Please go ahead.
Hey, thanks, and all the best to you, Dirk, from my end as well. Just a couple of questions for me. On the professional segment, ARPU growth, when you think of 2026, sort of how is that phasing over the course of the year, given the strong ARPU growth you reported in 2025? Some color there would be great. The second question is on PropTech. It was really good to see that you've got such a good penetration among the agent base with PropTech. Where do you think that can sort of expand to? And could you remind us again, how is it monetized? Is it a monthly subscription?
you know, how profitable is PropTech, and how much of revenue does it contribute at the moment, and what's your sort of objective for that segment going forward, given that's, you know, fundamentally different to the rest of the classifieds, business that you do? Thank you.
I thought, Dirk, maybe you take over with it. If it comes to PropTech, we are quite heavy, happy, on PropTech here. We are on the way to become clear market leader if it comes to agent software systems in Germany. Why is it important? For the agents, they interact, engage, they engage much more intensive with the CM solutions than they do with listing marketplaces, because they also save the customer data, for instance, in agent software solutions like PropTech. They prepare the listings there and so on. This is often the starting point of the agent. With the agent software product, we mirror the business of an agent in a software, in a digital format.
We have a clear strategy here. We would like to become market leader this year in terms of customer numbers, in terms of listing imports, so we can measure how many listings on the portal are imported via our own agent software, and we're making progress here, which is really, really great. The way how we monetize it is, of course, the agents, depending a bit on the size of the agent business, they have to pay a monthly subscription fee. Here, it's a classical SaaS model. We're also combining it, depending on the packages, with the membership products here and there.
Actually, what the product does is, we are creating much more loyalty, because if customer not just using the listing portal membership, they are also using our technology, our system for running their business, day to day. Of course, we're creating a kind of loyalty and pricing power, and this is actually a big advantage. What we achieved in 2025 is that the number of customers who are using more than just one product in our ecosystem increased. This is actually something we would like to continue. For this, we have PropTech, our agent software system, but also other products, valuation products, for instance, we have on the energy certificate side, we have products.
We are building and creating this product suite for the agents more and more, and it's well accepted, I have to say. Of course, agents, they see also benefits because, I mean, everything is integrated, it's connected, and if you sum up the product and the, let's say, the prices you would need to pay standalone for each product, there is an advantage if you bundle it, right? And therefore, we are able to bring over more and more customers also from competitors into this, into those products, and that's quite positive. The other question, Dirk.
This one was around ARPU development over 2026 and where do we see it? What I can say is that we, at the moment, envisage a similar development that we saw in 2025. We were very positively surprised by our customer numbers in January, February. I was a bit more conservative when I guided you on that in the Q3 call. I have to say, January, February start was good. We saw a very low churn. We're seeing good migration efforts from our sales teams, and we're seeing a healthy development on ARPU. All in all, I think that business should develop nicely over the course of the year, and expect an ARPU growth in the high single digits.
Thank you. Very helpful.
The next question comes from Giles Thorne from Jefferies. Please go ahead.
Thank you. My first question was on your partnership with OpenAI, and specifically, it'd be useful to know what the cost profile or the commercial agreement would have looked like for the powering of the HeyImmo products and then the app SDK integration if you hadn't done a migration from Anthropic to ChatGPT. Did you get a good price because of that migration? Secondly, we're clearly in a very volatile and febrile environment for investor sentiment towards the sector and to Scout24. There's no consensus on what happens here next. Perhaps the one area of consensus is that you want to be a well-invested platform with a progressive mindset. Given all that, does the current environment and the risks and opportunities around AI, Ralf, change your attitude towards the ideas of regional scale? Thanks.
Yeah, I can. A lot of questions, so be not sure whether, I'm not sure whether I really understood everything, but let me start with the ChatGPT agreement we had. I mean, as you, as you know, we, so we cooperated internally with Claude in topics in particular, and then, and also with other LLMs, and then for our HeyImmo product, we decided to go with ChatGPT, OpenAI here. What we then learned, I mean, of course, we checked a bit on the usage of the product, and we agreed on token prices at the beginning of the year.
What we learned over the year is that there is high competition between the different LLM systems, and we see also that some of the LLMs, they are specializing. For instance, OpenAI is really specialized more on the consumer experience side, so it's, for us, the right partner for HeyImmo at the moment, I would say. We're using the technology, we are fencing the data. That's actually the deal here. So as we checked in on prices, even for our internal, let's say, AI system we are using for the organization, we learned really that we can lower the prices by 30% if we are shifting to, let's say, OpenAI here.
Anthropic, and that is maybe something you would like to know then, is that Anthropic also reacted, and they dropped prices also, even higher. You can see that there's high competition. Everyone wants to have us as B customers, and that's actually what we see, that those LLMs, they learned over the last year, it's really hard to monetize the consumer, so sometimes it's better you just monetize the B side of the business, as Google did, as always did, right? They are really hunting for B customers, and that's what we see, and therefore, we see also competitive prices here. On the deal we have, if it comes to our SDK product, I mean, we've been one of the first companies who launched in an app here, an own app.
We interacted quite early after Zillow launched the product here in the U.S. We interacted with the team here in Germany to learn a bit more about how the app SDK really works. As I said before, it's an API wrapper, there is no really commercial agreement. I mean, we also have no really commercial agreement with Apple here other than if there is a payment aggregate to share, if we are using their payment methods, but what we don't do. There is no extra cost for this app SDK. That might change, who knows?
At the moment, I think what we see from those LLMs, they have high interest, really high interest to cooperate with companies like ours because they need to create a really unique user experience, for this, they need data. If you're searching for real estate, they cannot really give access to this data, for this, they need us, they need us as a partner so that we give them better access than maybe they would have if they just need to pull the data. That's actually also in their interest, that they cooperate with reliable, that they do collaboration with reliable partners. Therefore, it's more like a partnership than a dependency. That's I would say good and positive here. The other question, maybe remind me.
Giles, I think we're through, right?
No, the second question, I was using too many words, and let me put it simply: Does the current very accelerated innovation cycle change your attitude towards the benefits of regional scale?
If I translate this, Giles, your thesis is that if you are able to accelerate the amount of innovation and product development, there should be more international scale, right? Is that your hypothesis?
No, it's the other way around. It's the other way around. If you're a bigger business, then your capacity to invest is bigger, all else being equal.
Okay. Okay.
Yeah.
Got it. I think of, I'm starting with-
I think the answer is no.
... the answer to that question. I think we're well there with 3% of our revenues annually invested into product development and tech, and that is around EUR 20 million. We'll accelerate a bit, maybe to EUR 21 million-EUR 22 million, but we never had any year where we had the feeling we didn't have enough capacity to invest. The last time I recall that we had that discussion was when we migrated to the cloud, but that's four, five years ago. There we still had 7% CapEx to a revenue ratio. At the moment, I don't see any need to accelerate our spending on product development in AI. The answer to your question is no, we don't see scale from being bigger.
Yeah, I fully agree. I think, I mean, what we said earlier is that, of course, we also gain efficiency if it comes to internal processes, right? I mean, for instance, we shortened the process of developing a product. In the past, we had 6 months, now we've brought it down to 3 months, and we would like to bring it down to 3 hours. From an idea to an MVP. This is freeing up capacity. We never shipped so many new features, probably than we did in the last quarter. That's actually a positive sign. We can reshuffle capacity internally to cope with the additional need of shipping innovation to the market.
That's actually also what others now are starting to do, right? You can see it at Rightmove, at IA, and so on. They reshuffle capacity. They're using AI technology to become more efficient internally and to ship faster. I mean, the filter search It's around or has been around for 25 years. A bit change here and there, but not much. Now you see semantic search popping up everywhere within 1 quarter. I think it shows you a bit what's possible, even with the same amount of capacity and cost budget. It's actually quite positive here. Therefore, I'm with Dirk. On regional side, I think, no extra need.
Yeah.
Okay. Thank you.
Thank you.
Ladies and gentlemen, this was the last question. I would now like to turn the conference back over to... Excuse me. I would now like to turn the conference back over to Filip Lindvall for any closing remarks.
Okay, this concludes today's call. Thank you for joining and your interest in Scout24.
Ladies and gentlemen, the conference is now over. Thank you for joining, and have a pleasant day. Goodbye.