Welcome to the earnings call of Exasol AG regarding the preliminary figures of the full year 2025. The CEO, Jörg Tewes, and CFO, Jan-Dirk Henrich, will guide you through the presentation and the figures shortly, followed by a Q&A session via audio line and chat box. Having said this, I'm handing over to you, Mr. Tewes.
Thank you, Judith. Welcome, everybody, to this presentation from Exasol. We will share our preliminary, but yet unaudited results for fiscal year 2025 with you, and we will also provide an outlook both on our key financials as well as our strategy and product roadmap. I will start the presentation and give you a high-level overview about 2025 and the outlook for 2026, and then my colleague, JD Henrich, is going to take over and share more details on our financials. At the end of the call, we will be happy to take your questions either via chat or just verbally via audio. All right, so let's get started. That's us. So that's me, Jörg, and that's JD below. And let's go through our standard disclaimer.
And then let's get started on the recap for 2025 and our key priorities for 2026. So here is the summary of our financial highlights for 2025. As you can see, we have overachieved our revenue goal, so we got a 5.6 year-over-year growth, so we achieved EUR 49.9 billion. Our EBITDA came out at the upper end of our guidance, even slightly above our guidance. So we achieved a EUR 4.1 million of EBITDA. That was slightly more than doubled compared to our result in 2024. Therefore, our net liquid funds increased to EUR 18.7 billion.
So the company is solidly funded now, and we have been able to generate this cushion for the company over the last two years through our focus on profitability. Our ARR declined by 8% year-over-year. So we have—we've seen significant churn coming in earlier than we originally anticipated. So churn was the major factor on the overall ARR decline. The good news is that since this was earlier than anticipated churn, some of that we actually had only anticipated in 2026. We expect that the churn number is gonna be substantially lower than it was in 2025. So the headwind that we had to achieve ARR, overall ARR growth is gonna be way less in 2026.
So therefore, as we're gonna talk about in our outlook, we are very confident that we are going to achieve mid single-digit ARR growth in 2026. Our net income is EUR 3.1 million in 2025, and the equity ratio has increased to 34.2%. We've been talking about our strategic focus on certain key verticals. We shared that with our investors in previous calls. We've actually seen significant growth in those key verticals. We were able to secure two large upsells with our existing customers in the banking sector, most notably with Finanz Informatik, which we closed in June of 2025, which was a significant increase in that key segment.
We had other key customers where we also had increased volume and upsell in that segment, like LBBW, for example, which is also one of our largest customer here in Germany. So that actually, our strategy to focus on these key verticals worked well. As I just said, we had some of the churn from non-focus focus verticals coming in earlier. There are a few deals that slipped into 2026 in the focus verticals, so we do expect, as I just said, to go back to ARR growth by both minimizing the churn and closing these delayed opportunities. We will continue our path to profitability. We've been putting the company in safe territory and maybe go to-...
Our next slide here, which kind of illustrates the journey in which Exasol is. Starting in 2021, when our CFO, JD, came on board, where we started the turnaround, where we implemented restructuring measures to get the company back out of the deep red into profitability. We sharpened our strategy in the second half of 2024, with the articulation of our strategy to focus on key verticals, specifically in the Central European region. So we completed the successful turnaround by achieving an EBITDA break even end of 2024, which we then continued in 2025. We had the proof of our strategy shift by growing our share in our focus verticals and by doubling our overall EBITDA in 2025.
For 2026, we will continue that route. So we're actually very excited and bullish about 2026 to achieve both ARR growth and continue the journey of driving—staying and driving profitability in the long run. So I think we're on a good path right now for 2026. I'm gonna share more now about some of the activities we've started in 2025, and that will then, of course, carry over into 2026, as well as some of the product innovation that we're working on as a product company. So when we look at what we're doing, this is basically a recap.
And, this kind of gives you an idea on, on where we are from our, from our use cases, and also how customers are using the product. So on the left, you see four key use cases. There's the traditional data warehouse use case. So we do have a lot of customers that use us as their central data warehouse. We continue to do that. We have customers where we have a successful installation, where we are growing, where we are adding more, or customers are adding more data. So that's one of the key use case that we continue to drive. The data warehousing, specifically in these focus verticals, and I'll repeat, for those who haven't heard that, key focus verticals for us are financial services, so banking, insurance companies, public sector, and healthcare.
So these are the core focus verticals here in Central Europe. We're also very successful. I'll speak a little bit about this later in telecommunication in the U.S. market. But here in our core market, it's basically those four verticals. They are predominantly running their data analytics stacks on-premise or in their own clouds or in a hybrid environment. Some of them also look at SaaS, use SaaS, but the predominant installation sits basically in on-prem, in their own data centers. The other use case that we're traditionally strong is analytics acceleration. So these are scenarios where we're helping customers to accelerate an existing data analytics stack to add more, either for performance optimization or for cost savings.
Sovereign AI, and we're using that term over the general AI term. I'll speak more about this, but more and more customers are actually now moving on using AI together with their existing data analytics. And that in environments that are secured, that are sovereign and basically guarded and shielded from the outside world. So that's a very important trend for us, and we're actually capitalizing on this trend. Then last but not least, the Lakehouse Turbo. We also talked in earlier calls. This is similar to the initial analytics acceleration, with the focus now on the modern data stack, where we help customers reduce costs with modern data lakehouse infrastructure such as Databricks and Snowflake. So let me go to the next slide.
In 2025, we actually secured strong partnerships. Among those partnerships, a few that I want to call out here on the slide. It is MariaDB. We shared that in our previous earnings call already, but I just want to recap that here. MariaDB is a company headquartered in Silicon Valley. They were taken private by a private equity, a U.S. private equity firm, about one and a half years ago. And their main business is an open-source transactional database management system, which perfectly complements the analytic engine that we have with Exasol. So, we started the partnership last year.
We also got some initial upfront revenue from them, and basically they are reselling a product where they have their own transactional database, and then data gets migrated to Exasol Analytics Engine, and then customers can run analytics basically on Exasol. That's the key use case. MariaDB has millions, hundreds of millions, even I think a billion downloads of their free open source version, and they have 600, around 600 enterprise paying enterprise customers. So the initial target for us is together with them, to get in front of these enterprise customers paying customers, obviously, and generate incremental revenue.
But it's also an opportunity for us, and I'll share that later, that we get Exasol more visible in this wide world, and use the installation footprint that MariaDB has to get basically more awareness for our product in the market. We also started partnerships with STACKIT and Exoscale. STACKIT is a part or member of the Schwarz Group, the Dieter Schwarz group, who's the founder of Lidl. So STACKIT is very active in positioning themselves as a German alternative to the U.S. hyperscalers, AWS, Azure and GCP. They've actually seen good traction here, specifically in Germany, but in other Central European countries as well.
So our product is now available on their marketplace, and our free version that I'm gonna speak about, the personal edition, is also available on the STACKIT platform. So we've started that partnership with them, and we think it's a good alternative for customers that wanna run data analytics workloads in a regional cloud and in a more sovereign environment that is not dependent from U.S. hyperscalers. Similar play with Exoscale. Exoscale is also a European cloud provider. It's a subsidiary of the A1 Digital, Telekom Austria, so they are very strong in Austria and Switzerland, actually, so it's a good alternative.
It's a similar story, like STACKIT, so that we'll also provide our customers the ability to run workloads in the Exoscale environment. Beyond partners, so partners play an important role for us. We have brought on some people in our team internally on the partner management side, and we also have strengthened partnerships with other players, specifically here in Germany, most notably Adesso, for example, who we have a strong business partnership now as well. So that's on partners. Then a big milestone for us, we have launched a free version of Exasol. It's called Exasol Personal that we are going to use to generate product-led growth momentum.
So we will, of course, continue our sales motion with our go-to-market organization to drive awareness of the product through standard marketing channels through event participation. But in the future, we are going to invest more resources and, of course, also some financial resources to drive overall awareness of the product in the global market. So it's very important that we will get the visibility of the product to the outside world, and so we're not just talking to, let's say, decision makers in enterprises, or we are going to the developers and data analysts in well, large or even smaller companies. And the Exasol Personal edition is basically the same product as our standard Exasol product, with just one key, a restriction, it only allows one connection to the database.
So, a user can basically simulate, and benchmark, and test the full capabilities on Exasol with that, but they can basically not use it in a productive environment, because then they would need more than just one connection, and that's where we then would potentially have an upsell. So, we make this as easy as possible for users to get access to that personal edition. So we started launching it right now on several platforms. AWS was the first one, but as I just explained, we're also gonna have it available on STACKIT, on Exoscale, and on Microsoft Azure as well this quarter. And we continue to drive attention that we're participating in data meetups, so it's really a-...
I want to call it the grassroots initiative to get more visibility and mind share in the database market. The database market itself is a hype market, so I think it's important for us to get on this hype train, so people seeing us as a modern solution. We're using the Exasol Personal as a vehicle to basically transport that message to the outside world. So what are we doing on the product side? Where do we innovate, and how do we drive the product forward? We have two key use cases here that are already highlighted on the overview page. The topic, sovereign AI, that I want to dive a little bit deeper on, and the Lakehouse Turbo. We shared that with you in previous calls.
Again, this is a solution that lets customers who run on cloud-native data warehouses optimize cost and performance. So we're in a testing phase there with key reference customers. We continue to drive that to determine if and how we have a good go-to-market fit. But I think what's actually super exciting for us is the adoption of how our existing customers are using AI in conjunction with Exasol. We now have roughly 15-20 customers that are using our, let's say, AI enablement that we have in our product to implement AI use cases on our product, and who have been able to roll these out successfully. So we have a variety of use case reference stories.
Some of them are already on our website, and we're sharing that also, coming back to that product-led growth initiative, to gain more awareness and to help customers understand that AI and Exasol goes together really well. Sovereign plays an important role. We think that the need for data sovereignty becomes even more important when AI comes into play. So it's not just the data that can potentially leak out of an enterprise, it's also know-how, it's the way how prompts are being generated. So that's why, for us, sovereign AI is really the key topic. So let me dive a little bit deeper into this. We position ourselves as the fastest database for AI developers in that space. And we let developers use workflows that seamlessly integrate with AI develop models for rapid development.
As with standard Exasol, we have a high concurrency support, so the demands for agentic and AI workflows for real-world use cases, when there's a lot of data and when there's a lot of parallel queries entering the database, that's what Exasol is made for, specifically around agentic AI, where there's a possibility that a lot of simultaneous queries hit the data pool. This is where we shine, and this is where our solution really helps our customer. As I said, the data sovereignty aspect, secure governable workloads in your own environment, so customers can run this on premise or in their own cloud account. This is very important for a lot of customers, and we'll deliver the best and optimal price performance for our customers. Here's one slide that gets even deeper.
I'm not gonna go through all of this, but I want, I thought it's helpful for our investors to understand also the innovation that we're driving, and maybe you recognize some of the terms here, the technical terms. So we launched a MCP server last year, Exasol MCP Server. This allows customers that use Exasol to basically connect to any AI system, to any AI. So, for example, ChatGPT Enterprise version supports various MCP servers. So you can just connect Exasol as a data source into ChatGPT, and then you can prompt ChatGPT, and data gets automatically extracted out of Exasol. So we play in a larger environment, so the MCP server itself is a key interface, and we launched that in early Q4 last year.
This also helps us to do deeper and better integration into standard LLMs. And, as we run in secured environment, so customers can run it in a secure stack for regulated environments. So we continue to drive integration and, and full-stack AI integration. So it's no longer just one product, there's one thing. I mean, there are a lot of components being stitched together, and what our product and engineering team is doing, is working on, is to continuously support and add interfaces to the necessary environments.... Our AI and query engine is scalable. We continue to improve performance, scalability of our core database system. We now have a multi-language support for all relevant languages in the AI world, Python, Java, Lua, R.
We help customers to do SQL-based model training and execution at scale. The last bullet point here, value-driven AI applications. So, we've been working on, I think we also shared that in the past, we now have actually customers using that, converting unstructured data into structured data. Why is that important for us? It basically increases data volume. So we are able, with that, to get customers to put more data into the Exasol database. And that's a good thing because typically, we price on data volume. So we help customers in their infrastructure.
So, we've released GPU acceleration for key AI workloads, so customers get much better performance, specifically when they have invested in GPUs, and we deploy in flexible environments. We're in the process of developing, that's on the roadmap for 2026, even more modern infrastructure support. So for example, Kubernetes, Docker support, is on our roadmap for 2026. This is a feature that has been requested by a lot of customers and prospects that we're gonna bring to the product, this year as well. I brought one example. You might have seen T-Mobile as one of our key customers before. This slide goes a little bit deeper in what is T-Mobile actually doing with Exasol, and how they're using AI, to be successful.
So T-Mobile by now has become our second-largest U.S. customer, and the reason, as I just explained, T-Mobile is a good example for that. They're just adding more and more data into their Exasol database. And why are they adding more data? Because they're expanding use cases on what they're doing with us. So this slide basically describes that T-Mobile has about 120,000 cell towers all across the U.S. with multiple cells and sectors per tower. And what they're basically doing, or the business problem they're trying to solve is: how do they equip these towers with additional cells and sectors, and where do they put new towers in this country that, as we all know, is very large and very big?
So they're leveraging the historical traffic data over the last 52 weeks, and then have defined. We worked actually together with them, our team in the U.S., to build certain key AI machine learning use cases that creates an optimized planning based for each tower on how it will be best equipped on a going-forward basis. And that is being done with the data pool that sits in the Exasol database, and AI Lab, as it's been described here, is basically the key interface that we're providing to our customers. So the outcome is an optimized, improved network coverage. It's a really good use case that shows the value that we're bringing in.
As I said earlier, we now have about 15-20 of our existing customers actively using and deploying sovereign AI solutions from Exasol. Just to recap again, and then I'll hand over to JD the journey that we're on. We are very excited about 2026. It's an exciting time to be in tech in general. The shift to AI is gonna be a game changer. We all know that. We've been seeing that for now, almost 3 years. We believe in Exasol, we have the tools in our hand to capitalize on that, so that's one trend that's helping us on a going-forward basis.
And the other element, we do believe as a German company in this complicated world that we're in, in the geopolitical environment, that continue to hone in on our key strategy in key verticals, that customers want this more sovereignty from U.S. vendors, that's another big plus. Exasol is basically the only commercially available database technology here in Germany, so we're pushing on that, and we're working with partners. We're also getting closer to politics in Berlin and in Brussels, so I think it is an additional incremental opportunity for us to capitalize. So with that, I'll hand over to JD, who's gonna go deeper into the financials, and then at the end of the call, I'm more than happy to take any questions. Thank you.
...Thank you, Jörg. Let me shine a little bit of more light on what, Jörg, elaborated to you from a quantitative perspective. I think as he already hinted to in the beginning, in the executive summary, from a pure results perspective, it was a bit of a mixed bag for us, 2025. We had some things that went very well. I think the development of the focus industries and the validation we got in terms of market feedback, and opportunities, the financial strength of the company in terms of bottom line, on the one hand, the significant headwinds that we faced in the non-focus industries, which were also faster than expected, on the other hand, and let me give you some numbers around both of that.
Maybe starting with one of the pluses, I think in terms of pure financial strength, we made another significant step forward in 2025. EBITDA rose to EUR 4 million, which is more than double what we had in 2024. Likewise, we had again almost a cash conversion of one. Our liquidity increased by EUR 3.7 million to almost EUR 19 million. One thing I might add is we achieved both of that despite not insignificant headwinds from the weakening of the U.S. dollar, which obviously made the bottom line impact of our U.S. business significantly weaker. So I think we also have to look at the development of the financial results from that perspective.
Now, this financial strength, we don't look at necessarily as an end in and by itself, but as a comfortable position to handle the transition from that we've anticipated with the change in our strategy and the transition in our customer portfolio that we are observing, and that we have observed in 2025. So if we look at the overall ARR development, which is the key indicator for our business, as Jörg said, we faced a total net decline of 8%, which was largely driven by the significant amount of churn that we experienced in the non-focus industries. As you can see here, this churn amounted to a total of EUR 10.2 million, or almost 24%, of the beginning of year ARR.
Now, almost, well, 50% of that churn was basically driven by three large accounts in EMEA, in the non-focus industries, that also reduced their business faster than expected. So when we implemented the focus strategy, we anticipated a certain amount of churn over this and next year, or sorry, over 2025 and 2026. And a lot of that happened, or a bigger portion of that happened in 2025 than we originally anticipated. So this created a headwind on the churn side that we weren't able to make up with gross new business. We achieved gross new business in total of EUR 6.8 million. And looking at upsell and new logo separately, overall, on the upselling side, things largely evolved as we expected them at the beginning of the year.
On the new customer side, we are not yet where we want to be. We aim for higher values there. The EUR 1.4 million is good, and we achieved that also with the revenues that also our new partner, MariaDB, committed to, but it's not where we originally wanted to be. In the sum of those factors, we then were not able to achieve our original aim to achieve growth but faced the net decline. Now, that being said, with the big amount of churn impact being kind of preponed effects that we anticipated in 2026, this headwind will significantly decrease in 2026.
Overall, we are very confident that we will be able to decrease the churn by more than half in this year, which puts the churn rate back to 10% or maybe lower, which gives us a much better basis to get back to net growth in conjunction with the impact that our the initiatives and the partnerships and the slipped deals that Jörg talked about will unfold. One additional more technical remark, as you know, that we are reporting ARR during the year at constant FX, and also show you the results like for like, and also our guidance is against a like for like target. Then at the beginning of each year, we adjust the ARR value to the new FX basis.
We will do the same, so, the EUR 39.1 million end of period ARR translates into 38.4 on latest FX rates, and this is mostly coming through... or this is exclusively coming through a devaluation of our U.S. business, due to the weak U.S. dollar, that almost lost 20% in the course of 2025. So in the webcasts and results calls moving forward, you will see the 38.4 be the starting point for the 2026 reporting. Now, this overall situation also then led to net revenue retention be below 100%, for the first time, with mixed pictures across regions and segments, but I'll talk about that in a moment.
So, looking at the quarterly performance and how that unfolded, as mentioned to the churn or the headwinds that I talked about hit us primarily in the first half of the year, whereas ARR largely stabilized in Q3 and Q4. Now, overall, Q4 was not satisfactory for us. I think Jörg mentioned some slip deals, particularly on the focus industry side, which we're now expecting in Q1, Q2. So the traditionally strong Q4 that we typically had didn't unfold as much, neither on the focus industry or non-focus industry side. But as said, this we expect to recuperate in Q1 and Q2. That being said, overall, our focus industries continued to strengthen over the year.
You can see in the bubbles on the chart that focus industry ARR rose from EUR 24.2 to EUR 26.7. And that translates into then a portfolio shift, that we've experienced, that we are now roughly at 70% share of the focus industries in total ARR. That kind of represents the continued shift of the portfolio, towards the focus industry. So overall, 2025 was characterized by, kind of slightly slower dynamics than we aim for on, on the focus vertical side, although in principle, our hypothesis that this is the strategic field to go for is definitely confirmed, and faster dynamics on the non-focus vertical churn side. And this also came about if we look at the regional breakdown.
If you look at the industry split and the regional split in the full matrix, you can see the clear growth driver, and where do we take the statement from that we see the overall focus strategy validated. We can clearly see that in the core EMEA market, in the focus industries, where we grew by 14%, which also was the clear growth driver. The non-focus industries in EMEA were overall the key headwind sector, so to say, mostly driven by the three big customers that I referred to earlier. The Americas in total largely stagnated, with no clear difference in dynamics between focus and non-focus industries. I think what's clear here is also that the U.S. market is driven by slightly other trends than what's currently driving the market here.
I think AI plays a big role on both sides of the Atlantic, but obviously the data sovereignty discussion is dominantly a European discussion, whereas the U.S. is much more focused on AI, but then also on the compute and cloud cost topic. And because of these factors, the product development and innovations that we're working on, that Jörg elaborated on in his part, are important for us also to then reinvigorate our growth potential in the Americas moving forward, while we continue to focus on driving the growth in the focus verticals in EMEA as part of our focus strategy.
So all this together, if we look at some P&L detail and how the results unfolded, I think the very bottom of the page I already talked about, we were able to increase EBITDA to EUR 4 million. Also in conjunction, our net income rose significantly, even more than EBITDA, where we had an improvement of almost EUR 3 million. That came about through two effects. First of all, we were able to increase further our financial result that contributes below EBITDA, but then also our depreciations decreased further. I think we've talked about many times that we stopped capitalizing R&D expenses a while ago, and the amount of intangible assets in the balance sheet that's still related to old projects that were capitalized in the past is continually shrinking.
And as a consequence, the net income contribution which we generate out of our operational business is increasing. And this then also led to the increase in equity ratio beyond 30% that Jörg already talked about originally. Now, if you look at the revenue dynamics between 2024 and 2025, you kind of see what happens with the structure of the business. You see that overall our revenue increased, but we faced a decrease in recurring revenue, which is not surprising, given the 8% decline in ARR. At the same time, we were able to compensate that through one-off hardware and service business, particularly with our customers in the focus industries, where we see more and more demand also for bundled offerings. So not only the software, but basically, hardware, which has the software pre-installed.
One remark I would like to make on the gross margin development. You see here that on a reported basis, gross margin declined by EUR 1 million. At the same time, I'd like to point your attention to the fact that the IT infrastructure costs have decreased significantly one year over the other. This is related to a change in how we report our cloud cost.... In the past, in 2024, all our AWS cloud costs were reported under IT infrastructure.
Now, in 2025, we saw, we had a significant improvement in the level of detail we received from AWS in how the cloud costs break down, and as a consequence, we started allocating the appropriate amounts to, the SaaS product, which is roughly EUR 750,000, which are now reported under COGS for the SaaS product. So the delta that you see in gross margin is to 80% related to that effect. If you account for that, gross margin remained largely stable year-over-year on a like-for-like basis. So you can see that the decline in gross margin that was driven by the declining recurring revenue was compensated by the additional one-off business that we were able to generate, kind of compensating and buffering, these transitional effects.
Then on the pure cost side, obviously, once you account for the effect I just talked about, IT infrastructure cost even still decreased a little bit on a like-for-like basis, but the main contributions to bottom line came from savings in personnel expenses, based on a streamlined setup for the go-to-market and organization, and some other changes we made, over in the Q1... Q4 2024, Q1 2025. So these effects in total led us to the result where we are. And then maybe, taking a look forward to this year, what does that mean for 2026?
Obviously, in the end, we're expecting ARR growth back in the mid-single-digit percentage range, based on significantly reduced churn rates and the initiatives that we're having on the new logo side, and upsell side, along the lines that Jörg described to you, in combination. On the revenue growth side, in terms of pure recognized revenue in the P&L, we expect a single-digit decline. This is mostly then basically the delayed effect of two things. First of all, obviously, ARR is a leading indicator, so the ARR decrease still affects us in recognized revenues in 2026. But then also, we expect the one-off revenues not to be on the same total amount that we had in 2025, which was exceptionally high.
Lower one-off revenues, still some after effects from the decline in ARR before then the impact of the rebound ARR growth kicking in. As a consequence, because of that, we aim to keep profitability in the range as we aimed for in 2025. We again aim for an EBITDA of EUR 3 million-EUR 4 million, and also to support continued investment in the product, which we want to do, despite the revenue dip that we're going through. This is how we look at 2026. What I might add is that we, in our ARR perspective, did not yet factor in significant amounts for the new business we work on, like the Lakehouse Turbo, since this is in a proof of concept phase.
So this is kind of the growth we're aiming for with our focus industry and AI products that remain the core of our offering. And with that, I would conclude my part. In sum, what would we like you to take away with that? Exasol continues to perform strongly in terms of profit and cash flow. We've closed important partnerships in 2025, which will help us get back to growth in conjunction with significantly reduced churn. We focus heavily on supporting the big trend of AI and analytics that we see out there in the market, and we believe we are excellently positioned to benefit from sovereign AI tendencies out there.
And overall, although we haven't made the progress that we wanted to have in 2025, we do see clear validation of the focus industry strategy that we took, and expect to make inroads on that in 2026. So with that, I conclude. Let me also say that, I'm gonna be at the Hamburger Investorentage tomorrow, so maybe, hopefully, I can see as many of you as possible, and see you. If not, obviously, please pose your questions now in the Q&A section, or get in touch with us after the call, and we'll find a way to address your concerns and questions. Thank you very much for listening so far.
Thank you, Mr. Tewes and Mr. Henrich. Ladies and gentlemen, we are now moving on to the Q&A session, and you can post your questions in our chat box, as already happened, or use the Raise Your Hand button for a dynamic conversation via audio line. As we have no hands up, I will jump to the questions in our chat box. The first one: Could you shed some light on why contracts were postponed from Q4 to 2026, and what makes you sure that these contracts are going to be closed?
Yeah, maybe I can start. Can you hear me? Yeah. Okay. Maybe I can start, and JD, feel free to chime in. I think we're looking at two types of contracts. We actually have several-
... new logos in the financial services sector that we've been working on actually for quite a while. We are confident that we will be able to bring in these logos either in Q1 or Q2. So that's specifically some banks that have started working on POCs with us, and so that's one area. The other one is a larger customer that basically has the need to add more data to their existing installation. And we also expect that to materialize in Q1 or the latest in Q2. So I would say the confidence on these delayed deals is relatively high. Yeah. So that's maybe the first question. Second, from Mr. Weiland.
Do you expect new ARR in 2026 more at the beginning or the end of the year?
I think it's balanced. We have done significant improvements to our existing commercial organization. We're seeing initial traction of these improvements, but I think we will see increased ARR in Q1, as we also said ARR or net ARR is always gross new ARR minus churn, right? So that's basically how we look at it. And we think, or we know, this is actually relatively easy to predict, that our churn numbers will be substantially lower. So that helps us on the net new ARR side, because obviously we can—we don't need to do as much new ARR to case it—basically get back into the ARR growth segment.
But, I think we will also expect more new ARR coming in in Q3 and Q4.
Thank you.
Maybe, maybe an additional remark, because there was also, I think a question on seasonality from Lucas, from Lucas. And he also asked about earnings. I think Jörg already gave you the ARR perspective. I think from an earnings perspective, overall, it's gonna be a little bit. I expect a little bit of a weaker start into the year, simply because these delayed impacts on recurring revenue, which come from the revenue decline in 2025, hit us mostly or still sitting on our books mostly at the beginning of the year now, before we see net ARR growth again, which then also supports top line. Or maybe directly also answering another question that was raised when, for example, effect like the one of government grants kick in.
Yes, we expect, again, government grants in 2026 in the mid-six-digit range. And those typically are factored in, in the middle of the year, Q2, Q3. So, obviously, we still expect profitability every quarter, but it's the seasonality on profitability is gonna be a bit more second year, second half of the year, than first half of the year. Maybe to build on, make that additional comment on seasonality of earnings. I think on, on ARR side, Jörg already gave you the perspective.
Thank you. And I will move further with the questions from Mr. Spang. Can you share some insights of the partnership with STACKIT? Are you just available on the marketplace, or are there also some active sales activities from their side? And can you quantify the potential project sizes in the focus verticals which slipped into 2026?
Yeah, maybe I'll take the STACKIT piece, and JD, you can take the other part. So STACKIT, I think we tightened our partnership. They're not actively reselling us. However, as an example, we have their CEO, Bernd Wagner, as a keynote speaker attending our customer event, the Exasol Xperience, which takes place on March tenth in Berlin. So I think it's a commitment from STACKIT to work closer together with us. So getting visibility in the market as a local German solution, that's very important. What we do have actually, where we have now an active sales partnership, Adesso, which I think most people here in the room know, it's one of the largest German system integrators with 10,000 employees in Germany and another 2,000 outside Germany.
Adesso has assigned a dedicated salesperson that's selling Exasol to Adesso customers, so that we have something similar, very close tie with Adesso Turkey. So we have that momentum where a large partner, specifically in our-- Adesso is very much aligned with our core focus verticals. They are, they are actively asking or reselling us. So that's maybe the answer to that second question. And JD, you want to take the third one?
Yeah, on the, on the average project size, I think there were, as Jörg mentioned, there were a couple of new logo projects we're working on in the focus industries, which were on the low six-digit side, in terms of size. And then there was also a couple of upsell opportunities, which are in the low to mid six-digit area. So that's kind of in line with what we talked about earlier, that we saw slightly bigger project sizes in the focus industries than in the non-focus industries. But that's kind of the order of magnitude. But there was no mega deal like the LBBW or Finanz Informatik deals that we talked about last year and in 2024, that was among them. Yeah. So, those is...
So it's a kind of a bigger portfolio of mid-low to mid six-digit deals.
Thank you for the questions. Are there extra, extra hardware installation projects to expect in 2026, like it was in 2025?
Yeah, I'll take this one. At this stage, we don't expect the same volume of hardware and service projects as in 2025. This is also kind of factored into the revenue guidance that I, that I talked about. Why? Quite simply because we had one very, very big deal in there with one of our financial industry customers, which we don't see being repeated in the same way in 2026. Yeah. So we have an assumption in there, but it's lower, significantly lower than we had last year. Also, because quite frankly, this is much harder to predict than the software business.
Very often customers come with us and rather kind of a last moment idea of that they also want to procure hardware as part of that. So it's much harder to anticipate, and therefore, we also don't factor it as much, or only very conservatively in our financial plan for the upcoming financial year. I think-
There were two more questions, JD. Maybe we should try to get answers for those.
Yeah. So, did you lose customers in the focus industries? Yes. In EMEA, however, mostly very small ones, so, tail-end customers. There was one bigger customer in the U.S., in the focus industries. As I said earlier, the, the data sovereignty topic is not as much of a, of a topic there. So there was one bigger hedge fund client, that we lost, at the beginning of the year, which contributed to the churn rate, that you saw in the focus industries, which, however, overall, was still, with 7%, in, on, in a very good range. Yeah. And then I take the MariaDB question straight away as well. So MariaDB revenue, 2025.
So as part of the agreement that was signed, MariaDB also committed to a pre-commit revenue in the seven-digit, low seven-digit $ range, which translated into a high six-digit euro range, which kind of is in the numbers, or at least in the ARR numbers, and is gradually entering the revenue numbers. For 2026, I think, we are just starting the market motions with them. Obviously, we hope and aim to together develop this, then also to an upsell versus that value, to get that up. Their own plans for the next three years is clearly to drive this joint business in a double-digit million range on their ARR end, of which we perceive 20% royalties.
Yeah, so that's kind of how the deal is set up.
Thank you very much, ladies and gentlemen, for your questions. With no further questions, we come to the end of today's earnings call. Thank you very much for your interest in the Exasol AG, and a big thank you also to you, Mr. Tewes and Mr. Henrich, for your presentation and your time. Should you have any further questions, you can always reach out to investor relations, and also, as mentioned, meet Mr. Henrich at the Hamburger Investorentage, starting tomorrow. I wish you all a successful day, and handing over again to you, Mr. Henrich, for some final remarks.
Yeah, well, all that remains for me to do is thank you for your numerous participation in the call. Thank you all for joining. Thanks for listening. Thanks for your continued interest in our company and our share, and hopefully see you all soon, either in person or in some other form of catch-up.
[Foreign language] . Thank you all, and good afternoon.