Great. Good morning, everyone. Thank you for joining us. My name is Mohamed Moawalla. I cover the European software, IT services, and payment space here at GS. We are delighted to have SAP join us again at the conference. Representing the company is Dominik Asam, Chief Financial Officer. Dominik, welcome back. I know.
Thank you, Ben.
You made your debut last year.
I appreciate it.
Thank you for joining us. Maybe just to sort of kick off, the macro backdrop has obviously become a bit more softer over the past year. I mean, SAP's performance actually has been pretty good relative to that and versus the industry. Obviously, tariffs have come more into the picture. What's sort of the, when you speak to customers, give us your sense on the spending environment, digitization projects in the context of this macro.
Sure. The good news is that the pressure that is currently put into the system by these geopolitical tensions is, on average, helping because people are grappling for productivity and are saying, OK, how can I transform the company to really make savings, push the top line, and everything we can do? AI is, of course, a catalyst to really have a higher chance of bringing the benefits of the transformation to bear. On the other hand, in certain industries, there are just some boundary conditions which you need some clarity on. I always manage to mention the more complex manufacturing customers which have global supply chains. We had a little bit of ups and downs, I'd say, over the last weeks. The trade resolution with Europe was certainly a positive. We had also some pretty nasty discussions on India and Brazil and other regions.
I'd say that the overall picture in these specific buckets of the market has not changed. The public sector in the U.S. is still going through a lot of turmoil.
Yeah.
It is basically not unchanged. The challenge is, of course, that if you are pushing out some of these closings just because of the deployment time, the further you move towards the end, the less revenues reach our top line in the remaining to-do of the year. This is what we have to cope with. The good news is the general fundamental journeys are intact. On all of these situations, we have no debate as to that discussion being ended. It is all about when do we have enough clarity to really pull the trigger and continue because nobody can afford not to deploy AI, basically, these days to get the maximum productivity out of their workforce.
Got it. I think on the second quarter results, you had already flagged some of this booking softness in the industry, as you've mentioned. Can you give us an update on how you see current cloud backlog evolving? I know at the start of the year, you had flagged there would be some sort of deceleration, and we saw the first signs of that in Q2.
I mean, still, you talk about softness. It's a strong word for what we discuss here.
I mean, we have very high CCB growth rates. We did see a 1% deceleration in Q2. We have flagged that because of the anniversary of the WalkMe acquisition, there's around about a 1.5% technical deceleration purely from deconsolidating, or not deconsolidating, from having the comms in the prior quarter. We are also currently in the process of closing a smallish transaction, which is called SmartRecruiters. We have to see when exactly we will close. By the way, that one is probably more half a % on CCB given the small scale of that acquisition. These ones need to be properly embarked to kind of come to the underlying growth rates. Still, I always urge people to not be too much carried away by a 1% deceleration in that CCB growth number.
When we take the different buckets of revenues of SAP, cloud ERP suite, extension suite, the infrastructure service, the software license, software maintenance, and the services, and no matter whether we take 2024 or 2025 first half as a base, and we use the exact same growth rate and extrapolate them to the right for three years, in every single year, we would theoretically see a 3% acceleration, meaning that we can actually afford a certain deceleration and still accelerate the group revenue growth. We are really at the sweet spot of that mix effect. This is why I'd say it's still within the kind of boundaries that will allow us an acceleration of revenue growth in 2026 and 2027.
Got it. Got it. Maybe just moving sort of to SAP S/4HANA, the product cycle, obviously, momentum has been pretty strong. Can you talk us through where you are in the migration cycle as of now? I know you've alluded to while 40% to 50% of the customer base is either migrated or in process, but of the workload, we are still very early in that inning. How should we think of that over the next sort of two to the end of the decade in terms of how you think of that tool?
Sure. I would segment it in two ways. The first one is the move of some legacy software of SAP onto SAP S/4HANA, which is, frankly, the more complicated, challenging part of the transformation. The large enterprises spend quite some time on that. We talk sometimes about years of transformation as opposed to months. The easier part is then to take that SAP S/4HANA installation and to move it to the cloud on RISE with SAP or on a GROW with SAP greenfield type of implementation. In terms of the move to the cloud, if I look at the maintenance base today, I would say that very roughly a third of the maintenance-paying customers, if I look at the maintenance revenues ERP, which is about $10 billion, a third of that is paying both cloud revenues already and maintenance.
That means that even that first third is not kind of fully transformed yet because they pay both. That means two thirds still to go. If I look at the transition from ECC and older to SAP S/4HANA, the vast majority of our customers have some form of commitment to go as SAP S/4HANA already. It doesn't necessarily mean that they have completed the transformation. They might be in the middle. We think that's a very conducive, very favorable situation to start from because we still have a lot of runway on transforming to the cloud. We have already the commitment of the lion's share of our customers to stay on SAP and be on SAP S/4HANA. Of course, when they are on SAP S/4HANA, they are not facing the same time pressure because of the out-of-maintenance issue end of 2030.
There are also benefits from going into the cloud, like the AI they can use, like some other features they can use, and the lower deployment costs. Most of the customers don't move to the cloud because they like SAP and want to move to the cloud. Most of them do that because they have a better total cost of ownership doing that while us making more money with this. That's the beauty of the journey. It's actually progressing as planned and is giving us a high degree of visibility for that part of the growth.
Yeah. Obviously, once you are on that migration journey, you've got a broad suite of line of business solutions, now data products. How should we think about the kind of cross-sell, upsell opportunity in the install base then?
I would slice the answer in two parts. The first one is really the kind of traditional conversions of the install base. We are saying what we always said, that on the first signature of a RISE journey to the cloud, people tend to convert at 2% to 3x. They don't do that in terms of revenue. If you compare the maintenance they pay prior to that decision and then the subscription they pay in the cloud, it's kind of 2% to 3x. As you mentioned, there is cross and upsell happening, which actually kind of doubles that potential. If you think about four or five years to double that opportunity, it implies a net retention rate of like 120% or so because if you compound that for four years or five years, that's how you double stuff.
Yeah.
That is not completely unreasonable. It is really triangulating quite clearly here. We do believe that the platform idea, that we use the SAP Business Technology Platform to stitch all these applications together to alleviate the pain of our customers to integrate all these applications, to ensure that there is one source of truth behind the data and they do not duplicate efforts, is already a very high value. That is fully intact and no change to the story at all. I can only encourage you to revisit what we said at Sapphire in May this year, no change on that. Where we certainly have a little bit more confidence over time is on the other things that are more big opportunities towards the later years of this decade. I want to call out the three big ones. Of course, AI.
The monetization on AI these days is still relatively small, but it is growing fast. We have the SAP Business Data Cloud, great customer reception. We think this idea of allowing people to federate SAP data with all kinds of data in a seamless way, to not having to rebuild, maintain all these pipes between different sources of data, is a huge advantage. Last but not least, we have now with the GROW with SAP public cloud multi-tenant motion, we can go way further down into medium-sized enterprises. That revenue contribution is still quite moderate. It is completely realistic for us to see in each of these three buckets a billion-plus opportunity in, I do not want to kind of say, is it year four or five of a forecast period, but in that type of time frame.
If you compound that and then compare it to the revenue base that SAP might have at that point in time, you see that is already good for a kind of mid-single-digit plus acceleration. If anything we have on top of the migration cycle of the kind of bread and butter business with the large enterprises we run, we have this opportunity. Should at some point in time occur some deceleration in that, the runway on that one is still long. This is why our conviction about sustainable high growth rates for SAP is very high. Combined with our ability to leverage our own AI and also AI from other products to decouple the cost base more and more from the top line, we see the opportunity to expand the margin to drive the company forward.
Great. I think in Q2, you sort of alluded that over half of the order entry was a function from kind of AI use cases and deals. Maybe paint a picture for us of what are the kind of AI capabilities today that you're offering to customers and that's kind of driving, pulling the cloud migration through with it. How does that kind of contribute to the pipeline going forward?
If you think our business, what we actually sell is a transformation. To tell a customer, you have a certain way to run your business today, and you want to make that more productive. You want to automate. You want to get better insights. For that, you need an upfront investment, and then you reap the benefits after the transformation. What we need to achieve by AI is to lower the so-called non-recurring costs from the transformation and increase the benefit from what we get out of it.
Yeah.
Oftentimes, people forget the first part of it. If you want to deploy AI in the first part, first of all, you have to have all the tooling required, like process mining, like the enterprise architecture management. Also, now WalkMe, the in-app guidance of the user to adopt the tools. There are also partner apps we use for testing, which will be more and more automated. There is now SAP Joule for Consultant and SAP Joule for Developer. With SAP Joule for Consultant, we can basically put the best knowledge of the entire developer community of SAP at the fingertip of every single consultant to accelerate decision-making. We see really that the consultants save a lot of time when they don't have to do manual research or ask colleagues and call around, but just get it at their fingertips.
In these transitions, there is a lot of refactoring happening of legacy ABAP code into ABAP Cloud, which is compliant with the Clean Core approach of SAP. That is a massive undertaking for the install base customers, and that can be automated. We use GitHub Copilot for some more traditional coding tasks, but for our own ABAP refactoring, we can use SAP Joule for Developer. We tackle that part, reducing the non-recurring cost by infusing AI there.
Of course, it's all the automation on the outcoming side where we look over the shoulder of the customer and figure out what they are doing and learn from that, and then also train the models to basically do the same. We can go into topics like conflict resolution when the customer receives a wrong invoice, and then we can complain about it. The benefit there is twofold. It's not only the pure productivity. It's actually, I'd say, a little supercharged because if you bring more productivity, you can make certain business opportunities reachable, which before were not reachable. Think about my example of a wrong invoice. We invoice a customer that should have received something at time X, but they received it later, and they say, I don't want to pay for it because I've received it later. You have to correct the invoice.
Today, in the shared service center, that's it. You stop there. Now, with AI, you can then kind of dig into that transaction and say, who is actually responsible for that delay? Basically, then recover your funds by leveraging that intelligence you have in your AI-enabled conflict resolution engine.
This is a nice example because it shows that it's not only by bringing productivity, but by also bringing fruit that was too high to harvest. It's like a ladder. You know, it's like you put a ladder on the tree and say, that kind of high-hanging fruit I can reap now too. This is where the excitement lies on the output side. We have that wonderful knowledge of the problems the customer needs to solve in 25 different industries throughout the globe. We consistently and systematically exploit that and put that in our SAP Joule copilot. We also work very hard on the integration with other platforms. Be it a copilot of Microsoft so we can seamlessly book travel and do the expenses without the kind of user having to jump between copilots. We do believe there's not a lot of room to do that with hundreds of companies.
I mean, there will be some big guys doing that integration. We mentioned SAP Business Data Cloud already. If we want to do an intelligent app embarking third-party data, we now have a platform. It's not only that our customers can federate the SAP data with non-SAP data. When we develop apps, when we look over the shoulder of a customer, we can also use third-party data. I always bring the stupid example of a beverage company that wants to forecast usage of their products. Of course, you want to have the weather forecast in the system. That was not accessible before for SAP in an easy way. Now, with SAP Business Data Cloud, we can kind of link into our own inside apps, everything the customer can use. That gives us another opportunity to monetize what we call intelligent apps on top of that platform.
There is this sort of debate, growing debate around the sort of death of software, application software. Would love to get your kind of perspective on this at a high level. As you think about the business model impact, you know, the sort of currently the kind of perceived subscription-based model clearly is going to evolve into sort of probably a more of a hybrid, I would say, with some consumption-based. Would love to get your take.
Sure. First, I want to clarify that if I look at SAP's subscription and also software revenues, it's a misconception that that would be predominantly based on seats.
It's actually on some KPIs that are negotiated with customers. I just had an insurance company a couple of weeks ago. It was cross-insurance premium underwritten. There can be revenues. It can be seats too. Today, it's way less than half, which is seat-based. Now, on the AI monetization, we still have a fluid situation, I'd say. We are in the early innings of the market. The current model that's contemplated is a per-user per seat model with some overages if there's overconsumption. I'm sure vendors like us will not be stupid that if we kind of eliminate tons of jobs, that we stick to that. We will need to find something that's reflecting the value of what we bring to the table.
That's the key point. I'm really baffled by this assumption. Look, our customers now get these wonderful coding tools, and they would all code themselves. Of course, they will do that. Why wouldn't we do that ourselves too? The key question is not, will the customers do it? The key question is, will the customers or SAP get more out of that coding revolution? I would argue it's easier to transform these coding engines if you're 30,000-plus developers. If you need to recruit new talent in the market for a company like SAP, then it is for most of our customers, maybe with the exception of some, I don't know, huge bank or some companies which have tens of billions of IT budget, but there's very few of these out there. I'm really not convinced by that argument because it applies different paths, different velocity on transformation.
I have no inferiority complex that we can drive transformation on our own development faster than the customers because that's our day-to-day business. It's at the core of what we do. It's our raison d'être, the reason why we're there. The other thing is, I think it's actually a huge opportunity because before you were kind of confined to the IT budget of companies, now you suddenly tap that hugely expensive knowledge worker opportunity. I did some math. If you just take an assumption and you say 300 million-plus users and a typical mix of these knowledge workers in a typical company, and the merit itself is probably a mid-triple-digit billion money pool. You know our revenues. If I only can reap a tiny bit of that for SAP, why should I have any issue about AI? I don't see that risk.
I see that, of course, if we are stupid and we are not able to bring the technology to bear in our own shop, then we will be left behind.
Yeah.
I do confirm that AI will, for any vendor, SAP or others, be either great or horrible. It will be great for those who know how to do it well. This is not only the skill of the vendor. It's also where they sit in terms of data. This is where I think SAP has a pretty favorable position that we really sit at the nexus of the processes and the data. That's the treasure we can harvest. This is why I seriously have no inferiority complex on that question.
Yeah. No, we had a VC panel discussing this yesterday. It is exactly this, that the size of the pie is growing so significantly.
Yeah.
The other interesting point was companies, either vertical software companies or those with data, which brings me on to the point around what you just said, that your kind of presence in terms of the enterprise data ownership is sizable. At the Sapphire conference, feedback on SAP Business Data Cloud was very positive from the partners. I think I would love to kind of get your perspective on, you talked about this being potentially one of those new incremental billion revenue buckets. What is your partner with Databricks already? How do you see the kind of momentum there and the monetization opportunity?
Again, it's really a great start into that journey. It's really about making sure that this product is delivering exactly what the expectations of the customers are. I think the demand for it is very, very solid. I do think that this partnership is only one of several steps that will happen in the industry. There is a big waste in the industry of every single customer having to rebuild pipes between different key applications again and again. These pipes tend to be broken on upgrades. They need to be reconfigured, maintained. Data is shuffled back and forth at high cost, duplicated storage efforts. I think there will be more and more alliances where we find these partnerships. We did one with Databricks, as you described, which is off to a great start. We did others in different shapes.
I mean, we also have now an agreement with Palantir, how we work together. Whenever we have a certain degree of complementarity in the applications, we will try to move the effort of integrating these applications out of the SI customer space into our R&D lab and just do a kind of pre-configured plug-and-play solution. BDC is the first big step. You shouldn't be surprised if there are other steps following. Of course, there are obvious competitors we will never integrate with because I always jokingly say, if Workday says HR and finance need to be integrated, I tell them, yes, you're absolutely right. Why should I integrate with them? I let the customers happily waste money on trying to integrate these systems. If we have a reasonable degree of complementarity with other vendors, I think we will continue to negotiate these types of partnerships.
I think the cases we are showing so far show that these partners understand the data gravity we have in our systems and the value we can bring to the table to be open to these types of partnership discussions.
Got it. Maybe rounding all this off, you said that you're pretty confident around an acceleration in the total revenue growth.
Can you help us understand the building blocks between there's obviously going to be a maintenance fade in the maintenance growth rate? There is the cloud revenue, which is increasingly becoming a big part of the mix.
Yeah.
How should we think of those building blocks as you build towards the kind of low double-digit immediate?
Honestly, the best way to come to these numbers is probably to start with the many quarters you now have anyhow available on the key buckets of the revenues. For me, the biggest driver is the question on how high can we fly with cloud ERP suite for how long?
Yeah.
I mean, they're 14 quarters in a row, 30% plus. I mentioned in my simplistic extrapolation, kind of pull your spreadsheet to the right logic, we have quite some room for some deceleration. That's all that matters. Whether now the software is declining a percent faster or lower or the maintenance is not the biggest driver. The biggest driver is because this thing is getting bigger and bigger and has the highest growth rates. That is what really moves the needle biggest time. This is what I think is worthwhile watching. There is, of course, a natural asymptotic point at some point in time for that bucket because the market is growing at, what is it, 70%, 80% depending on what you look at. We are running at 30% plus. At some point in time, that either we continue to regain market share big time there.
Yeah.
We need to kind of converge to that. The good news is there is still a huge remain to do. All this math excludes SAP Business Data Cloud, excludes the mid-market opportunity because it's not really meaningful at this point in time, but will be in the outer years of a planning horizon. AI, that's the key question. How much will AI allow us to put some incremental market growth by absorbing that kind of transformation of personnel expense into IT expense?
Yeah.
How much will that be table stakes? I mean, that's a little bit anyone's guess. I think that question will be also driven by the differentiation of the product and also the competitive landscape. It's hard to imagine for me that if you look at our key competitors and think about Oracle and SAP, these types of embedded AI solutions, we will not have SAP AI solutions running on Oracle or the other way around.
Yeah.
There is a certain stickiness to the suite, and that gives me some hope that of that monetization, at least some share of that we can take for us. That gives another opportunity. This is why I'd say this kind of 2026, 2027 acceleration story is very much the extrapolation from the near term. What's happening in the outer years is very much, I mean, how much of these incremental growth drivers can we bring to bear? The good news, it's sizable. If you just would assume $1 billion for each of them, that's $3 billion on that revenue base, that is already kind of mid-single-digit plus percentage point. We could actually afford an attrition massively on the other stuff.
I have, I mean, given that we are running at 30% plus on cloud ERP suite, which is a core core, why should I kind of have any concern about not being able to grow better than the market?
OK. Just coming back to the margins, you've obviously outlined a number of efficiency initiatives in the past 12 months. It's about running the company more lean, more efficiently than cost cutting because at the same time, you've been investing back in the business.
You've also talked about kind of using AI internally. I think you referenced kind of an 80% to 90% OpEx growth relative to revenues. Where are we in that kind of optimization journey, and how much, you know, beyond is it something that can last beyond the next couple of years?
I think this is a very sustainable long-term theme. You've seen that in the more recent past, we've been better than that, of course.
Yeah.
Because we did the massive restructuring, which cost us also $3.2 billion. Now, what we really want to do is to come from that kind of one-off deep cut logic into a perpetual continuous optimization logic. It was quoted a little bit contentiously in the press when I said it's like brushing teeth because it sounded like I would kind of make it look innocent if we have to adjust parts of the workforce. What I really wanted to say with that is it's kind of better to anticipate. You know, you don't want to have problems and go to the doctor. You want to kind of tackle the problem before it even occurs.
That kind of proactive stance in a company that is growing fast, where you have more load on all the systems, that's the beauty of any strong growth SaaS model, is we don't need to have a bloodbath in our employee base because the growth is so high. Take finance as an example. We have more volume. We have more regulation. We have fast-changing business models. Of course, we want to be leading in productivity. I can tell my colleagues, guys, I want to see productivity from every single department of SAP. We can now adjust the workforce gradually. Also, we said that for these continuous adjustment efforts, we were not going to push that outside the non-IFRS operating profit, but embark that. You will see a couple hundred million roundabouts of expense in Q3 for that.
It will, of course, give us an even steeper gradient than to secure that we are hitting that 80% to 90%. We are not changing the numbers because of absorbing $200 million.
Yeah.
We can afford that. We create the room in the P&L by these measures to invest even more in productivity, invest even more in growth. That's the current philosophy. So far, touch wood, we have been on a very good trajectory and have been extremely close, if not above, the times we've set on these programs to start with.
Got it. We have time for a couple of quick questions from the audience if there are any. One right at the back or no? Cool. Maybe I'll continue. Free cash flow is another area of focus you've had since you've come to the company. Obviously, you mentioned restructuring as kind of perhaps masked already some underlying improvements in the business. There's obviously FX hedging. There are some of these migration incentives that have been transitory impacts. Can you help lay out those building blocks? Perhaps, are migration incentives something that are necessary? Is it simply just something that you need to sort of catalyze some of the customer behavior?
The most important message on free cash flow I want to leave you with is think of our free cash flow like we take the non-IFRS operating profit. We slam the tax rate on top of it, which is around about 30%. We should work on that to bring it down, but it is what it is right now. There is a structural delta between P&L on stock-based compensation and cash out because we equity settle a large share of that. That's the formula. I always say if we see each other five years down the road and you take the cumulative free cash flow of SAP and the cumulative non-IFRS operating profit, that formula should sit. That's the guidance to give you some very simple back of the envelope framework. Of course, in every single year, there are puts and takes.
For instance, this year, we are still burdened by $800 million of restructuring expense from what we pretty much concluded January 1, 2024, or early this year. We had a very beneficial tax position on cash this year because we were investing $3.2 billion in the restructuring, turning the company into losses in the home market, Germany, which then also resulted in some withholding tax we couldn't offset and some tax loss carryforwards which we can offset now, but cannot, I mean, these tax loss carryforwards will be used this year. There is nothing in terms of tax loss carryforwards for next year. You mentioned the transformation credits. The transformation credits by design have a cash conversion of 1. The EBIT net of tax will fall down to the bottom line.
They have phasing topics in it because depending on when exactly the customer calls off these credits, at that point in time, you will have a kind of lower cash conversion. In the other periods, you will have a higher cash conversion. It's a pure phasing topic. If you go back to my trend line, I would say 2025 is a quite favorable year or actually neutral year because we have these $800 million restructuring to digest. The other topics help us a little bit. 2026 will be a little bit more difficult. We also have some really operational improvement opportunities, for instance, on overdues on accounts receivable. You would be surprised how many customers pay SAP late. I'm really on a crusade to tell them that's not OK because if you want to have availability and timely service from SAP, you should pay me on time.
We introduce stuff like late interest charges and so forth to really chase that. I'd say the noise will be there around the years. We try to kind of stick to that yardstick as much as we can to not confuse people too much and keep it simple. We think it's the kind of cash conversion that's embarking the model by design. You don't need massive CapEx or massive working capital investment to grow the company. We're also not generating tons of working capital benefits like a retailer does. Our model is kind of neutral. It's really converting the cash. The biggest delta in cash conversion that's a little bit abnormal is anything that's equity settled on SBC, of course, is helping us on free cash flow.
Yeah. Maybe just to close out, obviously, the absolute cash generation of the business is still growing quite nicely. How do you think about the use of that cash between M&A, you know, returning cash to shareholders?
I mean, we like embarking more companies into SAP when it fits. I can tell you, while you have seen some smaller companies announced, so far since the 2.5 years I'm here, there was certainly a higher number of discussions where we had very serious approaches to companies and said, wouldn't it make sense to be sold to us? We couldn't agree on price. You also should have the confidence that we want to see a good business case.
Yeah.
The good news is we have no burning holes in the portfolio that we need to desperately do stuff to fix problems.
Yeah.
That puts us in the position that we are a tough negotiator and do the traditional deal math on NPV and some other metrics. Not always does it work. M&A, I mean, organic growth is the first. With the model we just discussed, we will throw off cash. M&A, if we have good opportunities and we have a reasonable agreement with the seller to see that that kind of transaction has the potential to create more value for our shareholders than simply share repurchases, that's the preference. If we don't manage to find these targets, share repurchases are the next logical consequence of that. There is also, of course, a recurring dividend. There is 40%+ of the recurring net income, or the non-IFRS net income. That's the pecking order. No change on that.
The consequence is if there is no big M&A, I have no good argument not to say we need to continue the share repurchase type of avenue. That's something we need to discuss with the Supervisory Board. I think as the old program is expiring end of this year, around the turn of the year will be the right timing to announce more on that front.
Great. I think with that, we're right on schedule. Thank you very much, Dominik. You had great insights.
Thank you.
Have a great day.
Thank you. For joining us. Thank you.