SAP SE (SAP) Earnings Call Transcript & Summary

So good morning, everyone. I'm Michael Briest. I head up European technology research at UBS. Delighted to be here in Phoenix, joined by Dominik Asam, who is the CFO of SAP. So we're going to go through some questions. There's an iPad beside me if you want to send in any questions. I'll put those to Dominik.

But Q3 result's about a month ago now. I think investors at this conference elsewhere, there's a lot of concern around macro. SAP is proving to be very robust. What do you attribute the strength of demand and confidence that you have in your outlook through not just this year but through 2025?

From our perspective, what really sustains our growth is the desire of our customers to future-proof themselves for their digital -- by virtue of digital transformation. We see that again and again. Of course, it's visible and clear to all the major customers that the old on-prem model has its shortfalls, which will position them not ideally for anything that's happening with AI. The on-prem model creates inertia because you benefit from the most recent technologies only when you do the upgrades. And these upgrades in the past have occurred maybe twice a decade or so and not every year or even every day or a quarter and of course, the need to have a very, very strong data management on all the transactional data that's in the ERP system. So that allows us to really turn all these large enterprises and some medium-sized companies. We have already on-prem into cloud customers. I'd say it's rarely a discussion as to if they have to do it. It's more about when exactly. And there's only a small part of the business, which is really strongly affected by macro, which is the transactional business where we don't get a subscription fee but where you're charging as per consumption. So when business travel goes down because you are in a soft macro environment and the overages are not paid because people are traveling a little less, when temps are sent home because you want to get rid of some fixed cost, it's variable cost in this case actually, that is hitting us a little bit. And that has been a part of the business which was probably less than 5% of the cloud revenues but not growing at all, but that's the only area where we felt it. And this is, by the way, also pretty much the explanation why we are growing a little bit lower in revenues than what we've seen at CCD in the forward-looking current cloud backlog because that transactional part is not committed. It's really paid as they consume these products.

Okay. I realize I haven't read the safe harbor, so I'll never get to jump in Investor Relations, but I'll just quickly do that. During this presentation, we'll make forward-looking statements, which are predictions, projections or other statements about future events. These statements are based on assumptions that are subject to risks and uncertainties that cause -- could cause actual results and outcomes to materially differ. Additional information regarding these risks and uncertainties may be found in SAP's filings with the SEC, including, but not limited to, the Risk Factors section of SAP's 2022 annual report on 20-F. Thank you for bearing with me on that. So just continuing on the near-term outlook, 23% to 24% cloud growth this year. The high end of that does require a very strong Q4. Can you talk through the building blocks in terms of maybe LeanIX, Litmos and those assumptions on the transactional business?

Yes. Obviously, last year, we divested a small business, I think, was effective 1st of December. So with 1 month -- 2 months of revenue still in there, and this will then be entirely eliminated. And that was a headwind, of course, in our year-on-year comparables on the cloud revenues. This year, I think we closed LeanIX a couple of days ago, a couple of weeks ago, very -- so we have a little bit of a pickup there. And more importantly, what is moving the needle is now, would there be a big snapback on the transactional side? So we can only reach the high end of the guidance if that was the case. Now you can judge from the most recent macro if that might be the case or not. So clearly, it would require a very, very positive development on that front. It is certainly the parameter within our guidance, which has the highest ambition level. But we are comfortable that we can hit that EUR 14 billion low end. And otherwise, we would need some kind of luck to boost it beyond that by having a bit of a tailwind from transactional.

Okay. And then sort of going to a micro level perhaps. I get a lot of questions on RISE. What's included, what's not, how it comes into your books? If we think of a customer maybe with, I don't know, a dozen ERP systems, maybe BW some line of business quite customized. If they signed a RISE deal with you today, when would you probably start to recognize revenues, when would it go into backlog? And perhaps in 3 or 4 years' time, how much bigger would the revenues be when they've completed that migration than in year 1?

So let me start with the last question, which is what's the conversion factor of a typical kind of software maintenance revenue base into that cloud revenue, it's about -- it's more than 2x, 2x plus to 3x, I'd say. And that, of course, shows you that in terms of gross profit from that relationship, we are improving because, yes, the gross margin on on-prem is very high, higher than in the cloud or private cloud in particular. But if you can convert a 2x plus the absolute euros of profitability of gross margin from that, of course, is significantly higher. Now in terms of the journeys, they're all very different. I mean these are complicated, large enterprises in many cases. And you're right, they might have actually dozens of ERP systems and some, because of lack of time, just to lift the shift and try to convert to the cloud as is others, the more ambitious one tried to make it a really comprehensive transformation approach where they redesign the processes. So as -- if you go all the way to a public standardized cloud, you need to come to what is called a clean core, a standard ERP system. They basically have to transform the actual company to fit the process, not the other way around. And that takes time. So with large companies, it can be as much as 7 years actually. So the portfolio granularity we have on that conversion journey from on-prem to cloud is layering not only of all the logos, which are already very diversified. But these logos also gradually ramp. So the revenues tend to be recognized quite quickly, but not the ACV, but something lower, and then it's ramping. We have some customers where we might have a pretty stable consumption over the term, but the more complicated customers are really starting to turn one instance after the other from on-prem to cloud, i.e. the ACV is kind of somewhere in the middle. They start with a lower consumption, and they end with a much higher consumption. And that ramp gives us also some visibility into the coming years. And the CCB, the current cloud backlog is actually looking at the rolling 12 months ahead. So whatever comes into that time window on these ramps is then shown in CCB.

And if that customer signed the deal today, that would go into CCB day 1? Or does it have to...

Only the -- it has to be -- it has -- well, we are activating. It's not a problem to provision, but we provision very fast, actually even our incentives to the sales team to make sure there's enough incentives to make sure once it's signed, it's also provisioned and then we can charge subscription. But there is a profile agreed with the customer, which might be very different in the early innings. I have one extreme example on top of my head where the CCB was, I think, a low teens million number and the TCV was EUR 300 million, EUR 400 million. So that's now an extreme example where we really have like that such a ramp because a very complicated company where they actually start easy pass first with some easy small instances, some, I don't know, sales companies abroad and then tackle the bigger companies later. And it depends on really how they structure the transformation journey, how that kind of comes together.

And the profitability profile, you mentioned private cloud is lower. That initial transformation where you -- obviously, the clients got a lot on their own place to deal with, but you're helping them. Does the profitability noticeably improve as they move through the sales of stage core even if they never get to public cloud?

It does. We see good margin expansion opportunities. I'd actually say the lion's share of the margin expansion opportunities we've materialized so far. And you can see that over the last 10 or 20 -- 10, 15 quarters, I'd say, is, A, the famous cloud conversion program where we really harmonized our infrastructure. That gives us a very significant boost because we had a lot of investments, which then turns into lower cost. And then there's a kind of economies of scale game where we really improve our own gradual operating model. There's a lot of work to be done between the pure infrastructure like the hyperscaler services we buy and the cloud service itself. There are managed services between security and so forth. And of course, it's getting easier and easier the bigger you become. And as the private cloud is growing so fast, it gives us a big opportunity to gradually expand the margin on the private cloud. And then ultimately, we also believe that because it's an even more attractive financial model, private cloud customers will ultimately migrate to public cloud. And there, again, they will start with a low-hanging fruit with the easier companies in their groups. With public cloud, we've seen even some more complicated manufacturing companies doing that already. So for me, it's more a question as to when that happens, not so much if it has.

That's helpful. And when we look at the cloud revenues, the SaaS element of it, I think a lot of analysts, myself included, would look at the S/4 growth, that's fantastic. BTP is comparable. But then the rest of the business, including the business network, SuccessFactors is growing sort of high single digits, maybe 10% in a good quarter. What's the outlook for that part of the business?

It's an excellent question, and I'm mulling over this myself also in terms of how we need to communicate to investors because I actually think there is a big upside in what you described. Let me explain why. I mean, we have that super fast-growing S/4, which is really what we call the lending. It means the customer decides. I put my whole ERP system on that basic infrastructure. And once it's in there, it's an extremely powerful tool for all kinds of transactional business. Then they buy the BTP, very high attach rate, I'd say 80% or so in the large enterprise space. And that means they also say, I use that BTP to come to the standard -- back to the standard to put my customizations and extensions in a way that will not slow down my upgrades going forward. But I can really upgrade a cleaner and cleaner core very fast to make sure I don't have that delay consuming innovation. And there you see -- I mean you've seen 46% PaaS growth, constant currencies in Q3. I think we've seen 77% S/4 growth. Now it's true that the rest of the sum of SaaS and PaaS is growing more slowly. But even there, you have to look into what's actually growing faster and what's growing more slowly. And there is one correlation which is kind of overwhelming. Everything where the transactional thing plays a big role, really ERP, where goods, people, money is moving back and forth and basically resulting in debits and credits. We're actually growing quite fast and make a lot of money. Everything which is more psychological, I call it, like learning some other fields of CRM where, frankly, we have small market share, and we have very strong competitors there, it's much tougher. Now why do I see that as an opportunity? If I now define a large kind of core ERP and SaaS suite, and this is really what we drive with the customer is this land and then expand, where there are synergies because it's all transactional. It's all embedded in the S/4 database. This accounts for about 3/4 of the revenue base. And that stuff is actually growing very fast, way above 30%. And the rest is actually less profitable, doesn't grow much, and we have a relatively [ earmarked ]. And that's actually an opportunity because what it does mathematically is if you look at that kind of large chunk of revenues, it's growing faster. So if you have 2 companies, one is basically 23% SaaS growth, everything equal throughout the portfolio and you have one which is actually growing in that core, more than 30% and that core is becoming bigger and bigger, the mathematical consequence is that the latter company will outgrow the former company. So this is why we are not too nervous about our growth opportunity because where it matters, we grow a lot.

And I suppose that would raise the question inevitably about the portfolio, you have exited things in the past, for example. Would you consider that more broadly for some of these other lower-growth assets?

Look, I mean, you have always as a CFO to think about where do I prioritize my scarce resources. And of course, it's more challenging to do that if you have slow-growing, low-profitability businesses with a relatively low market share. Now of course, there is also some bundling logic. For instance, on the HR side, you say -- you can say if I have payroll in the HR system, don't I want to have learning in there as we have. So it might not be an opportunity to kind of come to quick wins. But over time, as the big part of the transactional business is becoming bigger and bigger and will be augmented by AI to connect all the dots to some very powerful use cases, I'm confident that the overall growth trajectory will actually be very, very healthy even if we are not divesting or shutting down these type of assets because they might actually throw off a little bit of cash, and they don't burn a hole into our pocket. And it really depends on the actionable scenarios what you can do with it. But would I as a CFO put kind of a lot of R&D money into this, no.

Understood. And the question I get a lot is on the migration to S/4HANA. I think your predecessor said when S/4 was launched in 2015 that you had about 35,000 ERP customers. I think you've announced 20,000-plus sets for [indiscernible], which maybe half or a bit more on half and net new. Where are you on that ECC6 installed base in terms of how many -- how much of the dollar spend in maintenance is yet to migrate?

I think it's very roughly another 2/3 to go. We have already been progressing well. And don't forget what we see in the cloud revenues today is not necessarily cloud deals we signed. It's cloud deals which are now already up and running and built. So if you take the combination of that, and you can also [indiscernible] in some way, if you look at the kind of cloud revenues and you strip out what's not kind of ERP but other stuff, and on the other side, you look at maintenance, what's not ERP and other stuff and you assume kind of similar ratio just from the revenue numbers, if you assume a 2 to 3x conversion of maintenance into cloud revenue, that's also what you come up with. So that journey is still -- has a very long runway. And on top of that, we are adding -- now that we really have a powerful public product, which is also suitable for medium-sized enterprises, we start from an embryotic base, I'd say, on the smaller companies with GROW, which is the other big program we have. While the euro contribution numbers are not so meaningful yet, it's growing more than 100%. And if you look at the net new names, we actually have more than half of the new names on that kind of public cloud GROW journey. So the combination of the 2 gives us great confidence that the growth story we talk about is not a growth story to hit our 2025 ambition. But actually, the second half of the decade is benefiting -- still benefiting big time from the rollover. Even people who have gotten S/4 on-prem and can use that license through 2040, sooner or later have to think about how do I future-proof my business because the on-prem model itself has strong severe limitations, which are not -- I don't believe they are sustainable. There might be some exotic businesses or government customers who might do that through 2040. But I cannot imagine that everybody will sit there and not benefit from AI and all the things you can do with cloud. So that conversion is intact. We have then, of course, the advent of AI, where we are now starting to commercialize use cases. Of course, that's also starting from a very low base but with a very high growth potential. So I'm not really nervous at all about the sustainability of our growth journey in the second half of the decade.

And sort of following up on AI, you talked -- or Christian, I think, talked a couple of quarters back about sort of a 30% premium price point. Customers are clearly going to expect to have ROI. Can you talk to that and maybe how long it will take you to infuse AI through the portfolio?

I mean the premium offering is basically a little bit of a bundle logic. And by the way, that bundle logic is something we'll hear more about and -- because we have that unique capability of giving that flexibility to the customer without any negative implications for us because we have a broad product portfolio. So on that front, the 30% premium is really a reflection of a combination of products like sustainability being included, SAP Analytics Cloud, SAP Analytics Cloud being included and some AI consumption points. Now what has to happen is that we are now creating enough use cases that these AI credits are consumed. So for our 2025 ambition, honestly, I don't need much of that to get there. I think the story of the AI credits being consumed will drive our revenues in the second half of the decade. The current growth is very much driven by the platform choice, and that platform choice is super important for us because it drives that famous transition from on-prem into our RISE offering. And it lays the foundation to be able to monetize these credits going forward. So I would argue that a large share of our revenue base today is already driven by AI because people are not investing in this SAP platform just because of doing plain-vanilla ERP stuff, but because they are convinced that it's a future-proof platform that with the S/4 architecture and the granularity it can provide, they will get much better use cases embedded in that infrastructure than from other providers.

If we can sort of dwell on that for a little bit. I think a lot of investors might intuitively see the hyperscalers as the likely providers of most AI platform use cases. Can you address the concern an investor might have that a large customer might just say buckle their data out of SAP and drop it into Microsoft or AWS? Also what intellectual property protection do you have about the processes that transform that data that mean a customer can't just do what they want with it.

I don't want to hold any customer hostage to what they could theoretically do by shutting down the data for other users. To the contrary, we sell BTP, Datasphere and products like that because they enable the customer actually to get some data. And we sell BTP also because it has some powerful application programming interface that would allow our customers to feed in data from some Microsoft system or other system into our database and extract some data from our database. Now the way I try to simplify things for financial investors is I use an analogy. If you think about a complicated scenario analysis, you have to run. And all of you probably still do that on spreadsheets for your companies you cover. And you have really invested a lot of time to program a macro and really become very granular and run some scenarios on a nice Excel spreadsheet. You have to understand an ERP system is much more complicated than that. It's like a spreadsheet that's updated every millisecond where all the history of the spreadsheet is kept, every millisecond for every single transaction, actually, you can basically look at a spreadsheet, but you can also go into what the guy has been inputting into every cell. And you can say that cell can only be looked at by this person or that person or that person. Now you say, "Well, I need some data from that spreadsheet," and you copy that data. You open a new spreadsheet, which is now the analogy of a hyperscaler and they say, in a hyperscaler, I open new spreadsheet. I now paste column ABC into that spreadsheet and do some analytics on it. Artificial intelligence is all about trying to make the machine do the same things as humans can. If you as a human being have a choice between using that original spreadsheet for your scenario analysis or whatever analysis you do versus some cut and paste as of a certain point in time, I guess you would confirm and agree with me that the original spreadsheet has much more value. So you would embed that AI right there and think about opportunities like training our own foundation models on process improvement. I mean we have Signavio now. We can track the process landscape of all our customers. We have LeanIX now. We can screen the tech stack of our customers, and we can come up with recommendations to the customer on how to improve these processes. Now you tell me how a Microsoft or any other kind of AI vendor is going to do that in my system without the full power of the system. So how can you do that on the cut-and-paste values only data from that system? It's just not possible. There are examples where it works. And funnily, it's again a little segmented into that kind of hard transactional world and the more psychological world. If I want to write a job description, yes, we have in our SuccessFactors a tool where the customer can easily create a job description very efficiently. Could a customer also say, well, I take the API of SuccessFactors and I have some other pilot putting that language or that stuff into the system, yes. But that's a completely different ballgame from the hard transactional process-oriented stuff where the whole system is actually the ERP system of SAP. It's a little bit like you know that the formula says that number is 2x the formula to the left. You know there is this relationship. You don't have to run analytics on that. It's embedded in the system. So I think I tried to use that example to give you some tangible view why embedding the AI directly into our ERP system on the transactional side is just not something the hyperscaler would be able to do.

That's a good analogy, Dominik. Just a couple of questions now on margins and cash flow. So the cloud gross margin was nearly 74% last quarter. I think previously, Christian has talked about the lower margin profile of RISE private cloud. Given the mix shift to RISE and S/4, will it be a relatively straight run up to 76%?

I think it will be. I mean there's always fluctuations quarter-by-quarter, but I think on average, if you take a smoothing curve, it should be pretty much linear. Why? Because there was an acceleration in the last, say, 4, 5 quarters because of the cloud convergence project I have already mentioned, which caused a lot of extra cost. Now the harvesting is kind of complete. Now we are on the lower cost base already. So the rest is a more gradual grinding work on economies of scale and really making sure that everything becomes gradually more efficient, and that's a pretty steady process. So I don't have any better guidance on saying you connect the dot from whatever we print in 2023 for the gross margin for the year. I would caution you not to pick 1 single quarter because, again, there's ups and downs. But if you take the kind of full year gross margin and then you take what we've guided for 2025, I have no better idea than just putting the middle. Maybe there's a little -- well, if you have an exponential growth, of course, you have to be careful from a growth revenue side. So in euro terms, it's not a line, but in terms of percent improvement, I think should be quite linear.

Helpful. And then longer-term, beyond 2025, you mentioned AI. I think when you first arrived, you talked about efficiency opportunities in sales and marketing. Is there anything you can say about longer-term operating leverage beyond '25?

Yes. I think what our focus is right now in the planning process, which is a traditional 5-year planning process is to say we have 2 objectives. One is to make sure that we derisk and protect the 2025 ambition as much as possible. So we take all the measures that are required to make sure it's happening. And secondly, to make sure that the gradient in terms of revenue and earnings growth from '25 into the following years is as deep as possible. So that's what we're currently doing. And I gave you that acceleration mathematics. I think that's the biggest tangible argument I have for investors to explain why we are not shy about our growth numbers between -- beyond '25 because if you have a revenue which is currently kind of -- let's say 8%, 9% in the last quarters, and you have still a headwind from licenses and the license becomes smaller and smaller, then you look at the cloud revenues and in the cloud revenues, you have the infrastructure, which is becoming smaller and smaller. And then even in the SaaS, PaaS part, you have like 3/4 of the business which are growing like hell in the kind of transactional part. And there is some businesses which represent the remaining 25% in that bucket, which are not growing so much and are actually less profitable. The trend is clearly our friend on revenues and also on margin, by the way, because we have -- the base that's really driving our top line is super strong and healthy. And yes, we have some headwinds from certain businesses where the situation is not as rosy, but these will ease simply because these businesses will be diluted out in the mix. And so it's kind of washed out -- our problems are going to be washed out over time. And if we can achieve already today these type of cloud revenue growth numbers and the CCB you've seen over the last quarter, why should we be nervous to also reach that in '26, '27.

That's helpful. And on cash flow, I think you're planning to add EUR 3 billion in EBIT from '22 to '25, EUR 2.5 billion of free cash flow. The tax man or tax woman would normally take a good chunk of that. That's quite a high conversion rate you're assuming. What is the underpinnings or opportunities in, say, working capital or CapEx that you see?

There are really quite some opportunities. For instance, we collect cash quite late. It has not been at the core of our operational thinking. It was really more about profit maximization driving the top line. And sometimes people have been more accommodating with customers on that front. To the contrary on the accounts table, we've faced quite early. We do even sometimes big prepayments with hyperscalers because we get a little bit of a discount. But if I look at the IRR, it's actually at best at cost of capital. So I wonder why the hell should I just turn money. We do stuff like buying company cars for our employees. On CapEx, where you could also lease these cars and then you only put the capitalized lease value on the balance sheet for 3 years, and you don't have the full value of the car. So a lot of operational nitty-gritty stuff we can do. And I think SAP has been always very good when the company has been recognizing a problem and say, this is a problem. We have to take it seriously and think about the cloud conversion program. It was much more complicated, frankly, than what we need to do on these working capital items. And once people say, there is a point, and let's go for this, and I think we have it clearly under control. There's also, frankly, a little bit of a tailwind from the famous stock-based compensation, where actually the -- we decided to go for the lion's share now on equity settled. That doesn't help you much from a value creation point of view because ultimately, we need to repurchase the shares. But at least from free cash flow, it's actually making our life not so difficult. So I'm not super nervous about our, I'd say, my confidence level on the guidance for 2025 on cash flow is actually quite solid because it's actually the number, if you think about it, where we have the most opportunities to influence it still. Top line, you just can get orders in, but bottom line, you can do so much more. And that's -- I think there's plenty to draw on.

We're up on time, but just a final quick one, I think. I think you had this at other fireside chats as well, but you've been at SAP for not even a year now, but what are the biggest positive surprise how SAP's run, and where is the biggest opportunity, you think?

So I think really creating that consistent system of taking complicated large enterprises on that cloud journey and how well it works in light of the complexities these customers are facing. I've been sitting on the customer side for many years, seeing what a complication, what legacy I had to deal with that actually, we have the tools now to do it. I think SAP used to talk about it, but 3, 4 years back, it was not there. Then since October 2020, there was so much investment on that. And I think we can now say that while it might not be perfect yet, it's by far the best tool in the industry to do what I call slicing the elephant to gradually move these companies in the direction of private cloud first and then public cloud. And at the same time, the traction with the smaller customers is also very promising. So I'd say my investment hypotheses of coming to SAP are fully validated by what I've seen today. I think you hinted to it, the one thing where we can certainly improve is on operational excellence. I mean all these working capital items are really ultimately operational excellence topic. So I think a very strong strategic position with a strong product with some good self-help opportunities. That's what gets me excited about our company.

Okay. Well, we appreciate your insights, Dominik. Thank you, everybody, for coming.

Thanks for having me. Appreciate it. Thanks for being here.