Right, perfect. Let me open things up for this next session. My name's Adam Wood. I look after European software research at Morgan Stanley. It's a great pleasure to have Dominik Asam with us. Dominik, CFO of SAP. Thank you very much for joining us here.
Well, thanks for having me.
Thanks. So just to get the regulatory stuff out of the way to start off with. So during this presentation, this is on the SAP side. SAP will make forward-looking statements, which are predictions, projections, or other statements about future events. These statements are based on current expectations, forecasts, and assumptions that are subject to risks and uncertainties that could cause actual results and outcomes to materially differ. Additional information regarding these risks and uncertainties may be found in our filings with the Securities and Exchange Commission, including but not limited to the risk factors section of SAP's 2023 annual report on Form 20-F. From the Morgan Stanley side, please, for important Morgan Stanley conflict disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/research/disclosures. I can slow down now and there, going through the problem.
That's it.
Okay. So to kick off with, I think it's very interesting that you've been in the CFO seat just over a year, coming up for a year at SAP. You came in, you weren't from SAP, you weren't from a software background. So I guess that gives you quite a different perspective in what's possible, what could be achieved. Could you maybe just talk about, over the last year, the main two or three important things that you've seen, that you've thought we can do things differently, we can improve on things, that maybe that outsider perspective gave you the opportunity that wouldn't have come from internal?
Yeah. And I think I was lucky that basically the investment hypothesis, so to speak, I've made for myself by joining SAP turned out to be quite solid. The company, as you all know, has been in a very challenging transition from on-prem business on ERP to a cloud model, with the special part of it that ERP, for many large enterprises, is still a kind of private cloud model. And I think we've made extremely good progress on that front. So I'd say that is not something I can take credit for because I've only joined a year ago. But the fact that we really see now the top line inflecting because of that transition being successful was even more powerful than what I thought.
Now, in terms of margins and cost, it's not a surprise that in times where you go through this transformation, you push revenues to the right because of the switch from on-prem upfront licenses to regular subscriptions, recurring subscriptions, and the investment in our platform. These were very difficult times, but it's also very comforting to see that we are now really starting to harvest the fruit from that investment. So we can be now really looking at how can we bring the margin level to a more competitive level and also make the company better in terms of scalability. In that time where we had to go through this transition, we've been basically growing the cost base faster than the revenues, which is, of course, not sustainable.
And we want to do two things with our restructuring program, which we've recently announced, which is bringing in one big step through 2025 the margin to a higher level, and then also decoupling the cost base more. And I think this is where there's opportunity. And the last thing, if I may, I want to mention is the focus on cash. Because the company was so much focused on delivering these milestones that were already communicated in 2020, which were predominantly around turning the corner on growth and turning around the corner on double-digit non-IFRS operating profit growth, there was not so much focus on free cash flow. And I think there's a lot we can do over the next years to bring the cash conversion up.
So maybe going through those in order, you talked about the acceleration of the top line. The real engine of that has been the S/4HANA migration that's really accelerating, putting some incredible growth numbers up at scale there. Could you just talk a little bit about it from a customer perspective? What is it that's changed in that product? What are the benefits that customers are getting that's driving that adoption and that growth in S/4?
I think the first important discussion, and frankly, it even changed over almost a year I've been with the company now, is should we go into the cloud or should we stay on-prem? And of course, the challenge with the ERP system is that it is a quite complicated product. So every customer is different. There are many processes that have been created over years. They are related to some physical infrastructure, to the way a lot of people operate, how the supply chain is built. So we were always cursed by the fact that our software was procured, and then people have been spending 10 times the money they spend for our software to modify it.
That meant that every upgrade was a big pain for the customer, meaning that they would need to invest a multiple of the money in just making sure that all the customizations would be upgraded in line with the SAP product. That had, as a result, that the customer could consume our innovation just five, 10 years later than it was available by SAP. And with the move to the cloud, that is, of course, changing. Also, AI has created an awareness that we will not be able to train these models if we are not having access to all that data in the cloud, data of our customers in the cloud. And of course, they have to give us their consent. If they want to have the access to these learnings in the cloud, they give us the consent.
So I think the story that we should be on the cloud is now understood by almost everyone. And now it's about really showing them the value in that transition. And here we can leverage our process mining tools like Signavio. We can also use our enterprise architecture management tools like LeanIX to come from a move where we say, well, you have to do an upgrade. We have to go to the cloud to say, where's the real business value for you?
And this is a new value-added proposition, which is resonating very well. And if you look at the S/4 success, and I would say it's not quite right. I think our success is by now much broader than S/4 only. We have recently introduced a key metric, which we call Cloud ERP Suite. It's basically all the ERP products that are transactional in nature.
They represented in Q4 of last year 83% of our combined SaaS and PaaS revenues. And the biggest aha moment for me at SAP was looking at that number over the last two years because it was 31% in the first two quarters of 2022. And then it was, what, 33% ± a percentage point for the six quarters thereafter. I've never in my life seen such a high growth number with such a stability in terms of growth rate. And this shows it's not only S/4, which is a little bit more volatile, frankly. There's also the BTP in there. There's also Signavio in there. There's LeanIX now in there. And there is, of course, also the cross-sell and the upsell in there that once we have landed on S/4, we can also expand.
So the next leg of the journey is basically to transform these customers in customers that are consuming more of our products. And we have now really created that measure, which is for me the key metric on the revenue side quarter by quarter, to see how well can we sustain that super high growth rate, winning market share on the cloud ERP business. Because I think we can all agree that 33% is not what you have as a market growth. So I think it's the best testimony that this cloud story really works out.
I think when we speak to partners, you alluded to it. This integration that's happened between the core ERP and the adjacent solutions has really improved. I mean, is that the main metric we should be looking at, that Cloud ERP Suite growth? Is there anything else you can help us with around cross-sell of these products into the base?
Yeah, I think it's indeed the key metric because this number will give you, well, it's 83% of the revenues in the cloud. SaaS and PaaS, you know that infrastructure as a service is something we kind of de-emphasize. It's a highly capital-intensive business. There are very mighty competitors in the hyperscaler space. So we don't feel it's wise for us to invest too much money there. So that kind of 83% of SaaS, PaaS revenue will represent a larger and larger share of the SaaS, PaaS, IaaS, i.e., the entire cloud revenue base. And then the software and license business and, sorry, the maintenance revenues will become smaller and smaller. So I would say this is the key metric to focus on. I think, of course, there's also the forward-looking metric, which is quite interesting on the CCB.
I would say that the lion's share of the CCB growth comes from that kind of bucket.
When we think about, we know that a lot of customers are doing large multi-year deals with you, and there are step-ups in their consumption in years two, three, four. Often these are four, five, six, or plus-year deals. Could you just help us with the visibility you have and the confidence you can sustain those types of growth rates on Cloud ERP with the comfort that you have from how customers have been signing up already?
Yeah. I mean, of course, we've gone, as any company, through a comprehensive bottom-up five-year planning cycle. So I can confirm that we have a certain confidence that by virtue of still tackling the large base of customers we need to convert on the RISE journey, the RISE journey being the private cloud, large enterprises that we bring from on-prem to cloud, plus then the fast-growing emerging business, where basically before we have not really been present. I mean, it was almost prohibitive for a smaller company to install an SAP software because of the high integration cost.
But now with the public product, which we call the GROW Initiative, we see really high growth, I have to say, from an embryonic base. But the growth numbers are kind of triple-digit percent in 2023, so over 100% growth, so more than doubling last year.
So it's one of these flywheels we are starting to spin. Once that will have some size, it will also contribute to the growth going forward. Last but not least, yes, with the advent of AI, while certainly for our 2025 ambition we've earmarked, we don't need any miracles on AI, frankly, for the second half of the decade, that is something that should help us sustain a high cloud ERP growth.
Perfect. That's really helpful. You talked about support and that gradually declining. But actually, I think the support line has surprised investors with its resilience so far. Could you just help us a little bit with, are there any unusual factors in there that could fall off? How much is pricing? Are customers kind of paying premium support while they do the migration? And once they've migrated and things are running, it starts to fall off? Or just help us with what's supporting that.
In terms of the numbers we've shown in 2023, maintenance was remarkably robust. Also because we have been increasing prices, we had a price cap increase January 1, 2023 of 3.5%. We've increased that to 5% January 1, 2024. Now, there are some jurisdictions where inflation rates have been lower. The formula is basically max 5% or CPI if lower. I think we had, of the 3.5% last year, we had a little bit more than two percentage points, which ultimately ended up in our revenues. I'd say the ratio in terms of fall-through of the max, if you haircut for the countries like Switzerland and so forth, we didn't have that high inflation, should give us also some significant boost.
So if you strip out that inflation premium, so to speak, then simply adjusting for inflation—and by the way, these are contractual rights we have. They are in our contracts. For some reason, which I don't quite understand, we have not utilized that in the past. But since two years, we do. You see, there is actually a decline in the maintenance base. And now, how that will evolve, I think there will be a gradual, slow increase in the decline rates, of course. Because as we move more and more customers onto cloud, that will happen. But let's not forget, the lion's share of these losses on maintenance will be converted to cloud. I'm really deeply convinced about that. And the typical factor we have is between 2x-3x. So 2.5x, we say, on average.
That means that both from a revenue point of view and from a gross margin point of view, we are better off converting these customers to the cloud. And this is also why we basically have a little bit of a schizophrenic commercial policy. We are really, really robust on on-prem. Because I always say the hardest competition we have is our own bring-your-own-license-on-hyperscaler software versus the cloud. And it means that we are really using all the leverage we have and all the contractual rights we have. We will enforce that on maintenance. We will also become more robust on software in terms of commercial concessions. We incentivize our sales force less on software sales.
We even provide migration credits to customers to make sure that we are smoothening that journey and are not kind of having too much of a clocked system in 2027 when we're running out of maintenance on ECC. We're really trying to roll that custom base gradually into our on-prem business. That gives us tremendous boost because we have a 2.5x conversion. The devil's advocate logic would be, OK, these customers, they run out of maintenance. Then maybe they go third-party maintenance at some point in time. This is a pretty big risk to take because you would basically run systems that are not certified anymore. It's basically insurance policy, which you will lose. At some point in time, it's really a business continuity issue. You cannot play that game forever. They will need to run on some ERP system.
Because I'm deeply convinced that those companies who procrastinate most are the ones who want to spend the least money necessary on ERP, it's getting for them more expensive if they change the system. I'd say we are a natural kind of recipient of these customers. I'm not very nervous about not being able to ultimately really convert the lion's share of that customer base to the cloud.
Perfect. That kind of brings us on to another debate. I wanted to ask about competition from two angles. I was asked in a presentation of a competitor. They were kind of talking about this move of the ERP and financials to cloud as a huge opportunity for them. This was more in the HR and financial space. Could you just talk a little bit about how you see that battle with competition? If customers are being forced to choose going to S/4, does that open them up to competitors? What are you seeing? What are win rates like?
Well, I'm not very nervous about that conversion. I come back to the Cloud ERP Suite definition. This is really all about the transactional systems. I mean, you might ask, what is not in Cloud ERP Suite? And if you look at the businesses that we have now in the so-called Extension Suite, these are businesses it's basically SuccessFactors excluding the transactional part, which is the core HR and the payroll. And it's the CRM excluding the transactional part, which is the commerce and the customer data system. So basically the data for the customers we need also to run e-commerce platforms. So what is outside is actually what I would call more content business. And the way I would summarize it, that we are super strong on winning market share on everything that's transactional.
I'm not intimidated by any of our competitors with my 33% Cloud ERP Suite on that front. I would always challenge them if they say they eat into our lunch. I'd say, yes, we have lower growth on the Extension Suite, which is more content business. So what do we have in terms of sales, marketing, services in there? We have on HR learning in there, which are sizable revenue buckets, as I mentioned, well, 17%, the rest versus the 83% of our SaaS, PaaS offering. There, frankly, we had not much growth in 2023 because we divested Litmos, which was part of that. We will have a little bit of a snapback in growth. But this is growing much more slowly. So I'd say where our competitive strength lies is really on the transactional part of these systems.
I think we are now starting to occupy some space, which was previously not tackled by us, which is the mid-market. I'd say as long as we can continue to run at that kind of growth rate, I wouldn't be too nervous about losing market share because I know the market overall is maybe growing the 20% or so.
The other discussion I increasingly have with investors is people are maybe a little actually a little bit less worried about that battle at the application layer. People are more interested in what's going on at the platform layer. They're saying it's really important for SAP to sell BTP and get customers to use it and not allow a hyperscaler like Microsoft to put its technology underneath. Is that something that SAP thinks about? How do you see that battle?
Absolutely. You would not be surprised that SAP belongs to those who say that the analytics, wherever possible, should be embedded in the application as opposed to some enterprise analytics on a higher layer. In the financial community, I always try to find some non-technical way of explaining that. I have not found a better example so far than using an Excel spreadsheet as an analogy and say, look, you could argue that an ERP system is a highly sophisticated development of an Excel spreadsheet where every single time somebody is entering something into some cell, it's a new line in the general ledger, the universal ledger.
The access rights are highly modified, which, by the way, is already one of the things these authorizations, if you extract data somewhere in an obscure data lake, it's very difficult to preserve that integrity in terms of access rights.
So you think about you modeling a company and having smart logic as to how the volume and the pricing and inflation and the cost, all that plays together. And you've spent a lot of time creating that logic between the thing. We call it the semantics around the data. And this is something these guys don't have access to. So they would need to reinvent the wheel, basically, with their analytics to find out what is actually the embedded logic because we are not exposing these semantics to these types of higher-level layers. But that's exactly, in my eyes, the right discussion. This is the big battle. I mean, the new data, the data is really the kind of gold of the system. And how do you exploit it?
And of course, it's our ambition because we are so convinced that the value, the gravity of data we have and the value of the data is super high that we also monetize that by giving easy-to-consume use cases to the customers. So I think just charging customers for egressing some data is not really what makes the customer happy because customers say, I've paid for your software, your cloud product. It's my data. But while it's the customer's data, the semantics are our semantics. And we have the liberty to decide to whom I expose those or not.
And we will also need some data from some Snowflake or some other people. So what is currently ongoing is that everybody is kind of bartering and saying, I need that from you. You need that from me. Who has the gravity of the data?
I, again, feel quite solid on that front that there's a lot of good stuff in the ERP system a lot of people want to have. It's nice to get the data only. But it's not sufficient to run powerful analytics.
What's the best way for investors to track whether you're succeeding there or not? Is it attached rates of BTP to customers? Is there any other thing that we should be looking at to track?
Again, for me, the best KPI is the quarterly progression on Cloud ERP Suite because it's kind of summarizing all of these things. The attached rate of BTP is, of course, growing. We are promoting that by bundling. I mean, you might have seen that our backlog is growing faster than the CCB. The long-term, the overall backlog, the TCB, is growing faster than the current cloud backlog. That means that the customers are pushing for longer contract durations. While that gives us some stability, for customers, it's actually not easy to predict. The further they move out, the more error margin they have around their consumption forecast. What we do more and more is we bundle all these things together and allow the customer to migrate from one product to the other if they underconsume on one and overconsume on something else.
I think that's a huge advantage we can play vis-à-vis narrow competitors who are just in one product because it's so damn hard for these customers to commit. If they order too much, so to speak, they have too much capacity. They pay too much. And if they underorder, we cream them on overages, which they don't want either. So I think that story of integration, which has only been activated seriously by SAP over the last years, is really playing out nicely, that it's a better commercial offering. It can offer you more interesting use cases. It reduces the cost for the customer to do that integration himself or herself. So we really want to create a big domain around these transactional processes where we want to occupy a very, very high market share.
Perfect. And you started talking around—you've talked about data. Who owns the data permissions? I guess that brings us nicely onto a discussion around Gen AI and what can be done. Can we maybe start there with what data can you use? What data do you have access to? A criticism is, well, a lot of the data is on-premise. Therefore, it's harder to access. If you could just talk a little bit around how you see that data access and permissioning.
I mean, as you say yourself, one of the key reasons that has really changed the debate around why do I need to go to the cloud is that one. Customers understand that for us, it's pretty useless if the data is scattered in some systems outside our cloud for us to train our models. And I think the biggest opportunity for us to leverage the data is really on the process side. We have about 300 million users on the planet. And these users sit, for instance, in shared service centers. And they do, say, an accounts payable process every day. And today, they are just manually kind of they're already starting to automate.
But here, what we can really do once we have all the customers on the cloud and we ask the customers to provide their consent, we tell the customer, if you want to leverage that learning, we will share that learning with you. But for us to give you the feedback from that learning, you will need to also expose your data in return, of course, in an anonymized fashion. And then we can basically learn from the tens of thousands of customers we have, from the 300 million users we have, when these people run through our processes, what are they doing? It's completely analogous to a large language model. A large language model is basically sequencing words. It says, when a person is saying that word, what's the next one that person will say?
In our case, the analogy where we now ground our own foundation models on our data is on the processes. Basically, we want to anticipate what that person, for instance, in the shared service center is doing next, and to the degree that we can completely automate it away or that if it's too complicated, for instance, if the goods receipts in that process deviate from the invoice we've received from the supplier and that deviates again from the contract, that the machine knows, oh, there's a problem. What is that person going to look for now? And then that person doesn't waste time to extract all these pieces of information. It needs to solve the issue.
But it really has that and maybe even already with a draft email to the supplier and said, "We have realized that what you sent to us is different from what's on the invoice. Please rectify." So this is just a very simple example of showing what AI will mean for us. There's also one aspect which is very dear and near to the CFO community, which is assurance. I give you a super interesting example. It's super difficult for companies, when people are leaving the company, to delete the access rights in other systems of these people.
And I'd argue that half of the issuers on the planet have a material deficiency on their controls on that. So ask a CFO whether that works in their company or not. And they will mostly say no. It's a trivial thing. And this, by the way, shows the value of integration.
If you have the integration, you can basically automate that and say, whenever a person is leaving the company, all the access rights will disappear. It sounds easy. It's not, I can tell you.
No, that's really helpful. Again, another debate around Gen AI is monetization. And some people will say, well, putting these tools into the applications is table stakes. It'll be given away for free. You're talking about getting 30% price uplifts as you haven't.
Careful. I mean, the 30% price uplift is for the so-called Premium Plus bundle, which is a combination of some incremental software. So it's bundling software like SAP Analytics Cloud and BTP and other things plus credits for AI. So it's really just the start of the journey. And the way the monetization should work is that basically, once we have developed a process and we have two versions of the process, one is the kind of traditional process, the best we can do without AI.
And then the other one is we plug a model into that system that is really bringing the performance of the process up massively. And then we can demonstrate with tools like Signavio to the customer, OK, this is your process today. Hopefully, it sucks. So we can improve it. This is the process if you transform the way we run it.
This is the process if you click on top. It's like on an arcade game where you want to have the better weapons to play with. And you just have to click that button. And then you can reduce your headcount by another 30% if you click that button. And then we start to monetize. I think this is the only reasonable way to monetize. I think the basis of monetization is to sell software. I mean, the tools we use for that type of stuff is BTP and Datasphere.
So already, the fact that every customer who wants to do that needs to buy these tools gives us some monetization. I'm not a big believer in that kind of toll booth at the exit of the data when data is egressed because there, the customer will say, well, it's my data. I've paid for the software.
So why the hell should I pay again? But when I have that logic, OK, there is value creation. And I can measure it for you, customer. I can put a price tag on how much value you get. Then I can say, you just have to click that button. And then your process will be faster. Just do it. And this is, I think, the most solid way to monetize. And this is why we are really keen to step up the investment to bring more of these use cases to life as quickly as possible.
And then let's turn it to SAP itself. I mean, we've heard other companies. There was a big announcement the other day around lowering support costs to making support much more efficient, support to big line cost item for you. We're hearing about the benefits of using Copilot in development. Obviously, R&D is a big expense line. Could you talk a little bit about what you're doing internally to try to make SAP more efficient using these tools, maybe especially around those two areas?
I mean, I give one on automation. In my area, we will need to double the quotes to customers over three years. I want to do that without increasing headcount. So I just have to bring the so-called bypass ratio up. That means the number of transactions that are not touched by human being anymore but are fully automated. On coding, of course, we are also using tools like GitHub Copilot to make our coders more productive. It's always interesting in a budgeting process to see the people who have to commit to productivity have a different view on how much productivity you get versus the vendors.
The vendors say huge productivity. Then the R&D department is haircutting. So you have to find the right balance. We are also going to train our own models on ABAP. ABAP is a proprietary language to code SAP tools.
There is some scarcity of people who do that. So what we try to achieve with that is that we develop some low-code, no-code concept where basically our customers can describe their problems in a quite generic way. And then the code is generated automatically. And yes, you will also need some coder to fine-tune. But of course, it's much less work than if that guy has to write everything himself or herself. And then, of course, we use also these process mining opportunities to improve.
And last but not least, still an emerging discipline is the whole assurance area. I think in forensics, AI can do a lot. And now that we have a process mining tool in place, we can also define control points and the controls and then start figuring out how can we test these controls in an automation fashion.
Every CFO has the challenge that the regulation is becoming huger and huger. And you being part of a bank will know that also from your sector. And that regulation is really a pain for people. And how can we create a kind of killing two birds with one stone by not only making the process more efficient but by making it better controlled and improve the assurance on the process? So these are the type of themes we're investing into to improve our own operations and then show that to customers and hopefully sell it also to the customers.
Maybe to round out that discussion around Gen AI, I guess one of the big challenges for investors is to work out how quickly this all goes. When should we start to expect top-line benefits from these tools starting to bring better pricing? How quickly can these tools impact your margins? Is this something for 2025? Or should we be more thinking a little bit longer term?
I mean, the good news is in terms of de-risking our ambition, we don't need any miracles on AI to hit our 2025 numbers. But I'd say incrementally, it's similar with our growth journey. I mean, all these emerging businesses, they start from an embryonic revenue base. And if you double every year an embryonic revenue stream, it takes a while till that revenue stream becomes meaningful for a $30 billion-plus company. But at some point in time, it inflects. And I see that inflection point really more in the outer years of our five-year planning horizon. So that's been solidly in the second half of the decade. And honestly, it kind of nicely fits with the kind of different waves of growth we have. We still have a long runway on the whole conversion of our installed base on on-prem.
Then we have layering on top the GROW, i.e., the mid-market business, the public cloud, which over time will also eat its way into the large enterprises on simpler business within large enterprises. And then AI should be kind of the next stage of the rocket to really make sure I mean, I will always think about this Cloud ERP Suite growth number. And how can I keep that as high as it is for as long as possible and then see our growth accelerate because the headwinds are easing? And I can sustain that. So it's a little bit vague, I have to confess. But honestly, it's very tough to crystal ball gaze into the kind of 2027, 2028 time frame. But I would guess that by then, we would talk about a meaningful contribution.
You know we are still in the middle of the transition of our customer base in 2027 because ECC is just running out of maintenance end of 2027.
Perfect. That's helpful. I guess CFO on stage to me, I want to dig into a little bit on costs and thought process. I mean, the biggest question I get on costs is probably around gross margins at SAP. And the concern that people have is that a lot of customers, as they move to S/4HANA, are adopting a private cloud model. And I think you mentioned one customer that had maybe a low 60% gross margin on that. And where you're aiming for something in the mid-high 70s, that makes people nervous that that's just not going to be possible. So could you just help us with how does SAP get that customer base to that kind of profitability level on cloud gross margin?
Yeah. I mean, in the near term, i.e., 2024, 2025, the extrapolation to our 2025 guidance is quite straightforward because if you kind of if you're a little bit surgical on our improvement in 2023, you see that we had an uplift of, I think it was 2.4% points. And then you strip out what was what we called the Cloud Harmonization Program, where we invested a lot of money to bring everything on an integrated infrastructure. And that is probably accounting for a good percentage point or so of that. I'd say half of that 2.4% was from there. We still have a little bit of that benefit because we still had some costs in 2024, which will fall away in 2023, which will fall away in 2024.
If you just extrapolate the growth in margin we have demonstrated in the past years, you get to that 2025 ambition. And beyond, it's really about continuing to spin the flywheel on the economies of scale. I mean, this business is growing very fast. We get better conditions with hyperscalers because of that. We are also balancing the loans better. The public cloud is growing faster than the private cloud. Of course, in absolute terms, it layers in more revenue still. But it will flip at some point in time. So there's all these opportunities. We've not fully completed these initiatives. And we do see there's an opportunity to improve the margin on private cloud. So this is where the biggest absolute lift lies that we are not going to stand still at the margins.
We are currently seeing private cloud because by virtue of economies of scale, we have demonstrated these economies of scale in the past. We continue to harvest them.
Could you maybe give a couple of, because again, this is probably the biggest question I get, pushback on margins on SAP. Just on that private cloud, there are a couple of examples of how you drive so scale with hyperscalers, I guess, is one that you're getting better terms. But are there any other kind of specific key points?
It's also traditional learning curves. I mean, there's no magic around that. It's really, really spinning the flywheel, grinding on these opportunities. I mean, by the way, I also want to highlight that for me, the biggest, the most important KPI is the cloud gross profit, not the margin itself because, of course, sometimes we have trade-offs where we say, well, am I kind of now embarking that customer? And we are arm-wrestling around terms and conditions.
And do I give a better price and some migration credits and so forth? And we run the company really to maximize the growth in the cloud gross profit. I'm not maximizing on the margin because this is ultimately how I drive the maximum earnings growth. And this is what matters to me. I think the one guidance I give, I'm volunteering, so to speak, in terms of operating leverage as a whole.
If I put it all together, one of the key objectives of our transformation program is to improve our operating leverage. The Achilles heel I sense in the business model of SAP in the last years was this lack of operating leverage. We are going to achieve two things with our transformation program. A, really do a quantum leap in terms of margin in a relatively short period of time by eliminating 8,000+ jobs while then investing a lot in the future areas. The other is investing more in IT to decouple the cost growth more strongly from the revenue growth.
While I'm not prepared to give long-term guidance, what I can say is that if I look at our cloud competition, the good companies, the good competitors we have, they're normally between 80%-90% cost growth versus the revenue growth.
Of course, I need to acknowledge that that's something we have to get to too. We have not been able to do that in the past. One of the key objectives of the transformation journey is to get there so that once we tick the box on 2025, we can continue to a, accelerate growth. The growth acceleration comes very much from the favorable mix effects I've just described before, this Cloud ERP Suite becoming more and more dominant in the mix. The other thing is that we want to grow a bit faster in post-stock-based compensation, faster than the top line.
Then what I really want to achieve for SAP is that when investors add some reasonable metric of profitability in our consolidated growth numbers, they come to decent benchmark numbers and say, this is a good company.
The sum of the growth and the margin or cash conversion, whatever you pick, is actually at a very competitive standard. And then hopefully, also, that drives future cash flow growth beyond even the EBIT growth. So that's what's behind that restructuring program. And this is why we're willing to dole out EUR 2 billion in restructuring because I think it's the best redeployment of capital to really make sure that we come to that new model, which is much more in line with industry standards.
That SAP is fit for the next three, five, seven years. We've got a couple of minutes left. I'll maybe just see if there's any questions from the floor. Can we take a question just here, please? Thank you.
Hi. Could you share your thoughts on Asia-Pacific for 2024? How are things looking for your Korea, India, Japan, and China?
Oops. I think it was quite healthy last year, 2023. I think it continues to be quite healthy. I think China is a little bit of a special [MMO] because it's largely still an on-prem market. So we are not investing much in own infrastructure there. But it has been also growing quite strongly. I'd say the biggest growth engine for us is still the United States. I mean, in our revenue mix, it's really doing the heavy lifting. But Asia-Pacific is then a close follower in terms of doing the heavy lifting on our growth.
What was surprising in 2023 was actually how strong Europe was. We had a super strong performance in Europe. And overall, I can say that the IT spend in general is actually, if anything, accelerating, is my sense. I mean, you probably have much more data and research on that.
But in the customer dialogue, I'd say the first half of 2023 was a little bit tough. But towards the end of the year, with all the, I think, the customers, they really wake up to this reality. I always take SAP as the first example. We have 69% of our cost in personnel expense. And then comes the temporary workers, which are not even in that number. Then there comes the hyperscalers, where there's personnel expense embarked in that number, which we procure.
So there's a lot of customers who have tons of personnel expense. And they're kind of quite shocked by the merit hikes they've been suffering from. And then comes AI. And I see, ooh, the same story. We try to implement. I need to decouple from that. I need to decouple from that cost growth.
I cannot afford that a huge share of my cost base is inflated by 4%, 5% 1st of January every year. I need to find a more sustainable operating mode to be competitive. This is why we think that there's more money allocated to that type of ERP theme because people see a more tangible opportunity to make themselves more productive.
Perfect. Well, I've got lots more questions. I'm sure everyone else has. But unfortunately, we're bumping up against time.
No, thank you.
I want to thank you very much again for joining us. Appreciate it.
Thanks for being here. Bye-bye.