This is Mark Moerdler with Bernstein. I'm incredibly excited for this conversation. I've been watching for many years now as SAP has built out capacity in both the data, the analytics, and the AI capabilities, and they've done a lot of innovative things. So I'm really looking forward to the conversation. On the Zoom, I'm at Bernstein's Tech Conference in Palo Alto, California. We have Sebastian Steinhäuser, who is a member of the Executive Board and Chief Operating Officer of SAP. And we have Irfan Khan, who is the President and Chief Product Officer for SAP Data and Analytics. Frankly, if I spent the time going through their resumes, I would take the rest of the conversation to cover it. So I'm not going to do that. I'm going to go instead into some questions. Last week's SAP Business Unleashed event had some really interesting announcements.
I don't think people fully understood the impact of what you're announcing at that event. So Sebastian, could you walk us through the major announcements, and then we'll go into the specific ramifications?
Yeah, absolutely, Mark. And first of all, thank you for having us on this call. And hello to everyone joining live and virtually. So absolutely. And I think you set it up very nicely. The SAP Business Unleashed event we had earlier in February, in our perspective, was probably the most important launch event we had, at least probably since the inception of RISE in 2020. So if you look at what we announced, for us, we unveiled, first of all, the SAP Business Data Cloud, which is also combined with a landmark partnership with Databricks, and really to open up a new era for data and AI. And we go, I think, much deeper into that. In addition, we also announced a series of ready-to-use Joule agents, our agentic capabilities, and also that we'll bring out an agent builder capability.
So maybe let me quickly tackle some of the key highlights of the Business Unleashed event, and then I think we'll go deeper. So first of all, on SAP Business Data Cloud, which was the first big announcement, which is generally available now. So basically, it's a fully managed SaaS solution that I think only SAP can offer because it really breaks down data silos and it unifies data across SAP and non-SAP, structured and unstructured data, to offer one common data layer to our customers without replicating or extracting data. And with that, I think we can really give data engineering teams full control and access to shared data, bringing all SAP and non-SAP data together in a semantically aligned data model through our data products.
Now, of course, when you look at that, this will then finally and ultimately also power much more reliable and better Business AI for our customers. As part of Business Data Cloud, we also announced our landmark partnership with Databricks, so Databricks will be an embedded part, SAP Databricks, of SAP Business Data Cloud, and I guess we'll go deeper into that as well, but it's really Databricks is a market leader in data engineering, data science, AI, and ML, and I think by joining forces, we can really offer something truly unique to our customers with their power of data engineering and AI ML capabilities, and then finally, on the Business AI side, what we announced, as I said, so we really went into our agentic capabilities. We announced the availability of several new Joule AI agents, as well as then our AI agent builder capability.
And with that, I think in simple terms, we will make Joule the super orchestrator of agents for the AI world of tomorrow. So I think that gives you a good first overview. So Business Data Cloud and agentic AI capabilities and all of that coming together in what we call the SAP Business Suite. So really our combined set of capabilities across applications, data, and AI, creating what I believe will be a virtuous circle for our customers. And I always want to say vicious, but it's a virtuous circle, creating, leveraging the power of mission-critical data that sits in our applications, feeding that into SAP Business Data Cloud in context, context-rich, enriching it with third-party data, and with that, then feeding even more powerful AI that will be a virtuous circle to create value to our customers as well as for SAP.
Thank you very much. By the way, I apologize. I didn't introduce Richard, who's my peer at Bernstein. We co-cover SAP. He'll probably interject some questions along the way. Can we drill in a little more into the Business Data Cloud? What is the differentiation between it and the prior capabilities or the other capabilities you had, Datasphere, et cetera?
Yeah, Mark, I can certainly take that question. So one welcome from my side as well to all those who are listening online. So just to put a little bit of perspective here, so SAP's data strategy is really evolving to become the Business Data Cloud. So this is not a hard stop, kind of starting from ground zero. This is an evolution. Approximately two years ago, I'll just maybe describe what Datasphere was and how that then eventually graduates now into Business Data Cloud. So approximately two years ago, we made a significant market commitment to all of our customers to say that we want to make sure that you can use SAP data in conjunction with all of the data that you own within your landscape.
So if you're a very large conglomerate, maybe you have data in a data lake, you have it maybe in on-premise, of course, across all the different deployment environments. And in order to get that data to become meaningful to you, more often than not, what customers had to embark upon was a very significant IT sort of integration project, moving the data en masse, building a very substantial operational sort of oversight. And then once you land that data, you have to then continue to maintain that data. So the response of our SAP strategy was to really embrace the notion of what's called a business fabric, a Business Data Fabric. And that has the notion of connecting to data in a virtualized way. So imagine that you've got this blank canvas and you've got all of these disparate data sources that may be out there.
And then within a pipeline that you can virtually define, it allows you to be able to attach or connect to all of these different disparate data sets that are out there with SAP, of course, and the non-SAP data sets as well. And that is a very valuable attribute to have in any data strategy.
But what we came to realize, and this is why exactly Sebastian described it with the notion of a new SaaS offering, so a software as a service offering, as opposed to Datasphere, which was more of a platform as a service offering, the distinction between the two is that in Datasphere, this is a service that SAP provides, a first-party service, but it was incumbent upon the customer to then figure out exactly how do I connect to all this data, how do I build these pipelines, and how do I then service the needs going forward. And in contrast now with the SaaS offering of Business Data Cloud, it inherits the same fabric capabilities, but we also take on more of a first-party means of being able to share the data directly from our application. So think about the ERP as a good starting point.
If you have the SAP ERP, both S/4 and, of course, even from a historical ECC or ERP perspective, you have a substantial amount of data that is there. And in most cases, those environments have been heavily customized, which in itself is a very challenging situation to always maintain to understand, well, what data do I want to expose to my analytics or to my planning or even my AI type of use cases? So with the Business Data Cloud, we take more of a managed data products journey. So all of the underlying SAP applications will now reveal the most significant business processes and the underlying data associated with those business processes.
And then through this SaaS offering of the Business Data Cloud, we will perpetually persist that data in this new SaaS offering in a fully managed manner, plus any of the extensions that you may add or subtract on the core systems. And then we make that data available to a variety of different use cases, both AI and agentic, plus, of course, analytics and planning. And then we'll get into a little bit more detail on even the insight applications that we're building on top as well. So very quickly, in summary, this is an evolution of Datasphere to become a managed service. It provides integration with SAP Databricks. It provides business f abric, plus a full set of curated, governed SAP data products, which is underlining SAP data value, and making it available to the variety of different consumers that are out there as well, Mark.
Excellent. So effectively, what you're making is SAP the center of gravity for the data that may exist throughout an organization to be able to pull it together in order to be able to analyze it or run AI against it.
Correct. In addition to all the other line of businesses, not exclusively just limited to the ERP, we have the full spectrum of the full arsenal of SAP's data portfolio rendered and available in One Domain Model within the data cloud now.
So now all that information is fully available for any of the analytics or, in theory, to be able to be used externally if the client so desires without all of that historical process of creating a data warehouse or whatever to bring the data together.
Yeah, Mark, so what tends to happen is that let's call it the data tax that unfortunately has to be paid. If you want to share it to your point now with any of the consumers, you have to maintain pipelines. You've got to pay somebody operationally to maintain those pipelines, which is a very arduous and very kind of fractious process because they often break. So therefore, you end up with analytics and reporting and dashboards that typically are not representing what the business user is looking for. So this really solves a very substantial problem for most customers where you have that governed access to data.
So I think you're hitting something important. What is the key customer from the expanded, because it's not completely new, but expanded set of capabilities that you're bringing to offer specific relating to the data cloud? And then we can talk about some of the other things.
Yeah, maybe let me start, and Irfan, please add to that. So for me, I mean, if you look at most customers, actually, I've run our own IT at SAP, and we are no different. Many customers spend up to 50% of their IT budget on data and analytics. However, a lot of that money isn't spent on creating value through data science and AI and ML use cases or analytics. It's actually spent on actually data engineering, cleansing data, bringing data together across data silos that are in the company, SAP data, non-SAP data, structured data, unstructured data, bringing all of this data together. And here, I think what really happens with SAP Business Data Cloud is that we open up the SAP data treasure for the first time, really. And as you're well aware, SAP data typically comprises the most mission-critical data a customer has.
And now what has happened in the past. I mean, in the past, and I like the analogy. I take this sheet of paper. So in the past, I mean, you take an SAP system. Let's take a BW, a Business Warehouse. You actually shape the data structure over many, many years, if not decades. So in the past, if you wanted to get that data out, you could basically only do a flat file extract. So you lost all context, all this context you built over many years. You extracted basically a flat table into a data lake, and then you had to pay someone to sit there, look at the shape of the data, and recreate exactly this business context.
Now, what we allow with SAP Business Data Cloud is now that you can actually share SAP data in context through the SAP data products and through the embedded capabilities of SAP Business Data Cloud. Also, then, into a data lake and combining with SAP Databricks, then really do data engineering, combine it with third-party data in a much more seamless way so that you can really create one common data layer, one common data foundation across the company with a lot less effort, and that's something I have to say I'm very excited about. Our own IT team, who is customer zero, is super excited about this because they were going exactly through these painstaking processes.
Now that we are building hundreds of data products across all of our lines of businesses, also for the ERP and the BW, this will take a lot of the effort out and then really provide one managed SaaS experience to work with SAP and non-SAP data. I think this is really exciting.
Irfan, anything you want to summarize?
Yep.
Oh, please.
Go ahead, Mark. Go ahead, Mark.
No, so to summarize, it's not just the data you're making available. It's all the metadata, all of the contextual information, all the security information. Everything is being made available, which takes out all that work when you want to analyze, but even more importantly, if you want to use it with generative AI in any form of automated process.
Exactly. And then all of that without actually having to move the data physically, you can actually use zero-copy sharing of that data. So that's exactly. And that's why what I said in the beginning, we are really opening up the SAP data treasure. And I think that will create a lot of customer value through Business Data Cloud, especially then combined with AI. Because AI, from my perspective, you know over time, the progress we see on the model side will level off, and it will all become about the value of the data that you can feed into your AI that will determine the business value you will get out.
Excellent. Richard, did you have a question?
Yes. In fact, I would like to ask Irfan about something because you mentioned earlier about the Datasphere customer today. And for those customers, could you share with us about their migration path to the Business Data Cloud here? What could be the time frame for them? What kind of pricing do you have in mind for those customers? And are there additional features that they need to purchase to maximize their investment in the Business Data Cloud platform?
Yeah, thank you, Richard. I mean, that's kind of like a multi-part question. So I'll try and maybe recount what you said in the three parts, right? So firstly, the migration journey for customers is that we are effectively taking the same asset, which is Datasphere, as I mentioned earlier on, and we're graduating it to become a core part of Business Data Cloud. So if you're an existing Datasphere customer today and you've spent the last couple of years building these very ornate data pipelines with all of the different semantically onboarded data, when we now go to Business Data Cloud, it's effectively a rewiring exercise. So we preserve all of your existing investment, all of your existing efforts that you've put into this, and that in itself is one simple part of the journey there.
On a commercial construct, Sebastian alluded to this. In the Business Data Cloud, which is a SaaS offering, we have this concept of a fungible credit, a BDCU, Business Data Cloud Consumption Unit, and within that, you can apportion that to be used for Datasphere, for SAC, for Databricks, SAP Databricks, or BW, so it's an exchangeable sort of capacity unit that you can use across those different assets that have just been described, and as it relates to the transition for most customers, what the exact value will be is that now we're going to that model that I described earlier, which is fully managed SAP data products. And think about the primary producer of those data products being the SAP business applications and BDC, in other words, Datasphere, becoming a consumer of those data products, a producer and consumer.
And now Datasphere becomes a first-class consumer of fully managed SAP data products. And there's a big value add there because exactly like that example that Sebastian just shared with the piece of paper, you're not losing that semantic richness of the data. And it's exactly to Mark's point, fully governed with security, with profiling, with permissions. So therefore, we have that virtuous circular trust that's also created. So within SAP landscapes and also with the consumers as well, that we maintain GDPR and all the other things that you'd expect from a governed set of access to data coming from SAP.
Perfect. Thank you.
Impressive. Where does unstructured data now fit into this? Because historically, all you could really deal with was the structured data with most of these types of solutions. You're now expanding the reach. How do you see the unstructured data coming into the mix, and where do you see the value from that?
Yeah, Mark, so I think this comes back to our choices and these are very conscious choices that we've made by partnering with, in my opinion, the best of breed data engineering provider right now out there, Databricks. Databricks has a very strong profile of managing large amounts of structured and unstructured data and the manner in which that we now embed inside of our Business Data Cloud, an SAP Databricks tenant or customer account, if you want to call it that, means that we can manage SAP and non-SAP structured and unstructured data. We're kind of taking the benefits of really the expansion of the ecosystem, these highly distributed processing or kind of platforms that are out there.
By us embracing it with their tooling means that we have a customer often will end up deciding for themselves whether they want to land it in our greenfield Databricks environment, part of the Business Data Cloud, or they want to maybe share it in that zero-copy sharing format that Sebastian also alluded to, bidirectionally. So it's really giving the customer the choice of where is it that my data resides, both structured and unstructured, SAP and non-SAP. And then do I want to do primary first processing in BDC, Business Data Cloud, or do I want to do it in the existing brownfield? But there's no disadvantage because you can mutually share the data between the two environments as well.
Yeah. And maybe to add on, maybe a few, to make it a bit exemplary on a persona-based example, imagine you're a CFO, and we showed, I think, a very nice demo at the event that showcases now the power of what's possible now combining structured and unstructured data. I mean, CFOs have used SAP solutions, SAC planning for a long time for financial steering. However, in the past, when you wanted to combine the structured financial data with non-financial data like GDP, inflation, whatever market sentiment data, consumer data that you wanted to enrich your planning with, I mean, that was lying in silos separate from your SAP data.
Now, with SAP Business Data Cloud and SAP Databricks as an embedded component, you can actually access the data product that we will provide out of, for example, here the ERP system, like general ledger information or cash flow data with now external data. So you take the SAP data product, take general ledger data, you then use that, basically access that with SAP Databricks, enrich it with third-party data, let's assume consumer data you take from social media, whatever it might be. And then you can actually use this enriched data product immediately in SAC to enrich your simulation. And of course, then using Joule's simulation capabilities, you can adjust variables on the fly and have much more powerful simulation. And that's really then this virtuous circle.
And the same is true if you're a CHRO and you want to do workforce planning and you want to enrich it, our data products of an employee master, for example, with external data from, say, LinkedIn for recruiting purposes. Or if you're a supply chain leader and you want to enrich supply chain planning data with weather forecasts or logistics data that, again, today sits in a silo, you can again always enrich our data product that we provide standard out of the box with all the context, all the metadata through SAP Databricks with third-party data and then make it again accessible for your planning application, for example, with SAC planning, which also then is a part of Business Data Cloud.
Excellent. How does the new offering fit into your recently updated strategy of AI-first, suite-first?
Yeah, maybe I start, but Irfan, please chime in here. So indeed, we rolled out internally and externally at the beginning of the year our AI-first, suite-first strategy. And it really, for me, embodies the clear priorities that we have set, especially in our development organization. AI-first is very clear for every application, for every line of business development teams. We think about how to embed SAP Business AI, how to embed Joule, how to build agents to drive value for the persona behind this specific buying center. And of course, we do that in an aligned way. We do that in one planning rhythm by now with semester planning that's aligned across all of our lines of businesses. But of course, for the full power of AI and data, you also need the suite qualities.
That's why for us, suite-first is equally important because ultimately, the customer doesn't want to run a piece of HR software or a piece of financial software. They want to run an end-to-end business process. So that's why in the same planning rhythm, we are now really focusing and doubling down based on the huge progress we've made over the last four to five years on integrating our portfolio. We are even more aligning our entire suite of applications from a UX perspective, from a workflow perspective, from a technical architecture, and now with SAP Business Data Cloud also from a data perspective to create one common data foundation, and that really then comes all together in the SAP Business Suite and creating this virtuous circle that I think will create a lot of power for our customers, bringing applications, data, and AI together like never before.
But Irfan, anything you want to add?
No, I think you summarized it well. I mean, maybe my simplest response will be AI-first and, of course, suite-first harmonized with SAP Data Cloud, which in itself is really giving that added stickiness so SAP's data can be combined with non-SAP data. And the biggest investment is in that one domain model, which in itself is a huge value add because it opens up the opportunity for net new applications and, of course, GenAI and agentic cases to be built. Trying to wrangle with data from multiple sources, SAP and non-SAP, is a non-trivial exercise. And we know customers, unfortunately, see a huge level of frustration and overhead, right, in terms of having to maintain what's essentially gone on in the past.
This is why it's an exciting SaaS opportunity because we take away a lot of the heavy lifting that typically would be accompanied with those types of projects.
Excellent. Well, you touched on something that's the buzz in the world today, and that's Agentic AI. Richard has done a lot of work looking into the whole, talking to the system integrators, IT services providers on the whole questions of implementing Agentic AI and the complexity of implementing Agentic AI. At Business Unleashed, you started talking, you launched Joule agent building capabilities. Can you explain how Joule helps drive productivity for clients? Can you touch on the Databricks and where that goes into this? And maybe adding a third to it is how do you deal with that pain point of all the complexity of trying to implement an agentic solution?
How much time do we have? Irfan, do you want to get started?
That could be the whole conversation, yes.
Yeah, we can probably split it out, so maybe I'll address the Databricks components and why that is such an incredibly strong ingredient for us to have now. In the model training, in the grounding, in the inference that you end up doing, you need a bunch of tools, right, to be able to do that. You need a knowledge graph, which we put in terms of an organic development. We have a knowledge graph from an SAP perspective. You need to be able to have RAGing facilities, right? This whole ability to be able to things like vectorization. We have that in our stack. The parts that we don't necessarily or didn't have were the very specific tooling, right, so this is the whole AI ML tooling on the one side, and on the other side, when we now look at our first-party apps and first-party data.
And why we believe that we're probably going to be the most successful at the agentic movement is because you need business context, right? Without the business context, you can't have an agent being able to communicate with another agent without that context. And in the past, the context wasn't really available. Or if it was, it was probably out of sequence, or meaning the latency was so poor that you couldn't really be using it in a real-time example. So for argument's sake, if I'm bringing up a call center and I've got a call handling problem and you want to now replace that as a physical agentic use case, you want to be able to check from accounts payable to dispatch to, of course, inventory to make sure that the effective, the response of what the client is looking for was met.
That in itself demands to have that stronger level of orchestration. Joule is our orchestrator with a natural language extension on top. We have, of course, Databricks for our advanced tooling now. We have a lot of the organic work that we've been doing. And then in terms of what the agentic framework, as it sits inside of the broader strategy, maybe you want to pick it up from there, Sebastian, in terms of what your points of view are on that.
Yeah. And look, it's exciting times, I have to say, because now, look, we have the data products. We have the data structured and unstructured. And this will now become the foundation for Joule to open up, I think, what I think will be much more powerful AI and only SAP can offer that, especially when thinking about the agentic world. But maybe before we go to the agent side, a quick recap on SAP Business AI, which is, I mean, the collection of our AI capabilities. I mean, we have delivered only last year over 130 GenAI scenarios through Joule to boost productivity, to boost growth for our customers. We have some groundbreaking innovations like Joule for consultants and Joule for developers, which will fundamentally accelerate the migration to the SAP Business Suite for our customers and to the Clean Core.
And now I think what's happening is with the emergence of agentic AI capabilities, we will open a new chapter because agents are much more than like a simple prompt to a large language model. An agent can support and mirror human collaboration, autonomously working together to combine the expertise that is in our applications, the deep process know-how to execute multi-step workflows end-to-end across applications in one singular business process. But if you think about what I said, what's incredibly important, and I feel reminded a bit of the rise of RPA when for a certain time all the talk was about how many thousand bots do you have in a company. Honestly, that doesn't matter if it's 100, 1,000, or 10,000. What matters is that you can orchestrate the agentic AI capabilities across multiple applications and process steps to ultimately achieve a business outcome.
That's where I think Joule now is coming in to be a super orchestrator of tomorrow because we know the deep semantical business context. We know how the processes work, even down to industry-specific processes across end-to-end workflows. We can now combine that with making available our very context-rich business data through SAP Business Data Cloud and with that supercharge our agents. We have now announced the first set of agents that we develop ourselves in areas like customer service. And we have announced the agent builder capabilities. And I think this is not dissimilar from what Irfan described on the SAP Business Data Cloud. For very obvious things, we build on the data side, data products. So for the things we think every customer will need, we build it. But then every customer's business is unique in its own way. Otherwise, there would be no competitive advantage.
We are very aware that every customer will also build customer-specific agents. Now, what the agent builder will enable you to do is to configure your own agent, but keeping all the context that we have knowing our applications and data structures. If you look at the Business Unleashed event demo that we did, you could also see then that with the agent builder, you will, among other things, be able to discover the SAP Business Data Cloud data products as an input feeding into the agent builder, into the agents you're configuring. Then Joule, of course, will be able to orchestrate multiple agents across and say a quote-to-cash process that you are running. I think that's super exciting. I'm very excited for what's to come in 2025.
Of course, now I can't say too much because Sapphire is already around the corner and we need to talk about something there. But it's a great time, I have to say.
How much of a buzz are you getting from clients on the agent builder? It must be very, the whole agentic must be very exciting for the client base.
I have to say, absolutely. I mean, there's a lot of interest, I mean, generally into SAP Business AI capabilities. I think the number that we've disclosed is that already in Q4, 50% of our new orders were having AI components as part of it. So there's a lot of interest. Of course, now we have to deliver on that, building our own pre-configured agents as well as then GAing the agent builder. But yes, there's a lot of interest. But I think the interest now even got much bigger knowing that with Business Data Cloud and with much even stronger doubling down on building a unified suite, this will become this virtual circle that will reinforce the value of SAP's portfolio for the customer.
Wonderful. Richard?
Yes. Thank you, Mark. Sebastian, I have a question. In fact, if I understand you correctly, Joule will play the role of the orchestrator for the SAP ecosystem, right? And then how about other agents like Salesforce, Agentforce? Because they have also the ambition to play this kind of role of orchestrator. And then your old customer, they can customize, they build their own agents. How do you see the roles of Joule here for the overall ecosystem?
If I'm so honest.
Would that be this kind of machine-to-machine talk, et cetera, to try to orchestrate everything, and then where are we in this journey?
So first of all, I think there are some unique assets that we will bring to our Agentic AI capabilities, like our knowledge graph and then the SAP data products that will be discoverable for our agents and the deep process context. So I'm not afraid of competition here. Anyhow, I think competition is always healthy. And then on how to cross-orchestrate across multiple, I'm not deaf and blind to the fact that customers have heterogeneous installed bases. I think that's something time will tell how this will play out. But I'm very confident that our Agentic AI capabilities put us in a very good position here.
Yeah. And maybe building on, Sebastian, your answer there, I mean, one of the fundamental dogmas that we've addressed here is to embrace the openness of the ecosystem. Now, every vendor historically has always tried to retain as much as it can, the platform, the data, the whatever, the processes, et cetera. And part of the Suite-first, AI-first mantra that we have is also embracing the openness of the ecosystem. And I can give you two explicit points that I will share here. One is on data products. So we are going down the path, let's call it the SAP data products economy, but it's not only SAP data. It's all customer data products that are out there.
If you're already subscribing to a particular platform or whether even if it happens to be a OneLake coming from Microsoft or it happens to be Collibra that's being used for managing your data governance, by virtue of us opening up our partnerships, as an example, with Collibra and customers that consider to sometimes be the catalog of all catalogs, we know that their data products are already resident within the Collibra catalog. Our Business Data Cloud to Collibra integration means that we can now consume the same catalog information of data products. Therefore, our orchestration and our ability from a Joule perspective to become that super agent, so to speak, means that we're actually fueling this with an openness of approach here.
Arguably, and I'm not disparaging the saying this against a Salesforce or a ServiceNow or whatever else, SAP has the widest spectrum of business applications across the multitude of different lines of business and, of course, industries as well. So arguably, if you were to looking at a vessel, how full is our vessel? I would maybe attest the fact that our vessel is very, very full in comparison. And therefore, the natural process gravity or even the agentic gravity would go to where the majority of your mission-critical applications, processes, and your data reside. So therefore, that's why the openness and the valuation now putting in there the Business Data Cloud means that we become much more of a fit for that environment.
Yeah. And I think maybe one last thought on this because I care a lot about openness. I always like to say to be AI-first, Suite-first, we also have to think partner-first. And while maybe on the agentic side, it's still early days. I mean, look at what we've done with Joule more from the GenAI capabilities. I mean, we were, I believe, one of the first to announce at last Sapphire the integration of Joule with Microsoft Teams. We've now delivered that. So I fundamentally believe in openness. Ultimately, it's all about creating value for the customer. And I'm not shy of generally collaborating here with strong partners to make sure we maximize value for the customer because that, I mean, is the ultimate goal and yardstick from my perspective.
There's a lot I could follow up on, but we can't talk all day. So I'm going to, let's talk about data gravity, okay? Historically, you arguably or likely the most important data you had in an organization sat within the SAP environment. You're now creating this connectivity, this spider web that pulls in and connects to other people's data. Can you just explain at a high level how that benefits the clients and what impact you think that will have on your partners who hopefully more of them will sign up and integrate into the data cloud?
Sebastian, I can maybe start and then finally, please, yeah, then perhaps add your comments here. So data gravity is a relative term because I think ultimately, and I've used this analogy a number of times, Sebastian's probably tired of me using this one, but I met with a CIO of Mondelez who gave me a very specific example. He said, "It's never your strategy, your data strategy, or any other vendor's data strategy. It's the customer's data strategy. And we decide where our data goes." So the gravity is dictated by the customer. And he gave me this analogy of golf. He said, "Look, think about golf. When you play golf, you try to hit the ball straight down the middle of the fairway. And wherever the ball lands, your second shot needs to be played from where the first ball landed.
It's not as if you can then move it to wherever you choose to, off the rough, into the green or whatever. You've got to play from where it lands." And the data strategy of SAP has evolved, as I said, not in that dogmatic way where we have to grab, we have to hold the gravity to our side. Of course, we believe that there is a strong value in SAP data living within an SAP ecosystem, now more importantly with SAP Databricks, giving the benefits of the tooling and, of course, the means of being able to do the unstructured management. But then with the zero-copy sharing approach that we've adopted as a principle means that we can share the data with whomever the consumer would be.
And as I also explained, with the openness of the ecosystem, of course, in the fullness of time, we'll give customer choices for them to dictate where the ball lands in that golfing analogy. And they could use whatever tooling they choose to in that zero-copy sharing format. So that's probably the best way, at least I can articulate to you that we will absolutely end up retaining more of the data gravity because when you have semantically rich data and it's available for generative AI use cases and, of course, the means of being able to drive better analytics and insights, you're going to have a value association there with a much shorter time to market and value for those use cases without any unnecessary penalization in taxes for having to maintain pipelines.
But equally so, we can share the data, right, as and when the customer chooses to do that.
Makes sense. So you're really making it the customer's decision as to where they're going to do what type of data analytics or AI or whatever they want. You're going to make your data available. At the same time, you're going to give them the best products you can in order to be able to answer those questions within the SAP world.
Yeah. I mean, I'll give you one specific example around BW, right? There's a whole host of BW installations that we have for thousands of our customers. And BW is a simple sort of case of, "I need to modernize my BW for AI and, of course, these more cross-analytical use cases." And having the means of being able to modernize BW directly in Business Data Cloud, by definition, in my opinion, will be a very strong lure to keep the data gravity because it means that you're actually having the shortest timeframe to evaluation of the data without having to incur unnecessary overhead and costs, right? So I feel as if there is an inherent value association, not being biased here, but of course, we still leave it to the customer choice with a zero-copy sharing approach to wherever they may choose to consume the data.
So it's not copied, but it's shared. So we still, inarguably, we're kind of still owning the gravity, but we can share or replicate if the customer chooses to do so.
Excellent. Change gear a little bit. I have to say I was very happily surprised when I saw the announcements with Databricks. Can you give us some more details on that partnership, how that's going to work, how clients are going to procure if they're not a Databricks client, if they are a Databricks client? I'd appreciate it. Thanks.
I think maybe you start and then I give some more details.
Should I start, Irfan, or do you want to?
Okay. All right. I'll go. So firstly, let's just put it into the two dimensions, right? So you are an existing SAP customer who currently also happens to have Databricks, right? So therefore, there's a kind of a strong intersection between the two install bases. And in order for us to cater to support both install bases, let's for argument's sake, let's call a greenfield SAP Business Data Cloud, SAP Databricks instance, and then a brownfield, which is a customer's existing environment. And the approach that we're taking, why is it that we're very passionate to work with Databricks is because we see that they are the proponents of this notion of zero-copy sharing. They are the ones who authored the notion of Delta Sharing.
This is really a testament to them that philosophically, they are on the same page as us, not moving data unnecessarily to let the data gravity reside where the customer chooses to do so and share the data. So in terms of the commercialization now, as it's a fully embedded component within SAP Business Data Cloud, so think of it as SAP Databricks in the compartment, if you want to call it, in the foundation of the Business Data Cloud, it means that we have this commercialization opportunity where we call them the Business Data Cloud Core. And in the Core, it has a set of entitlements, right, which includes SAP Databricks, includes SAC and Datasphere and BW. And the credits that we give you through the subscription, it's a subscription-based approach, you can fungibly use your credits independently of whichever service you'd want to use.
100% on Databricks, 40% on Datasphere, 20% on SAC. It's your call. Then as and when you want to share the data with your existing brownfield environment, we are doing the interoperability via data products. That's what we share. We can zero-copy share the data products. So they land in the Unity Catalog of your existing brownfield Databricks account. That means that bidirectionally, you can share the data. In fact, the exact example that Sebastian shared earlier, if you wanted to do an enrichment and you had data which arrives from SAP and you want to then do some modeling, machine learning, grounding, you can now do so in the greenfield SAP Databricks. If you want to add additional blending for other SAP and non-SAP data, you can do a Partner Connect effectively to the brownfield existing Databricks instance, and you can bidirectionally share the data.
So it's actually quite an elegant solution, and it's very harmonious for most customers because we cater for both ends of the spectrum, existing new greenfield and also brownfield and blending the two together.
Very impressive. Very impressive. And that's available today?
Yeah. So we actually announced Business Data Cloud in a controlled release. That was the announcement that we made as of February the 13th. And as of April, we'll have the general availability. We will have all the bits, all the capabilities that I'm describing with SAP, brownfield, greenfield, all converging in the April timeframe. So this is the timeframe that we're working towards.
I think what's important to call out is, and Mark, I believe you already alluded to that. I mean, and Irfan, we're not starting from zero here, right?
Right. Right.
The components, SAC, BW in the cloud, as well as Datasphere have been there. We've had a Databricks partnership already previously to Datasphere, so when we now built Business Data Cloud, we, of course, also had in mind, I mean, A, of course, the plumbing, so to bring this all together into what I believe is uniquely integrated, a SaaS application, so a managed product by SAP, but also the migration path to it to make it as easy and simple to consume for the customer and to move from whatever they are already using to Business Data Cloud in a very seamless fashion and in a very customer-centric fashion, I believe, and this is really, I mean, on the Databricks side, I believe two leaders coming together to redefine the future of how enterprises are run with data AI and applications coming together in a unique way.
Makes a lot of sense. SAP's clients, a lot of them, most of them are still on-premise, at least within the ERP environment. How do we think about how this will impact those? Is this that critical thing that's going to make people want to move even more quickly, to RISE with SAP? How do you think that will impact the transition? How will it in any way impact the clients that are not running S/4 on-premise, but maybe running prior versions?
Irfan, do you want to start or should I?
I can start, yeah. I mean, look, so Mark, you asked a very logical question, so yeah, it's great if you're in the cloud and you have all of the different moving parts already likely there, the data platform, the engineering piece, et cetera. But what if you're in the on-premise world? What happens then? This is the advantage that we have, and exactly to Sebastian's point, we have a number of the lived experiences already exactly with Datasphere where we provided a means of being able to semantically onboard even ECC, ERP 6.0, or even BW directly into the cloud with that governed manner. We call these things customer-managed data products. So just a quick sort of nomenclature here. Business Data Cloud has the currency of data products. That's the currency that you fuel Business Data Cloud with. And there are two kind of currency denominations.
One is called SAP Managed, which is SAP taking a fully governed approach. All of the core applications, ERP, HCM, et cetera, will render those data products, will persist them, will make them available to all the different use cases. But if a customer has an environment that is not in the cloud, then we use this customer-managed denomination of the same data product concept. And then they will semantically onboard whatever they want from BW. It could be modernization of BW. And we can actually accelerate their journey to the cloud.
In many ways, it will preempt them to want to actually go in a more of an aggressive way to the cloud because when they start seeing the connective tissue between modernization, new tooling, agentic AI on top of that, and the means of actually going to the on-premise world, then maybe the business case becomes even more justifiable, maybe accelerating it from that perspective. It's just, I mean, obviously, it's a thesis, right, so to speak, right? Obviously, my assumption would be, given the lived experience working with many of our customers, they're looking for the business case to justify that. That becomes even another value add to be able to justify it.
Maybe I would be a bit more direct here. From my perspective, I mean, of course, yes, SAP Business Data Cloud will also work with on-premise estates. The true value of SAP Business Data Cloud comes through the SAP Managed Data products and the true SAP Managed experience. You will only be able to benefit from that if the underlying applications, including S/4HANA, are in the cloud as well, including BW private cloud. The full power will only be available for customers where the underlying application is in the cloud or the BW is moved into the cloud as a component of SAP Business Data Cloud. I believe that will make the case even more attractive.
I mean, of course, we already have a strong momentum on RISE with SAP of our customers migrating their core ERP estates, their financial systems to the cloud. And I think Business Data Cloud will be an accelerant on that journey. Plus, will be an accelerant for the Business Warehouse, the BW modernization in the cloud as well because you're probably aware, we have tens of thousands of customers using the BW, but only very few actually migrated the BW into the cloud in conjunction with their RISE with SAP move of their core ERP estate. But now with BW as part of the Business Data Cloud, we deliver unique value to actually, I mean, because that's where typically the most, for many customers, the most mission-critical and most semantically rich data resides and where the most effort is spent on extracting and replicating data in context today.
I think that will be a big cloud migration in itself, a motion that will accelerate that has already existed, but will accelerate because now we can give you a seamless path of your BW estate to the cloud. Then also because of the fungibility of the capacity units within Business Data Cloud, you can over time modernize your BW while running it in the cloud and getting the full value of the semantical context you have in the BW already without replicating all of that logic somewhere else outside in a data lake. Look, maybe just to give you a sense because BW is a, I feel sometimes overlooked and undervalued part of our application portfolio.
Not only do we have tens of thousands of them from an investor perspective, it's also a bit of an off-balance sheet item because we in the past gave the BW as an added component away almost for free or for free. But it has a lot of value, and it typically hosts a lot of data in a similar ballpark, if you think of it in a simple terabyte manner, than the ERP system itself, sometimes a bit smaller, sometimes a bit bigger, but it's a sizable estate in itself.
Richard?
Yes. Sebastian, Irfan, I have a question here regarding what Irfan said earlier about the managed data services because that really struck me, so now with Business Data Cloud, does that mean that for the customer, do they need to purchase also additional products like SAP Analytics Cloud, BW for them to maximize the value of BDC if they don't have those products today?
Yeah, Richard.
You are opening more cross-selling opportunity here?
Yeah. Richard, what we've taken is some of the what was formerly BTP services, SAC, Datasphere, and they have been moved into a new commercial construct and a new service called Business Data Cloud. So there is an additional service that they would need to acquire. Just like Sebastian just summarized, you can fungibly use your credits, your subscription credits that you have in BDC and apportion them to using as SAC tenant if you want to do that. So in other words, there are no additional pieces that they need to purchase because we've assembled within one single SaaS offering all of the functional components that you would need, including SAP Databricks, right? And the BW entitlement is also there. So you could modernize, use the tooling, be it on SAC for a low-code or no-code insight application.
Plus, of course, you could use the SAP Databricks to be able to do the modernization at the same time as well with ML and AI type of tooling.
And maybe it would be good to maybe go a bit more into the, how does this work commercially and also from a market opportunity perspective? Because, I mean, A, this is a significant opportunity not only for us, but particularly for our customers residing, I mean, up to 50% of IT spend, enterprise IT spend is spent on data and analytics. That's a $200 billion market according to IDC for data and analytics already in 2025. I mean, compares roughly to a $380-ish billion enterprise application market. And now in that, how do we show up? With SAP Business Data Cloud, you basically have two components of the offering. One is the core BDC subscription, and the second one are LOB-specific insight apps and insight packages. Think of it HR analytics, for example. Both will be available as subscription offerings.
There's the core SAP Business Data Cloud subscription that is offered where you get access then to Datasphere, SAC, BW in the cloud if you want to move your BW to the cloud and SAP Databricks. You basically decide on the size of your subscription, and then it's a per user per month. It's not per user, but it's a monthly subscription. The second component is an insight app per LOB. That houses the SAP Managed data products applicable for the line of business, as well as then more and more sophisticated insight applications that we are building. Think of it like in the SAP Business Unleashed event, we showed, for example, the working capital dashboarding capabilities that we are building to have out-of-the-box working capital analytics in SAC. Those are the two components.
And with that, then I think we are unlocking a significant opportunity for the customer to receive value, but also a significant market opportunity for SAP. And of course, you touched on cross-sell. This, in my mind, will A, accelerate the move to cloud for customers and B, also foster cross-sell because, of course, the broader the suite of SAP applications you have, the more data products will be available for you that are semantically creating immediately without you needing to spend any money on data science. One semantic data layer across your core enterprise processes.
Excellent. We could hold this conversation for another hour easily, but unfortunately, I'm being told we're running up on the time, and in fact, I have another event I have to run here at the conference. Sebastian, Irfan, I really appreciate. Richard, thank you for helping on this, coordinating on this. Again, I really appreciate that everyone was able to make the time for the conversation. I think it's extremely valuable. Frankly, impressed with what you guys are doing in such a relatively short period of time to deliver so much value. Thank you.
Thank you.
Thank you very much.
Thank you, Mr. Irfan.
Bye-bye.