Clearly an outstanding quarter for what we delivered in Q3. I mean, you see, you mentioned it: the backlog is up 26%. So it also shows the resiliency of SAP, especially in times of macroeconomic volatility. Also very important that we see the impact of AI on our numbers. I mean, all the deals we closed had AI embedded. We closed AstraZeneca, where we are now developing new AI-native solutions for life sciences. We are selling a lot of AI in the supply chain area and the HR area for workforce transformation. And also, you see that we translated the great top-line results into the bottom line, because we are model agnostic, we are optimizing for price and outcome. So indeed, yeah, the strategy is working, and it resonates now also in the numbers.
Can I ask you about that — whether the top-line growth is actually converting to the bottom line? Because there have been a lot of questions about whether you can quickly migrate customers to a model where there are subscription revenues that are enduring, or whether in fact you don't have the right AI product. But in the numbers you can see that the cloud ERP suite revenue is up 29%. Is that revenue growth converting to profitability?
Absolutely, Karen. I mean, the good piece is — I guess you also have seen it in the markets over the last months, also in our stock price development. I mean, the market is realizing, first of all, that the ERP is a mission-critical system, and the agents need the ERP to have the business context, to have the governance. You cannot just build an agent and then let the agent run your business. So that's very important. The second part obviously is also very important: that now customers are really going into applying AI in their business. And yes, it's not a subscription model, it's a consumption-based business model, but they are seeing the value. And of course there is always an adoption curve, like it was also for software, and now we are ramping this up, and customers are going into autonomous financial closing, to do workforce transformation with AI, to build an autonomous supply chain. And we are seeing this impact. Obviously there is a ramp in the adoption of AI, but it already has an impact on business. Customers are seeing the value and they are on a journey. And again, it also then translates into the bottom line, because at the very beginning we were using OpenAI and Anthropic; in the meantime we have over 70 models. We are also seeing an increase of open-weight models, which is great, because the value on the one hand side is important, but you also need to of course make sure that the token spend is in check.
Christian, one of the big questions for investors has been whether you can convince those customers who have been on legacy systems to spend the money to upgrade to AI models and hosted software that you have. Is there a cost? Is there a hurdle here for some of those businesses that don't want to spend the money still because of all the volatility that we've been talking about in international markets?
No, I mean, Karen, I can tell you: you see me hopefully very energized by our numbers, but I'm even more energized, like our customers, by the opportunity we have with AI. I mean, SAP was always running the world's most mission-critical businesses, and now everyone is realizing: wow, we can really reinvent how businesses run. Last week I was in Asia and I talked to a customer who said: hey, for treasury cash collection, in the past I really waited for the next release to come. Now I'm building an agent for cash collection, and it's there the next morning, and it helps to really impact positively the cash flow of the company. And when I talk to my engineers, they are saying: wow, we are getting better and better on our own AI platform with Joule, our new UI for AI — what can we do for our customers to run their business with AI? So, I mean, this is really a once-in-a-lifetime opportunity for this company, and we are really gaining great traction.
Christian, it's Ian here. I just wonder what that means for your headcount. I mean, you've said in the recent past that you weren't hiring as many people in research and development because of AI productivity. Does that remain the case?
Absolutely, and you have also seen that we have the intent to acquire TechWolf. It's a European-based AI startup for skill management, because every customer actually is transforming their operating model and the way people work with AI. And so that was for us very important, to invest also into the ontology, to understand how can AI help you to transform your workforce. The same is happening inside SAP: the way we develop, the way we sell, the way we market, the way we run the company is all done in the meantime with Joule, with our own AI. And people are on a journey of reskilling, of training, of having an open mindset to adapt to a new reality. And we are constantly working on our workforce, but it was very important also for our own people to see: yes, we are taking them on this journey, and we are transforming the company together with them. Of course, the workforce and the skills are massively changing.
