There is another financial report worth paying attention to today, which is SAP. It also has significant implications: the AI business of software companies is starting to become clear 1. The financial report growth is very good, with several sub items exceeding expectations SAP's total revenue for the third quarter was 10.27 billion euros, a year-on-year increase of 13%; Cloud revenue of 6.668 billion euros, an increase of 25% under fixed exchange rates; The current backlog of cloud orders is 24.144 billion euros, an increase of 26% under fixed exchange rates. After adjustment, operating profit increased by 18% and free cash flow increased by 36%. Compared with the consistent expectations before the financial report, the total revenue is about 1.5% higher, cloud revenue is about 1.1% higher, and operating profit is about 3% higher, which can be considered a stable answer. 2. In terms of future guidance, the official outlook for 2027 is optimistic, with fourth quarter orders being key. SAP has adjusted its outlook for full year cloud revenue and cloud and software revenue to lean towards the upper end of the original guidance range, without any increase in the range itself. Maintain the operating profit target and free cash flow target of approximately 10 billion euros. The company also maintains the judgment that the total revenue growth will accelerate in 2027. This indicates that the management has confidence in revenue, but is still leaving room for investment in products and platforms. 3. AI is starting to change the reasons why customers purchase SAP. During the conference call, the management revealed that approximately 90% of the top 50 transactions included AI and Business Data Cloud. This indicates that AI and data platforms have entered the procurement decisions of major clients, and when companies consider updating their ERP, they begin to include what can be done with AI in the future. This is also SAP's strength: its system has business objects, process rules, and transaction interfaces, on which AI can perform tasks. The official architecture also clearly integrates identity permissions, approvals, and operation tracking into the intelligent agent execution system. I personally think this is also a reason why the value of some enterprise software may be enhanced by AI (there are proprietary business and data flows that AI cannot obtain) The stronger the model, the more complex the tasks it can handle, and the more reliable the system is needed by the enterprise to implement these tasks. Of course, it cannot be said that all software and SaaS have this potential: the importance of a tool that mainly provides a simple interface and a software that manages core transactions and masters complete processes is vastly different 4. The most noteworthy change is that SAP has started to design fees around "executing work". In SAP's latest business model document, there is a straightforward sentence: commercialize the AI execution layer through AI Units. In the past, many software programs charged based on user seats. After AI improves efficiency, the number of people needed for a job decreases, which may put pressure on seat income. But the daily orders, invoices, and procurement requests that enterprises need to process may not necessarily decrease synchronously. If the fees can be linked to the actual execution volume, software companies will have an additional growth path for their revenue. l Overall, SAP deserves to be included in the key focus list of enterprise AI applications. As a representative of the "AI execution layer", it has good conditions for undertaking actual business execution and can already generate considerable cash flow from existing businesses. For the entire software sector, this financial report provides a more practical research direction: to see why companies are willing to pay and which systems are still indispensable when AI completes its work. From a personal perspective, SAP is providing an increasingly clear business path: AI helps customers increase their willingness to purchase cloud software, and then expands the scope of contracts through data platforms and business execution Previously, here was https://(x.com)/qinbank/status/2076839550119444702? S=46&t=k6rimWSEbo2D2TXolYcM-A has talked about "vertical SaaS mastering work entry points, business objects and semantics, user permissions, historical operation data, system records, final action execution, customer result feedback login elements, it can package cheap models into high-value business results. Of course, this only applies to SaaS that truly possesses workflows and core exclusive data. Thin applications that only provide a simple interface for generic models are easily replaced by model manufacturers or CSPs. ”(The screenshot below). SAP is a SaaS vendor that owns workflows and core exclusive data. Also last week here https://(x.com)/qinbufark/status/2110238858587004953? The AI mainline discussed in the last part of s=46&t=k6rimWSEbo2D2TXolYcM-A is gradually shifting. With the commercialization of the model, the deployment of AI to B and to C applications are deepening, and to C is becoming increasingly popular. In September, the strengthening of some software and the emergence of AIDD are the beginning of AI applications becoming the main trend.