Backlog Growth Is Less About Demand, More About Deadlines
SAP’s current cloud backlog of €22.9 billion is the contracted value of future cloud services that customers have already committed to buy, driven largely by forced SAP cloud migration from aging ECC systems ahead of support cutoffs and intertwined with new AI offerings that will change how enterprises pay for software. SAP’s latest quarter looked strong on paper: total revenue reached €9.878 billion, up 9% year on year, with cloud revenue at €6.281 billion, up 22%. But the more telling number is that backlog jump of 27%, which shows enterprises signing deals now to avoid running out of runway later. The reason is blunt. Mainstream maintenance for SAP Business Suite 7 core applications ends in December 2027 for most ECC customers. Typical migrations take 18 to 36 months, and many estates are more complex. If you are still debating your move in mid-2026, you are not being strategic; you are racing the ECC cutoff deadline with a shrinking margin for error. That pressure shows up in the backlog more than in the glossy revenue slide. It confirms that waiting is no longer neutral for ECC shops. Every quarter of delay means more compressed testing, more rushed custom-code clean-up, and less time to rationalize process change before the migration project hits finance and operations.
Mandatory SAP Cloud Migration Collides with AI Ambitions
The backlog surge is not some organic rush toward innovation; it is a forced march. Only about 39% of ECC customers had licensed S/4HANA by the end of 2024, leaving more than 60% of the installed base without the platform they need for the next phase of SAP cloud migration. Meanwhile, on-premise license revenue fell 32%, and support revenue declined 8%, underlining SAP’s structural push to make cloud ERP the center of its business. Christian Klein framed the quarter around an Autonomous Enterprise strategy powered by the Autonomous Suite and Business AI Platform, arguing that customers are choosing SAP for “accurate and compliant AI outcomes” anchored in core business processes and data. The hard truth is that those AI promises are now inseparable from the migration story. Klein has said that with current data quality and ERP complexity, “AI is going nowhere” unless customers modernize their core environments. In practice, enterprises face dual pressure: move off ECC before support erodes, and modernize data so that AI agents can function. That strips away any illusion that AI adoption is optional. For SAP customers, it is increasingly bundled into the cost and risk profile of the migration itself.
Enterprise AI Pricing: From Seats and Modules to Outcomes and Agents
The earnings call made clear that the next battle will not be about whether AI works, but how enterprise AI pricing will be set. In the Q&A, Klein described AI as an opportunity to move away from traditional ERP pricing logic toward value-based and outcome-based pricing tied to autonomous agents. He emphasized that SAP does not want to monetize the model alone, but “the value of our agents,” and said this could “completely reset the price level.” That is a sharp break from subscription economics built around user seats, modules, and predictable maintenance. SAP expects to launch Business AI Platform and Joule Work in Q3, with close to 50 assistants by then and more than 400 Autonomous Suite agents by the end of the year. These agents will do tasks that used to belong to teams and individual users—closing the books, reconciling orders, flagging compliance issues. For customers, autonomous agents cost is both an opportunity and a hazard. If an agent shortens financial close or cuts manual reconciliation, outcome-based pricing might be easier to defend than another seat license. But it also introduces the risk of opaque metrics, unexpected overages, and a gradual escalation of AI service tiers far beyond today’s subscription baselines.
AI Economics Are Already Reshaping SAP’s Own Financials
SAP’s financials already show the strain of building an AI-first model while maintaining cloud growth. IFRS operating profit rose 8% to €2.64 billion, and non-IFRS operating profit rose 7% to €2.74 billion, slower than the expansion in cloud revenue. SAP reset its 2026 non-IFRS operating profit outlook to €11.8 billion to €12.2 billion at constant currencies, explicitly reflecting dilution of more than €100 million from recent acquisitions Dremio and Prior Labs. These acquisitions are not side bets. Dremio brings an open data lakehouse platform aimed at analytical and AI workloads across SAP and non-SAP data, while Prior Labs adds tabular foundation model expertise tied to SAP-RPT-1 and business-data intelligence. The near-term effect is margin pressure; the strategic goal is a Business Data Cloud and Business AI Platform that can support agentic workflows at scale. Customers cannot ignore this cost story. SAP is funding an AI transition while still protecting the cloud growth narrative, which means it must keep pushing migration, data modernization, and AI adoption together because it needs a cleaner cloud base for agents to work. The more the company invests in agents, the more it will need to recoup that spend through new pricing structures tied to AI value rather than old maintenance logic.
What This Signals for Enterprise Software Economics
Taken together, SAP’s Q2 numbers and its AI pitch mark a turning point in enterprise software economics. Cloud revenue rose 22% to €6.28 billion, and cloud ERP Suite revenue climbed 25% to €5.53 billion. Current cloud backlog grew faster than cloud revenue for the first time in several quarters, giving a clean signal that contracted cloud demand is outpacing in-period billing. But behind that signal, the shift is stark: cloud subscriptions are becoming the baseline, not the profit engine. The new engine is AI, priced around outcomes and autonomous agents. That combination—deadline-driven migration plus experimental AI monetization—changes the software purchasing calculus. Buyers must now assess not only whether S/4HANA and RISE contracts meet ECC cutoff pressures, but how Business AI Platform and Joule Work may lock them into long-term, value-based charging they cannot easily benchmark. The practical takeaway is plain. Enterprises should treat SAP’s backlog story as a warning, not a comfort. The rush to meet the ECC deadline is locking in multi-year commitments at the same moment SAP is trying to “completely reset the price level” via enterprise AI pricing. Those who sign under time pressure without a clear view of autonomous agents cost and AI tiering may find that the real bill for this migration arrives years after the project goes live.
