The Cloud Backlog Is the Number That Matters
SAP reported Q2 2026 cloud revenue of EUR 6,281 million, up 22% and 24% at constant currencies. The figure that carries more signal for enterprise buyers sits one line down. Current cloud backlog reached EUR 22,929 million, up 27% and 26% at constant currencies. Backlog is contracted future revenue, so it reads as a leading indicator of demand that has already cleared procurement and legal. When a company that sells the systems of record for most of the Fortune 500 grows committed future cloud spend at that pace, it tells you how enterprises are actually allocating budget rather than how they are talking about it in board decks.
That distinction matters right now because the market has spent a year debating whether agentic AI would gut enterprise-software budgets. The theory ran that autonomous agents would replace seat-based SaaS and compress the value of application vendors. A 26% jump in constant-currency backlog is a hard data point against that thesis, at least for the core ERP layer. Customers are not pausing multi-year cloud commitments while they wait to see how agents shake out. They are signing, and they are signing at scale. For technology leaders modeling their own spend, that resilience in the demand signal is worth more than any single quarter's revenue print.
Cloud ERP Suite Is Doing the Heavy Lifting
Underneath the headline, Cloud ERP Suite revenue grew 25% to EUR 5,525 million, up 27% at constant currencies. That line is now the clear center of gravity in SAP's model. It captures the migration of legacy on-premise ERP estates onto the cloud platform, and its growth rate outpacing total cloud revenue tells you the mix is still shifting toward the strategic, stickiest workloads. For CIOs, this is the part of the SAP story that maps directly onto their own transformation programs. The S/4HANA cloud move has been a multi-year slog for many shops, and these numbers suggest the cohort finally clearing that migration is large and still expanding.
The counterweight sits in software license revenue, which fell about 32% at constant currencies to roughly EUR 131 million. That decline is a mechanical feature of the transition. Upfront perpetual licenses are giving way to subscription contracts, so revenue that once landed in a single quarter now spreads across the life of a cloud deal. Total revenue still grew 9% to EUR 9,878 million, which shows the subscription base is more than absorbing the license runoff. Leaders evaluating their own vendor mix should read the license collapse as the mechanical cost of a model that trades lumpy deals for predictable recurring revenue.
The Profit Cut Is an Acquisition Story
SAP lowered its 2026 non-IFRS operating profit guidance to EUR 11.8-12.2 billion, down from EUR 11.9-12.3 billion. On its own a guidance trim would spook the market, so the reason behind it deserves scrutiny. The company attributed the cut to a dilutive impact in excess of EUR 100 million from the Dremio and Prior Labs acquisitions, both of which closed in July. Notably, 2026 cloud revenue guidance stayed unchanged at EUR 25.8-26.2 billion. The demand outlook held firm while the profit outlook softened, which locates the entire adjustment on the cost side and specifically on deal integration rather than any weakness in the underlying business.
That framing changed how the number landed. A profit cut caused by fresh acquisition costs reads very differently from one caused by slowing sales or margin pressure in the base. The quarter itself was healthy: non-IFRS operating profit of EUR 2,743 million was up 7%, non-IFRS basic EPS came in at EUR 1.59, and free cash flow of EUR 3,002 million rose 27%. Cash generation growing far faster than reported profit is exactly the pattern you expect from a maturing subscription book. The dilution is a deliberate investment charge, and SAP is signaling it will keep spending to expand its capabilities.
Dremio and Prior Labs Point at a Buy Strategy
The two deals driving the dilution are worth understanding because they map SAP's intent. Dremio is a data-lakehouse and query engine that sits close to the analytical data layer AI workloads depend on. Prior Labs sits on the AI research and modeling side. Together they show SAP reaching outward for the pieces of the AI stack it wants to own rather than assembling them internally on a slower timeline. Spending more than EUR 100 million of near-term profit to close both in a single month is a statement about urgency. SAP would rather take the dilution now and hold the capability than wait to build equivalent technology in house.
For CIOs, the pattern is the useful takeaway. Expect the large application vendors to keep buying their way into the AI stack, folding data engines, model tooling, and agent frameworks into platforms customers already run. That has direct consequences for your build-versus-buy calculus. Capabilities you might have scoped as internal projects can arrive as native platform features through acquisition, often faster than an in-house team could ship them. The risk sits in integration debt and roadmap dependency, since acquired products take time to knit into a coherent platform. Track what your core vendors acquire as closely as what they announce on stage.
What the Executives Are Actually Signaling
CEO Christian Klein tied the demand back to trust in outcomes: "Customers are choosing SAP to enable accurate and compliant AI outcomes grounded in their critical business processes and data." Strip away the polish and the claim is about grounding. SAP is betting that enterprises want AI anchored in the governed, auditable data already sitting inside their ERP rather than in free-floating models operating outside the system of record. For regulated industries and finance-adjacent processes, that grounding argument is a real differentiator. It positions the ERP layer as the compliant substrate for agentic workloads, which is precisely where the Dremio data engine acquisition starts to make strategic sense.
CFO Dominik Asam framed the internal side: "We aggressively drive our own transformation into an Autonomous Enterprise, leveraging AI to boost effectiveness and efficiency." That is SAP applying its own AI thesis to its own operations, and it doubles as a proof point for the sales motion. A vendor telling customers to run their business on AI-infused processes carries more weight when it is restructuring its own cost base the same way. For technology leaders, the pairing of these two quotes is the strategy in miniature: sell grounded, compliant AI to customers while using AI internally to defend margins against the very acquisition dilution the guidance cut reflects.
The Read for Your 2026 Roadmap
Markets endorsed the quarter. Shares rose about 9% on Friday July 24 as investors treated the cloud numbers and the unchanged revenue guidance as evidence that the core business remains resilient. That reaction confirms the profit cut was read as an investment choice rather than a demand problem. For technology leaders, the equity move is less important than the underlying signal: the enterprise-software demand curve for cloud ERP is holding through the loudest phase of the agentic AI debate. Budget pressure is real, but it has not translated into customers walking away from long-dated platform commitments. Plan your own vendor relationships around continuity, not disruption, in this layer.
The practical implication is to treat vendor acquisitions as a live input to your architecture decisions. When your core platform vendor buys a data engine or a modeling shop, capabilities you were about to build may soon ship natively, which can strand in-progress internal work. Build a habit of mapping each major acquisition to your own backlog and asking whether the platform will now cover it. The counterbalance is discipline on lock-in, since a vendor buying the whole AI stack deepens your dependency on one roadmap. Own the grounding data, keep integration points open, and let the vendors absorb the cost of assembling the stack while you stay free to move.
