SAP Q3 Preview: 27x For An AI Turnaround That Hasn't Happened Yet

SAP shares remain about 30% below their 2025 peak after falling roughly 50% amid AI-disruption concerns and weaker cloud-backlog growth. Q2 was better than feared, with current cloud backlog up 26% and cloud revenue up 24%, but guidance depends on Middle East de-escalation, which has not materially occurred. The article identifies execution risk for SAP’s AI plan, citing its age and more than 110,000 employees, and says the company is not part of the AI build-out.

Analysis

The market is treating AI as a potential threat to enterprise software pricing and seat growth; the less obvious counterpoint is that AI agents still need governed business data, permissions, and auditable workflows. That may favor systems of record such as SAP over lightly differentiated application layers—but only if SAP can make its own AI useful without weakening its cloud migration economics. The key test is whether cloud backlog converts into revenue while customer adoption broadens, not the existence of an AI roadmap.

Near term, the Middle East assumption embedded in guidance is a material source of estimate risk: renewed escalation could defer customer decisions and weaken conversion, regardless of underlying product demand. Over 6–18 months, execution and product integration matter more than company size alone; failure to simplify implementation or show measurable customer productivity gains would strengthen the substitution and pricing-pressure case. Conversely, credible adoption could support a rerating without SAP participating directly in AI infrastructure spending.

The prior drawdown may have absorbed some disruption fear, but the rebound and better-than-feared quarter do not establish that backlog quality or guidance is durable. Without valuation, regional exposure, and backlog-conversion detail, there is not enough basis for an unconditional directional position.

Key Decisions for Investors

• Keep SAP on a catalyst watch rather than chase the rebound. Reassess after the next update on cloud backlog conversion, cloud growth, and guidance assumptions tied to Middle East conditions.
• A conditional relative-value expression is long SAP against a short position in a broad enterprise-software peer basket, sized modestly and only if SAP sustains cloud growth while demonstrating backlog conversion. The thesis is relative resilience of mission-critical systems of record, not immunity from AI disruption.
• Falsify the relative-long thesis if cloud growth or backlog conversion decelerates materially, guidance is cut, or management cannot show customer adoption of AI features translating into retention, migration, or monetization. Renewed regional escalation is a near-term downside catalyst.
• Before increasing exposure, verify geographic revenue and customer exposure, the definition and duration of current cloud backlog, conversion rates, and whether AI functionality is incremental revenue or bundled into existing contracts.