The 85 percent revenue illusion
SAP stock reached record highs this week. On October 21, 2025, CEO Christian Klein announced that 85 percent of projected 2026 revenue is already committed. The market reacted with euphoria. This reaction is premature. The committed revenue is not a sign of customer enthusiasm for new features. It is the result of a coordinated extraction strategy. Most of this backlog consists of multi year contracts signed by legacy customers desperate to avoid the 2027 maintenance cliff for ECC 6.0 systems. This is a hostage negotiation, not a growth cycle.
The numbers hide a structural decay. According to the Q3 2025 financial disclosure, the cloud backlog has hit 16.5 billion euros. However, the conversion rate from backlog to active software usage is declining. Organizations are paying for cloud subscriptions they haven’t switched on because their internal data is too fragmented to migrate. SAP is booking the revenue today for transformations that may never actually happen. This is a massive pull forward of demand that leaves very little room for organic growth once the 2027 deadline passes.
The AI credit utilization gap
The technical mechanism of the ghost credit trap
SAP sells air. They call it Business AI Units. To qualify for the best licensing tiers in the Rise with SAP program, enterprises must commit to high volumes of these credits upfront. The revenue is recognized as “booked” the moment the contract is signed. But the actual implementation of these AI features is stalled. IT departments are currently drowning in technical debt. You cannot run a predictive supply chain model on messy data from 1998. Consequently, companies are paying for AI credits they are not using. This creates a ghost revenue stream that looks healthy on a spreadsheet but provides zero productivity gains for the client.
The discrepancy is widening. Per market data from late October 2025, the gap between SAP cloud bookings and actual consumption metrics is at an all time high. If customers do not see a return on investment from these ghost credits by the time they renew in 2028, the churn will be catastrophic. The current growth is a debt that SAP must eventually pay in the form of proven value. The 14 percent utilization rate shown in the chart above highlights a growing rejection of AI features that require a “Clean Core” which most companies simply do not have.
| Metric | SAP (Oct 2025) | Oracle (Oct 2025) | Salesforce (Oct 2025) |
|---|---|---|---|
| Cloud Backlog Growth (YoY) | 28% | 19% | 12% |
| Operating Margin (Non-IFRS) | 31% | 43% | 33% |
| AI Credit Utilization Rate | 14% | 22% | 29% |
| Maintenance Cliff Exposure | High (2027) | Low | Negligible |
The hidden cost of the 2027 maintenance cliff
Fear drives the current stock price. SAP has set a hard deadline for the end of mainstream maintenance for its legacy ERP 6.0 applications. This has created a bottleneck in the consulting market. Companies are rushing to sign contracts now just to secure implementation partners for 2026. This artificial scarcity is inflating the booking numbers. It is not a sign of market health. It is a sign of a stampede toward the exit. Large enterprises have the capital to migrate, but the mid market is failing.
Smaller firms are being forced into the Grow with SAP program. This version lacks the customization they need for their specific industries. When these firms realize the cloud version of SAP is more rigid and expensive than their old on premise servers, the backlash will begin. We are seeing the early stages of this in the SEC filings of secondary software providers. These companies are picking up disgruntled legacy customers who refuse to move to the cloud and are instead opting for third party maintenance providers like Rimini Street.
Regulatory headwinds and the EU AI Act
Brussels is watching the data. The EU AI Act is now entering its critical enforcement phase. SAP is heavily exposed. Their Business AI relies on processing massive amounts of proprietary customer data across borders. The compliance costs associated with these new regulations are not yet fully reflected in the 2026 revenue projections. If SAP has to localize its AI processing to comply with strict sovereign cloud requirements, the margins on that 85 percent booked revenue will collapse. The overhead of compliance will eat the profit before the first line of code is run.
Investors are ignoring the liability. Every AI generated insight carries a risk of hallucination or bias. In an ERP system, a wrong prediction about inventory or payroll has legal consequences. SAP has not yet clarified who bears the liability when their AI makes a multi million dollar error. This legal ambiguity is a ticking time bomb for the enterprise software sector. The next critical data point arrives on January 22, 2026. That is when SAP will release its full year 2025 results and provide the first concrete guidance for the 2027 transition. Watch the contract cancellation clauses in the fine print. If the utilization rate of AI credits remains below 20 percent by mid 2026, the current valuation will be unsustainable.
