SAP's AI Math: Why Cheaper Models Alone Won't Move the Margin Needle
Published on 10/30/2026 at 07:21 | Editorial boerse-global.de
SAP bills customers for tasks completed by artificial intelligence rather than for the computing power consumed—a distinction that shifts the investor's focus from usage growth to the cost of delivering each of those tasks. As the trading week closes, that billing model offers a yardstick for judging the company's AI strategy: rising demand does not automatically translate into rising profitability.
Finance chief Dominik Asam has placed the protection of healthy margins front and center, according to Reuters. An absolute reduction in AI spending is not the goal. That nuance matters for valuation. Higher outlays would not, on their own, signal an uneconomic strategy; what counts is whether SAP can cover the added expense through the business generated by completed tasks.
To keep that equation in balance, the group draws on more than 70 models from providers including OpenAI, Anthropic and Google. Selection increasingly hinges on which model delivers the required performance at the lowest cost. Under task-based billing, cheaper technical execution can support the margin—provided the output is good enough for the use case at hand. A lower model price by itself says little about the economics of a given customer task.
Control Tools Fold Into the Product Pitch
The tools unveiled yesterday at TechEd take aim at governing AI deployment as well. New functions are designed to track agent behavior, model usage and spending, while the planned Enterprise AI Gateway would centralize policies, monitoring and control of AI traffic. SAP is thus selling not only new applications but the instruments companies need to supervise them—tying the product roadmap directly to the cost question.
Should investors sell immediately? Or is it worth buying SAP?
Provenance adds another layer. Regulators and regulated industries scrutinize where models come from, Reuters notes, and European customers more frequently request local solutions. The cheapest vendor is therefore not automatically suitable for every deployment. SAP's AI strategy cannot be reduced to a savings pledge; it has to reconcile performance, cost and customer requirements. For investors, the metric that matters is the relationship between the task business and the expense required to sustain it—not the number of available models or announced features.
A 3.5% Gain With No Clear Trigger
Separately, SAP presented new lakehouse and Business Data Fabric capabilities built on the completed Dremio acquisition. The shares rose 3.5% yesterday, though media reports identified no concrete catalyst for the move—leaving the advance unattributable to the product launch or any other single news item. The new offerings include native support for Apache Iceberg and a universal catalog based on Apache Polaris. That is tangible product progress, but no additional revenue contribution can yet be derived from it.
Cloud revenue is the yardstick to watch. In the third quarter of 2026 it climbed 26% to EUR 6.668 billion—a reading that describes the existing growth pace, not a contribution from yesterday's functions. After quarterly results roughly a week ago, SAP said it expected 2026 cloud revenue at the upper end of its existing guidance range of EUR 25.8 billion to EUR 26.2 billion. That expectation is the financial reference point; a fuller assessment would require evidence of whether product development merely sustains the current trajectory or unlocks fresh demand.
Where Data and AI Could Converge
A favorable scenario rests on the interplay of several initiatives. On Tuesday at TechEd, SAP announced new AI functions for Joule Studio, Joule Work and the SAP Business AI Platform, including tools for building and extending AI agents. Yesterday's data features complement an AI expansion already flagged. The commercial opportunity would lie in customers wanting to use both areas together, lifting the appeal of the broader portfolio—a possible path from announcement to additional cloud business, though so far only a scenario.
SAP at a turning point? This analysis reveals what investors need to know now.
Hard evidence of customer wins or statements on commercial uptake of the new functions would be the next concrete operational catalyst, carrying more weight than a further expansion of the feature set. The counterargument is equally plain: a broader product range need not lead to higher customer spending. Even a technically convincing addition might mainly support existing applications—relevant to competitiveness, without triggering a standalone growth spurt. Completing an acquisition and unveiling new functions are distinct steps, and neither substitutes for proof of additional revenue.
For now, as long as the cloud business carries its current growth direction, the Dremio features can be filed as an added opportunity within that trend. Should evidence of new customer demand accumulate, the bullish reading gains support; if the economic contribution stays limited, valuation would have to lean more heavily on business already booked. Either way, the case turns on the return the expanded offering can generate—not on whether SAP has enlarged its portfolio, which it has.
Ad
SAP Stock: New Analysis - 30 October
Fresh SAP information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
