SAP Bets Enterprise AI Wins on Price, Not Power

SAP's second-quarter earnings call went beyond cloud growth and updated financial guidance. Executives laid out how the company expects artificial intelligence (AI) to be deployed inside businesses over the next several years.

While much of the AI industry stays focused on larger language models and new agents, SAP executives returned repeatedly to a different priority: lowering AI costs and giving customers freedom to switch between foundation models.

Those themes appeared alongside another quarter of solid cloud performance. SAP reported cloud revenue of €6.28 billion ($7.16 billion), up 22% year over year, while current cloud backlog reached €22.9 billion ($26.1 billion), up 27%.

Total revenue increased 9% to €9.88 billion ($11.26 billion). The company lowered its full-year non-IFRS operating profit outlook to reflect the dilutive impact of its acquisitions of Dremio and Prior Labs.

SAP also completed its acquisition of Reltio during the quarter, though it expects to offset that deal's effect on operating income separately. Those investments are part of SAP's effort to build an enterprise AI platform capable of delivering reliable business outcomes across finance, supply chain, and customer operations, CEO Christian Klein says.

AI Economics Is Becoming the Competitive Battleground

SAP placed heavy emphasis on the economics of enterprise AI during the call. Customers increasingly care less about which frontier model powers an application and more about achieving a strong "price-outcome ratio," Klein says.

SAP's Business AI Platform, introduced at Sapphire in May, is designed to let enterprises choose among multiple large language models, including those from Anthropic, Google, OpenAI, Cohere, and Mistral AI, as well as open-weight models. The goal is to let SAP select the best-performing model for a given price point rather than lock customers into one provider, Klein says.

SAP does not always need the most expensive frontier models for enterprise workloads and can instead optimize inference costs while maintaining business accuracy, Klein says. SAP is applying similar cost discipline internally as it expands AI use across its own operations, CFO Dominik Asam says.

Inference spending has become a growing concern for technology leaders as organizations move AI from pilot projects into production. SAP's broader vision for enterprise AI rests on business context and governed data as the advantage.

SAP Wants to Sit Above the Foundation Models

SAP also used the call to reinforce its position as a model-agnostic enterprise AI platform, not a developer of proprietary frontier models. Enterprises face barriers beyond model performance, and Klein names governance and vendor lock-in among the concerns preventing organizations from scaling AI across mission-critical operations.

Those concerns underpin SAP's Autonomous Enterprise strategy, introduced at Sapphire in May. This year's acquisitions help explain that strategy.

Dremio expands SAP's ability to access SAP and non-SAP data without copying it. Reltio strengthens master data management across enterprise systems, and Prior Labs contributes tabular AI models designed to improve prediction accuracy across structured enterprise datasets.

SAP intends to embed those technologies inside AI agents that perform tasks such as forecasting and procurement, rather than sell them as standalone products, Klein says. Prior Labs will let agents generate accurate predictions from enterprise data without requiring customers to build complex data pipelines, he says.

Together, the acquisitions form the infrastructure behind SAP's AI platform, an evolution already visible in SAP's recent AI roadmap.

SAP Is Preparing to Sell AI Outcomes Instead of Software Seats

The call's most forward-looking discussion centered on how SAP expects enterprise AI to be commercialized. AI creates an opportunity to rethink how enterprise software is priced, Klein says.

Instead of licensing applications mainly on a per-user basis, SAP increasingly sees value in charging customers for completed business outcomes delivered by autonomous agents. AI agents could autonomously complete financial closes or generate forecasts, letting enterprises pay for measurable results rather than additional software licenses, Klein says.

He calls outcome-based pricing "a unique chance" to reset how SAP prices its solutions as AI adoption accelerates.

That approach carries into SAP's own organization. The company is restructuring engineering around AI development and rolling out Joule Work internally, executives say.

SAP is also expanding AI training with a target of reaching more than 90% of its employees in the coming months, while slowing hiring in several areas as developer productivity improves through AI-assisted software development. Development teams are already seeing average productivity gains of about 30%, Klein says.

Those internal changes build on the company's recent work around its Business AI Platform and Autonomous Suite, where enterprise workflows increasingly run through AI agents rather than traditional software interfaces. SAP's financial guidance still reflects significant investment in that transition.

The earnings call's real story was about where SAP believes enterprise AI is headed. The company is betting on lower costs and outcome-based pricing to define the next phase of enterprise AI.