Philipp Herzig, SAP's Chief AI Officer, revealed at the Bank of America Global Research C-Suite TMT Conference that despite industry buzz, customers show little interest in token-based payment models for AI. Instead, SAP is doubling down on a commercial strategy centered on delivering clear business outcomes and value.
- SAP divides AI features into base and premium tiers with a single AI unit SKU for premium access.
- The company uses a consumptive revenue model focused on business outcome metrics, not token usage.
- Example: Joule for Consultants decreases billable consulting hours, demonstrating measurable customer savings.
What happened
At the recent Bank of America Global Research C-Suite TMT Conference, SAP’s Chief AI Officer Philipp Herzig clarified the company’s approach to AI commercial models. He acknowledged that while tokens are often discussed in AI pricing debates, customers generally dislike token-based systems. SAP instead focuses on delivering measurable business outcomes, which drives their commercial strategies.
Herzig explained that SAP AI capabilities are split into base features, which are included at no extra cost with standard subscriptions, and premium features accessed via AI units. These AI units function under a consumptive revenue model where customers pay based on actual usage of premium capabilities across SAP’s portfolio, including supply chain, finance, and IT functions.
Why it matters
The shift away from token-centric models toward value-based commercial frameworks reflects a growing demand from enterprise customers for greater transparency and clear return on investment. Enterprises want to understand exactly what outcomes AI investments deliver, such as cost savings or efficiency gains, rather than dealing with abstract usage metrics like tokens.
SAP’s model exemplifies this by quantifying value in terms of reduced operational costs like fewer billable consulting hours or improved cash flow metrics. By sharing 70% to 80% of the created value with customers while charging 20% to 30% as a take rate, SAP aims to create win-win relationships based on trust and measurable benefits.
What to watch next
Market observers should track how SAP’s consumptive AI unit model evolves and whether more enterprise software vendors adopt similar value-driven frameworks. The transparency and simplicity of this approach could influence how AI as a service is packaged and sold across the industry.
Additionally, monitoring customer feedback and adoption rates for premium SAP AI capabilities like Joule for Consultants will reveal if these tools deliver scalable benefits that justify their costs. Success in demonstrating direct business impact may encourage broader acceptance of outcome-based AI pricing, replacing less intuitive token models.