SAP Eyes Outcome-Based Pricing for Joule Agents After Q2 Earnings
SAP CEO Christian Klein signalled at the Q2 2026 earnings call that Joule agents will move toward outcome/value-based pricing, potentially resetting ERP economics.
SAP Eyes Outcome-Based Pricing for Joule Agents After Q2 Earnings
SAP CEO Christian Klein used the Q2 2026 earnings call to signal that Joule agents will be monetised through outcome- and value-based pricing rather than the existing AI-Units seat model.
(See the list of Joule agents.)
The development
SAP's Q2 2026 earnings call drew attention not just for a recovery in cloud backlog, but for Christian Klein's comments on how the company intends to charge for agentic AI going forward.
According to erp.today's analysis of the call, Klein indicated that the current AI-Units model — a consumption-based seat construct introduced alongside Joule — is not the end-state for agent monetisation. Instead, SAP is pointing toward outcome- or value-based pricing, where customers pay relative to measurable business results delivered by agents rather than by usage volume or user count.
The framing erp.today uses is notable: a potential "reset" of ERP economics. The argument is that if agents execute multi-step business processes autonomously, the logical charge is tied to the outcome (an invoice cleared, a procurement cycle completed) rather than to who or what triggered it.
What it actually means
What is genuinely new: The public, CEO-level signal that outcome-based pricing is the direction is a meaningful step beyond the AI-Units model SAP launched. It suggests SAP is aware that seat- and consumption-based metrics are an awkward fit for autonomous agents that act across process steps and modules.
What remains unresolved:
- SAP has not published a pricing framework. Klein's comments are directional, not a committed model with rates, contract terms, or go-live dates.
- Defining a measurable "outcome" in ERP is harder than it sounds. Whose system of record confirms the outcome? Who arbitrates disputes? These are unsolved commercial and contractual questions.
- Most SAP customers running heterogeneous or partially-migrated landscapes will struggle to isolate agent-attributable outcomes from existing process automation, making value measurement contested.
- The shift also carries risk for customers: outcome-based models can be more expensive than consumption models if vendor and customer disagree on what constitutes a successful outcome.
Context on the cloud backlog: The headline backlog recovery is important as a business signal — it suggests demand hasn't stalled despite pricing uncertainty — but it doesn't validate that agentic features are in wide production use. Pipeline and production deployment are different metrics.
What to watch for
Watch for SAP to publish concrete outcome-based pricing pilot terms — likely first in specific agent categories such as accounts payable or procurement — within the next two to three quarters. If no framework appears by SAP TechEd 2026, the "reset" framing should be treated as aspirational positioning rather than an imminent commercial shift.
Related reading
For the full current list of Joule agents and what's actually live, see our Joule & AI Agents hub. For the environment prerequisites this assumes, see our AI Readiness hub.
Source: SAP AI pricing: outcome-based ERP economics — erp.today
Whether your own SAP environment is actually positioned to use what SAP just shipped is exactly what our free AI Readiness Check scores in about 10 minutes.