Building the SAP AI Business Case — Value Hypothesis, AI Units and Proof Before Production
As of 2026-09-25
Teaches SAP Analytics consultants moving into SAP AI how to build a business case that survives a finance or risk review, not just a demo. Covers the falsifiable value hypothesis (baseline, target, decision, owner, kill condition); the two SAP AI cost currencies — AI Units for Premium AI inside packaged applications (SKU 8019164, activated via SAP for Me) versus GenAI tokens converted to capacity units for custom orchestration and AI Core builds (SAP Note 3437766); the Joule Base / Premium AI boundary; dated market anchors from SAP's Value of AI 2026 study (Oxford Economics, 2,600 leaders, 13 countries) including the ROI trajectory and the governance-readiness gap; pilot-vs-production sizing; and a four-line risk register (model deprecation, governance readiness, adoption vs activation, evaluation before scale). Three exercises and a self-assessment gate the move to M379.
What you will learn
- Open a business case with a single falsifiable value hypothesis (baseline metric, target metric, decision, owner) instead of a feature description
- Distinguish AI Units (Premium AI inside packaged SAP applications) from GenAI tokens/capacity units (custom orchestration and AI Core builds) and classify a use case correctly
- Explain the AI Unit activation path (SAP for Me, package then feature activation, SKU 8019164) and where actual current rates live, without inventing a price
- Distinguish Joule Base (included, no AI Units) from Premium AI (AI-Unit-metered) to avoid pricing something already licensed
- Cite SAP's Value of AI 2026 ROI and governance-readiness figures as market context, never as a client's own number
- Build a four-line risk register (model deprecation, governance readiness, adoption vs activation, evaluation before scale) that a finance or risk sponsor will accept
Module overview
Who this is for. You can build a HANA Cloud model, an orchestration pipeline or a Joule use case. This module is about the fifteen minutes before any of that: the meeting where a sponsor asks "why should we spend on this, and how will we know it worked" — and where a vague answer kills a good idea as surely as a bad one. It gives SAP Analytics consultants moving into SAP AI (M334) a repeatable method for building a business case a finance or risk stakeholder will actually approve, sized correctly in the two currencies SAP AI actually bills in. It feeds the delivery lifecycle in M379 and the AI use-case discovery workshop in M377, and it is the output a freelance consultant prices against in M380.
Prerequisites
- M334 (From SAP Analytics to SAP AI — the 90-day reskilling path) or equivalent working knowledge of Joule and the generative AI hub
- Comfort reading a simple KPI or cost model
- Optional: access to a SAP for Me account with Business AI visibility, to see the real Business AI Consumption Dashboard referenced in section 3
Outcomes
- Produce a one-page SAP AI business case with a falsifiable hypothesis, a correctly classified cost currency, pilot and production sizing, and a four-line risk register.
- Explain to a sponsor, without notes, the difference between AI Units and GenAI tokens/capacity units and why mixing them collapses an estimate.
- Use SAP's own Value of AI 2026 figures as dated market context without misrepresenting them as a client-specific result.
- Hand off a business case in the shape the Explore/Discover gate of SAP's AI Golden Path (M379) expects to receive it.
Full module available to members. The full module adds: the decision framework · the end-to-end scenario walkthrough · the KPI scorecard · the anti-patterns · the code blocks · the knowledge check · the diagrams.