SAP AI Readiness Checklist: A 20-Point Self-Assessment
Is your SAP system ready for Joule agents? Run this 20-point checklist across clean core, data, licensing, process, and skills before you commit budget.
20 items, 5 weighted dimensions — most SAP customers land on "not sure" for at least a third of them.
A 20-point self-assessment across the five dimensions that determine whether Joule Assistants and Agents run reliably: Clean Core & Technical Foundation, Data Quality & Governance, Infrastructure & Licensing, Process Readiness, Skills & Governance.
Each item maps to a dimension from the readiness framework: Clean Core, Data Quality, Infrastructure & Licensing, Process, Skills & Governance. Clean Core and Data Quality gaps tend to count for more, since agents reasoning over unclean code or ungoverned data fail in ways that aren't obvious until something's already gone wrong — exactly how Kernstein's SAP AI Readiness Model weighs each dimension isn't published.
Clean Core & Technical Foundation
- Current extensibility level known. Proportion of system at Level A (classic ABAP modifications), Level C (key-user extensibility), or Level D (ABAP Cloud, released APIs only).
- ABAP Test Cockpit clean-core check run in the last 12 months, with a result on file.
- RISE Methodology dashboard KPIs for custom-code debt tracked, not just checked at renewal.
- New development follows released APIs by default.
Data Quality & Governance
- Master data governance runs as a formal, owned process, with a named responsible team.
- Duplicate and inconsistent vendor/customer records are a measured number.
- Current view of how data reaches reporting and AI systems: batch extract, real-time feed, or a modern platform (e.g., SAP Business Data Cloud), mapped by area.
- Business users don't list data quality as their top complaint when asked.
Infrastructure, Licensing & AI Enablement
- Running S/4HANA, not ECC. Joule and agentic features aren't available on ECC.
- Exact Joule contract entitlement known. RISE customers typically get ~2,500 messages per Full User Equivalent per year; many Order Forms just say "available in relevant cloud subscriptions" with no figure — verified against the actual Order Form and Bill of Materials.
- BTP services relevant to planned agents are provisioned and in active use.
- At least one Joule Agent or Assistant deployed to production already.
- Current S/4HANA release and Feature Package Stack known — agent availability is often gated by FPS.
Process & Organizational Readiness
- Process to automate is standardized across regions and business units.
- Exceptions to the standard process are documented and routed to a specific owner.
- SAP modules in scope for AI identified and named.
- AI rollout has an actual, named owner as a job function.
Skills, Change & Governance Readiness
EU AI Act high-risk-system obligations reach full enforcement 2 August 2026 — applies directly to HR and financial decisioning pilots.
- Basis and ABAP teams have working knowledge of BTP and Joule Studio.
- Formal AI governance framework approved (not draft), covering data access scope and exception review.
- Business users have existing exposure to AI copilots, inside or outside SAP.
What your score means
| Score | Read |
|---|---|
| 15–20 | Reasonable shape to plan a Joule pilot with real scope — verify Infrastructure & Licensing items specifically; most-assumed, least-checked dimension. |
| 8–14 | Identifiable gaps. Most common pattern: strong Process/Skills, weak Clean Core or Data Quality — the two dimensions that tend to matter most. |
| Below 8 | Common: SAPinsider reports only ~17% of organizations have AI embedded in core workflows despite ~91% using AI at some level. Next step is diagnosing which dimension to fix first. |
Frequently Asked Questions
How many points should a "ready" SAP system score on this checklist?
No universal passing number — a raw tally isn't the full picture. A system can score well overall and still be less ready in practice than a lower-scoring one, depending on where the weak items sit. See the 5 Dimensions framework for how Clean Core and Data Quality tend to matter more than a simple point total suggests.
What's the single most commonly failed item on this checklist?
Item 10: exact Joule contract entitlement. Many RISE Order Forms describe access only as "available in relevant cloud subscriptions" without a numeric allocation.
Is a low score on this checklist a bad sign?
"Not sure" is itself a readiness signal — most SAP customers score low on at least a third of this list. Use the result to prioritize which dimension to fix first.
Key Takeaways
- 20 items across five weighted SAP AI readiness dimensions: Clean Core, Data Quality, Infrastructure & Licensing, Process, Skills & Governance.
- Clean Core and Data Quality gaps matter more — agents reasoning over bad code or data fail quietly.
- Item 10, actual contract entitlement for Joule, is the item most teams have never verified directly.
- Low scores are common: ~17% of organizations report AI embedded in core workflows despite ~91% using AI at some level.
- Use the result to prioritize.
Want a scored version of this checklist mapped to your specific environment, with a prioritized upgrade list instead of a self-graded tally? Run the free SAP AI Readiness Check and get your personalized PDF report in about 10 minutes.