SAC Predictive Planning Hands-On
As of 2026-09-25
A hands-on module on SAP Analytics Cloud Predictive Planning, sourced directly from SAP's help documentation. Distinguishes the lightweight table-cell forecast tool (Automatic Forecast, Linear Regression, Triple Exponential Smoothing; monthly-minimum granularity) from the full Predictive Planning scenario integrated with a planning model (up to 1,000 predictive models, entities from up to 5 dimensions capped at 1,000, a documented 5:1 history-to-forecast ratio, a 1,000,000-cell query limit). Covers the SAP BPC exclusion, target and aggregation-type restrictions, spreading policy, and the mandatory private-version write-back path with its governance gate before promotion to public. Closes with two planning-specific data-quality risks: adjustment periods and currency conversion.
What you will learn
- Distinguish the table-cell predictive forecast tool from a full Predictive Planning scenario and choose correctly between them for a given request
- Apply the documented entity limit (5 dimensions/attributes, 1,000 entities max) and the 5:1 history-to-forecast ratio to size a real predictive scenario
- Explain the SAP BPC exclusion, the target and aggregation-type restrictions, and why they matter to the design before configuration starts
- Diagnose and resolve a predictive scenario that fails on the 1,000,000-cell query limit, in the documented priority order
- Trace the write-back path from private version to public version and justify the review gate between them
Module overview
Who this is for. You build and run SAC planning models — versions, data actions, allocations — and you now need to add a forecast that a planner will actually trust and use. This module is hands-on: it follows SAP's own documentation for Predictive Planning, states its real limits precisely (so you do not discover them mid-delivery), and gives you two working techniques, not one. It assumes M333 (AI & LLM Fundamentals) or equivalent AI vocabulary, and basic SAC planning-model experience (versions, data entry, the account dimension).
Prerequisites
- Working experience building and running SAC planning models: versions, data entry, the account dimension
- Module M333 (AI & LLM Fundamentals) or equivalent working knowledge of forecasting vocabulary
- Access to an SAC tenant with a planning model containing at least 24 months of monthly actuals, for the exercises
Outcomes
- Run a table-cell predictive forecast on real planning data and interpret its confidence interval and retrospective fit.
- Design an entity structure and history window for a predictive scenario that respects the 1,000-entity and 5:1 ratio limits.
- Explain to a client exactly which planning-model configurations Predictive Planning does and does not support, with the documented restriction as evidence.
- Design a governed write-back path from private version to public version for a client-facing forecast.
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.