Discover how new entity selection capabilities make working with hierarchies and large member sets faster and more intuitive in SAP Analytics Cloud's predictive planning.
Streamlining Entity Selection for Predictive Planning
When configuring predictive planning models in SAP Analytics Cloud, entity selection defines which business elements you want to forecast. For many organizations, this means working with hierarchical dimensions like stores organized by geography, products arranged by category, or customers grouped by region.
As forecasting has evolved to support increasingly complex organizational structures and larger datasets, we've enhanced the entity selection capabilities to make working with hierarchies and large member sets more efficient and intuitive.
New capabilities now available: Starting with wave 2026.12 (Q3 2026 release), SAP Analytics Cloud's predictive planning includes enhanced entity selection with hierarchical level selection and include/exclude logic, making it easier to precisely define which entities you want to forecast.
What's New: Three Powerful Selection Enhancements
The entity selection dialog now includes three key enhancements that streamline how you work with hierarchies and member sets:
1. Members by Level
Select all members at a specific level in your hierarchy with a single action.
Example: In a geography hierarchy (Country → Region → Department → City → Store), you can select “all members at Store level” to automatically include all stores, regardless of which country or region they belong to.
Fig 1. Members by Level selection capability is available as a contextual option on the Members by Level item.
2. Members Below by Level
Select all descendants below a chosen member, down to a specific level.
Example with level-based hierarchy: Starting from “Paris” (a city), select “all members below down to Store level” to include all stores under Paris.
Example with parent-child hierarchy: Starting from “France” (a country), select “all members 2 levels below” to include all entities two levels down in the hierarchy.
Fig 2. Members Below by Level selection capability is available as a contextual option on all the members.
3. Include and Exclude Logic
Build your entity selection using both inclusion and exclusion rules, making it easy to specify “all except these few.”
Example: Include all 1,000 stores, then exclude the 3 Paris stores that are undergoing renovation—no need to individually select 997 stores.
Dynamic Selection: Adapting to Your Evolving Data
An important characteristic of these new selection capabilities is that they create dynamic selections—the precise set of entities included is evaluated at runtime when you train or apply your predictive model.
How Dynamic Selection Works
When you use level-based selection or hierarchical patterns, your selection rule is stored, not the specific list of members at that moment:
- Selection rule: “All members at Store level” or “All members below Paris down to Store level”
- Evaluation time: When you train or retrain your model, the system evaluates this rule against your current data
- Automatic adaptation: If your hierarchy has changed—new stores opened, stores closed, organizational structure adjusted—your entity set automatically reflects the current state
Example: Dynamic Selection in Action
Scenario: You configure a forecast with “all members below France down to Store level”.
Initial training (January): France has 150 stores. Your model trains with 150 entities.
Retraining (July): France has expanded to 155 stores (5 new openings). Your model automatically retrains with all 155 stores—no configuration update needed.
Benefit: Your forecasting configuration stays current with your organizational reality without manual maintenance.
Dynamic vs. Static Selection: Choosing the Right Approach
SAP Analytics Cloud supports both dynamic and static entity selection, each suited to different scenarios:
Dynamic Selection
What it is: Using Members by Level, Members Below by Level, or hierarchical node selections. The entity set is determined at runtime based on your current data structure.
Best for:
- Forecasting scenarios where your entity set evolves regularly (stores opening/closing, products launching/retiring)
- Hierarchical selections that should adapt to organizational changes
- Configurations that need to stay current without manual updates
- Large entity sets where maintaining explicit member lists is impractical
Example: “All stores at Store level except those in Paris” automatically adjusts as stores are added or removed from your hierarchy.
Static Selection
What it is: Explicitly selecting individual members one by one. The entity set is fixed to the specific members you selected at configuration time.
Best for:
- Forecasting a specific, fixed set of entities that shouldn't change even if the hierarchy evolves
- Historical analysis where you want to maintain consistency with past forecasts
- Scenarios where you need precise control over exactly which members are included
- Small entity sets where explicit selection is manageable
Example: Selecting exactly 10 specific pilot stores by name—even if new stores are added to the hierarchy later, your forecast will continue to use only those original 10 stores.
Real-World Scenario: Retail Store Forecasting
Let's explore how these enhancements work together in a practical forecasting scenario.
The Business Context
A retail company operates 1,000 stores across multiple countries. The stores are organized in a hierarchical geography dimension:
- Country
- Region
- Department
- City
- Store
The forecasting team needs store-level forecasts for inventory planning and staff scheduling. However, three stores in Paris are undergoing major renovations and should be excluded from the forecast model since their patterns during construction won't be relevant for future planning.
Previous Approach
Creating this entity selection previously required:
- Expanding the geography hierarchy level by level
- Navigating through countries, regions, departments, and cities
- Manually selecting each of the 997 stores to be included
- Verifying that the correct stores were selected
While functional, this approach required significant time when working with large member counts, especially when dealing with deep hierarchical structures.
Enhanced Approach with New Selection Capabilities
With the new selection enhancements, the same configuration becomes remarkably straightforward:
Step 1: Select All Stores by Level
Use the Members by Level option to select all members at Store level. This immediately includes all 1,000 stores with a single action.
Fig 3. Select all the stores (dynamic selection).
Fig 4. Selection update
Step 2: Exclude the Paris Stores
Navigate to the Paris city node in the hierarchy (you can browse the tree or use the filter to find it quickly). Once Paris is selected, switch to exclude mode and click Members Below by Level > This Member to exclude all stores located in that city.
Fig 5. Exclude stores in Paris
Fig 6. Selection update
Step 3: Configuration Complete
Your entity selection now shows “All stores at Store level except Paris stores”—exactly the 997 stores needed for the forecast. The selection is clear, maintainable, and easy to modify if renovation plans change.
Fig 7. Member selection in the model settings
Key Benefits of Enhanced Entity Selection
These selection enhancements deliver meaningful improvements to your forecasting workflow:
Value of Enhanced Entity Selection
- Faster configuration: Select large member sets with a few clicks instead of extensive manual navigation
- Clearer intent: Selection rules that express business logic (“all stores except renovations”) are more maintainable than long member lists
- Reduced errors: Level-based and exclusion-based selection minimizes the risk of accidentally missing or including the wrong members
- Easier maintenance: When entity sets change, updating selection rules is more efficient than revising manual selections
- Hierarchical efficiency: Work naturally with hierarchical structures using level-based selection instead of expanding every branch
- Flexibility: Combine inclusion and exclusion logic to handle complex business scenarios
Backward Compatibility
Enhanced entity selection is available now in SAP Analytics Cloud's predictive planning. Entity selections created before these enhancements operate in static mode (explicit member selection), maintaining the specific list of members originally selected. When you edit existing models, you can keep the static selection or migrate to the new dynamic selection capabilities.
Conclusion
Enhanced entity selection with hierarchical level selection and include/exclude logic makes configuring predictive planning models faster and more intuitive, especially when working with large member sets and complex hierarchies. These dynamic selections automatically adapt as your data evolves, keeping your forecasts aligned with your current business structure without manual maintenance.
These capabilities are available now in SAP Analytics Cloud's predictive planning, starting with wave 2026.12 (Q3 2026 release).
Do you want to learn more on predictive planning?
