Summary
We need to configure a month-, quarter-, and year-level headcount and attrition time-series analyses over a history in SAC, based on the SuccessFactors fields Date of Joining and Last Working Date. However, under a live OData connection from SuccessFactors, SAC treats those fields as plain dimensions—because OData maps ABAP’s date types to Edm.DateTime and SAC’s live connector doesn’t infer the SAC “Date” semantic type—which prevents SAC’s native time hierarchies (Year → Quarter → Month → Week → Day) from being available. The following post outlines the business context, the precise problem face :
1. Business Scenario
To build time-series analyses for headcount and attrition metrics leveraging live data from SAP SuccessFactors. The end users require:
Headcount snapshots of total active employees at each period end, and
Attrition rates calculated as leavers during the period divided by average headcount,
broken down by calendar year, quarter, month, and week, with drill-down capability and filters by department, region, job family, and other organizational dimensions.
This must operate entirely over the existing live OData connection—we cannot import or replicate data to any other platform (BW, HANA, Datasphere or WFA).
2. Problem Statement
Under SAC’s live OData connection to SuccessFactors, fields such as Date of Joining and Last Working Date arrive as Edm.DateTime properties in the OData service and are ingested by SAC as generic dimensions without any date semantics. As a result:
SAC’s predefined calendar hierarchies (Year → Quarter → Month → Week → Day) are not generated for these dimensions, because SAC only auto-creates them on dimensions explicitly tagged with the SAC Date semantic type in import models.
Model-level configuration is read-only in a live model, so you cannot reassign semantic types or define hierarchies for these fields within SAC.
Native drill-down functionality in charts—users clicking from Year down to Month or Week—is unavailable for these date attributes, blocking standard time-series analytics.
I’m looking for suggestions or ideas on how to tackle a specific requirement using a live OData-only SAC model. If anyone knows of any workarounds, lesser-known features, or upcoming updates that might help in this scenario, I’d really appreciate it. Thanks!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.