Configuring Key Figures, Aggregation Modes, and Time Profiles in Planning Areas

Explanation

A planning area is only as useful as the key figures and time behavior configured inside it. Once the master data structure (attributes, planning levels) is established, the next design layer is key figures: the numeric fields that hold quantities, values, or calculated results such as forecast, sales history, supply plan, or inventory targets. Each key figure in a planning area is defined with a data type (usually quantity or value), a unit of measure or currency reference, and critically, an aggregation mode and disaggregation mode. Aggregation determines how values roll up from a detailed planning level (for example, product-location-week) to a higher level (for example, product group-region-month) when a planner views summarized data. Common aggregation modes include SUM for additive measures like sales quantity, and AVG or LAST VALUE for measures like stock levels or percentages where summing would produce meaningless results. Disaggregation is the reverse: when a planner enters or adjusts a value at an aggregated level, the system must spread that change back down to the detailed level using a defined method, commonly proportional to historical values, existing plan values, or equal distribution. Choosing the wrong aggregation/disaggregation combination is a frequent design mistake that causes planners to see distorted numbers when navigating between views. Time profiles are the second critical dimension: each planning area is associated with a time profile that defines the granularity (day, week, month) and the horizon (past and future periods) over which key figures are stored and calculated. Time profiles are shared or planning-area-specific depending on configuration, and they determine storage volume and calculation performance—finer granularity (daily) over long horizons multiplies the number of time-series data points dramatically, impacting both database size and calculation runtime for operators like forecasting or supply planning heuristics. Many implementations use a coarser granularity (weekly or monthly) for statistical forecasting and demand review, while using daily granularity only where operationally necessary, such as short-term supply or response planning. Key figures can also be defined as stored (persisted in the time series database) or calculated (derived on the fly via formulas referencing other key figures). Calculated key figures reduce storage but increase calculation load at runtime, so the trade-off must be evaluated based on how frequently the value is queried versus how often its inputs change. Global time series (used across multiple planning areas) versus planning-area-specific key figures is another design decision affecting reusability and governance. In production, changes to key figure aggregation modes or time profiles after go-live can require significant data reload or historical data realignment, so these decisions should be validated early with business stakeholders and tested against realistic data volumes before being locked into the productive planning area. Understanding these mechanics is essential for consultants who configure planning views, diagnose why summarized numbers look wrong, or tune performance for large planning areas.

Code example

-- Illustrative key figure configuration table (not an actual SAP object) -- Key Figure: SALESHISTORY  Data Type: Quantity  Aggregation Mode: SUM  Disaggregation Mode: Proportional to History  Time Profile: Weekly (Planning Time Profile: PTP_STD) Key Figure: TARGETSTOCK  Data Type: Quantity  Aggregation Mode: LAST VALUE (non-additive; stock snapshot, not summed)  Disaggregation Mode: Equal Distribution  Time Profile: Weekly Key Figure: FORECASTERROR (Calculated)  Formula: ABS(STATFORECAST - SALESHISTORY)  Stored: No (calculated on demand)  Aggregation Mode: AVG -- Design note: SALESHISTORY uses SUM because weekly quantities can be added-- across periods or locations meaningfully. TARGETSTOCK uses LAST VALUE-- because summing stock snapshots across weeks would be meaningless.

Real project scenario

During a regional S&OP rollout, planners reported that when they switched a planning view from weekly to monthly, total inventory figures appeared inflated by a factor close to four. Investigation traced the issue to the TARGETSTOCK key figure being configured with SUM aggregation instead of LAST VALUE, causing the system to add stock snapshots across the weeks within each month rather than showing the ending balance. The consulting team had to correct the aggregation mode, validate against a parallel Excel reconciliation, and communicate to business users that historical reports generated before the fix needed to be regenerated, since the underlying calculation—not just the display—had been wrong.

Common mistakes

• Using SUM aggregation for non-additive key figures like inventory levels, percentages, or ratios, producing misleading rolled-up numbers. • Choosing a disaggregation method that does not match planner expectations, causing confusing top-down adjustments. • Setting daily time granularity across the full planning horizon without evaluating storage and calculation performance impact. • Treating calculated key figures as free, when heavy formulas evaluated frequently in large views can slow down planner navigation noticeably. • Changing time profile granularity after go-live without a clear plan for realigning or reloading historical time series data.

Best practices

• Classify each key figure early as additive, non-additive, or calculated, and choose aggregation mode accordingly before configuration. • Align disaggregation logic with how planners actually work in the tool to avoid confusing top-down edits. • Use the coarsest time granularity that satisfies business need; reserve daily granularity for short-horizon operational key figures only. • Prefer calculated key figures for simple, infrequently queried derivations; persist (store) key figures that are heavily used in dashboards or reports. • Test aggregation and disaggregation behavior with realistic sample data and multiple planning levels before locking configuration for productive use. • Document time profile and key figure decisions so future extensions do not silently break existing rollup logic.

Interview angle

Interviewers assess whether a candidate understands that aggregation is not automatic or generic—it is a per-key-figure design decision with real consequences. Strong answers explain the difference between additive and non-additive measures, describe disaggregation trade-offs, and can articulate why time profile granularity is a performance and storage decision, not just a reporting preference. Candidates who can describe a real troubleshooting scenario involving incorrect rollups stand out.