TABLE OF CONTENTS

• Overview
• How It Works
• Worked Examples
  • Example 1: Overriding a Product Family Total
  • Example 2: Handling High-Precision Decimals
• Key Rules & Behaviors

Overview

When you update or override a demand forecast at an aggregated level (such as Product Family + Region), the platform proportionally disaggregates that total down to the lowest planning level (such as SKU + Customer).

Because proportional calculations naturally produce fractional numbers, unit-based planning requires discrete integer quantities (e.g., whole units). The platform applies an automated integer rounding redistribution process to ensure:

• All detailed records display as clean, whole numbers.
• The sum of all detailed child records always matches the parent override total exactly.
• Proportional balance is preserved, with zero units lost or gained.

Note: 
• This functionality applies only to the Planning time series. 
• It applies only if the implementer enables it during implementation.
• Do not confuse this with the measure settings in the Planner workbench, which are display settings only

How It Works

When you enter an override at a higher level, the system processes the disaggregation through the following steps:

1. Calculates proportional values: The system computes the baseline disaggregated value for each detailed record using full precision based on historical or previous proportions.
2. Splits integer and decimal portions: The system separates each calculated value into its whole number (Integer) and its fractional remainder (Decimal).
3. Sums the decimals: The system calculates the sum of all decimal portions belonging to the parent group.
4. Determines additional units: The system establishes the total additional whole units to distribute by calculating:  Additional Units = sum (Decimals)
5. Ranks detailed records: The system ranks records in descending order by decimal value (the highest decimal remainder receives Rank 1).
6. Distributes whole units: Starting at Rank 1, the system adds +1 unit to each record in order until it exhausts all additional units.
7. Finalizes values: The remaining records retain their base integer value. The resulting values sum up precisely to the parent override total.

Worked Examples

Example 1: Overriding a Product Family Total

1. Higher-Level Override

You update the forecast for Product Family PF1 in Region Europe from 1,000 to 1,219.

2. Proportional calculation and splitting

The system disaggregates 1,219 proportionally across the child SKU and Customer combinations:

3. Summing decimals and ranking

• Sum of Decimals: 0.750 + 0.156 + 0.154 + 0.940 = 2.000
• Additional Units to Allocate: 2.000 = 2 units

4. Final Disaggregated Values

The system generates the final whole-number quantities:

(Check: 305 + 151 + 446 + 317 = 1,219)

Example 2: Handling High-Precision Decimals

1. Higher-Level Override

You set the override for Product Family PF2 in Region APAC to 943.

2. Proportional Calculation and Splitting

The system computes the proportional share for each record:

3. Summing Decimals and Ranking

• Sum of Decimals: 0.985833 + 0.705833 + 0.512500 + 0.795833 = 3.000
• Additional Units to Allocate (3.000) = 3 units

The system ranks the items by descending decimal value:

4. Final Disaggregated Values

The resulting planning items display as integers and match the total:

(Check: 378 + 416 + 82 + 67 = 943)

Key Rules & Behaviors

• Independent group processing: Redistribution calculates independently for each parent combination you override.
• Ties in decimal values: If two records have identical decimal values, ties resolve consistently based on standard system sorting (e.g., alphanumeric key ordering).
• Locked records: If specific child items are locked, the system preserves the locked quantities and distributes remaining values across unlocked items.