ABC / XYZ segmentation is the method of grouping planning objects based on the values of a selected key figure. The two types of segmentation serve different purposes, as follows:
- ABC segmentation is the prioritization of planning objects based on their relative importance. For example, you might want to categorize the combinations of product and customer based on the revenue key figure.
- XYZ segmentation is the classification of planning objects based on their demand volatility.
Benefits of Segmentation
Segmentation helps you define more specific alerts and reports, and generate more accurate results for demand planning and inventory planning. For example, it allows you to do the following:
- Tailor your forecast strategy and inventory optimization to the characteristics of segments.
- Assign more appropriate forecast models to planning objects.
- Analyze forecast accuracy by product segments.
- Identify planning objects with a relatively high or low level of forecastability.
- Identify inventory items that require closer attention.
- Define alerts for specific segments only.
- View analytic charts with regards to specific segments.
The ABC segmentation methods are named in the Manage ABC/XYZ Segmentation Rules app.
ABC Segmentation Method Names
- By Pareto Principle (Sorted and Cumulated %)
- By Pareto Principle (Sorted and Cumulated Values)
- By Number of Items (Sorted %)
- By Number of Items (Sorted Values)
- By Segmentation Measure (Single Values)
- K-means
ABC Segmentation Method notes:
Method (5) By Segmentation Measure (Single Values) works differently than other methods as the values are not ordered and not cumulated. Instead, they are summed up for each item and the sums are compared one by one to the predefined thresholds. The segments are then calculated based on the total value produced by each item in terms of the segmentation measure.
This is useful if you do not want to compare the planning objects to each other, only to the thresholds. For example, if you want to assign every product bringing more than EUR 1,000,000 revenue to segment A, you can choose this new method for your segmentation profile.
Method (6) K-Means uses machine learning to create segments as homogenous as possible with regards to the values of the segmentation measure. This is useful if you are not sure what thresholds should be defined for the segments.
XYZ Calculation Strategies and Methods:
You can now choose between two different calculation strategies when setting the rules for XYZ segmentation: Calculate Variation and Aggregate over Periods. The main difference between them is that Calculate Variation calculates variance values during the segmentation runs, while Aggregate over Periods works with values that were previously calculated by other tools such as the Manage Forecast Error Calculations app.
If you choose the Calculate Variation strategy, you can choose Coefficient of Variation (CV) or Coefficient of Variation Squared (CV Squared) as the calculation method. CV Squared is considered more convenient than CV for evaluating demand fluctuation.
If you choose the Aggregate over Periods strategy, you can choose Minimum, Maximum, Average, or Sum as the aggregation method. This calculation strategy is useful, among others, for classifying planning objects based on a forecast error measure such as MAPE. For example, if your key figure used as a segmentation measure contains a previously calculated forecast error and you want to assign every product with a 3-month rolling forecast error of less than 20% to X, you can choose this new strategy in your segmentation profile with the aggregation method Average.
Time-Independent Key Figures As Segmentation Measures:
You can now choose time-independent key figures as segmentation measures in ABC and XYZ segmentation. This is useful if you want to use a measure that is different for each planning object, but the same for each time period. Usually, it is a single value that can be recalculated or updated regularly by another process. For example, MAPE and the inventory turnover rate are time-independent properties that can be used using key figures as segmentation measures.
Calculation Methods for ABC Segmentation:
Example Method: (5) By Segmentation Measure Single Values
The segments are calculated based on the total value produced by each item in terms of the segmentation measure, and the sums are compared one-by-one to the predefined thresholds.
Calculate Variation and Aggregate over Periods:
The main difference between the two methods is that Calculate Variation calculates variance values during the segmentation runs, while Aggregate over Periods works with values that were previously calculated by other tools such as the Manage Forecast Error Calculations app.
Thresholds:
X: 0.3
Y: 0.7