Segmentation Creation and Run

Create segmentation

To create a new segmentation, click on the Create Segmentation button. This action will navigate you to the Segmentation Creation screen, where you must fill in all the mandatory fields to create a segmentation.

Segmentation summary: When creating a new segmentation, the following mandatory fields must be filled:

• Segmentation Name: The name of the segmentation should be unique.

• What data to segment: Choose the measure, the period during which the time series data is to be considered, and the hierarchy level under this section.

  • Output Measure: Select a measure/time series for which the segmentation is to be generated. 

    • This dropdown shows the list of configured time series that includes both Units and Revenue.

    • The output measure includes both historical and forecast measures.

    • If you don’t find the time series for which want to perform the segmentation, you can reach your customer service team to enable them.

  • Start From: Select the start date that the application should use to consider the data when generating segmentation results.

  • End At: Select the date till which the application should consider the data.

  • Choose a hierarchy to segment Data: Select the hierarchy level at which the application should generate the segmentation results.

    • Firstshift allows the flexibility to generate segmentation at any hierarchy level.

Segmentation classification: The Firstshift allows the planner to tweak the classification rules for ABC and XYZ classification.

ABC Classification

The ABC classification categorizes items based on their revenue or total usage impact. It follows the Pareto Principle (80/20 rule), where a small percentage of items contribute to the majority of the value.

Default ABC Classification Thresholds:

• A-Class (80%): High-value items that contribute the most to revenue (typically the top 20% of items). These require close monitoring and accurate forecasting.

• B-Class (15%): Medium-value items that contribute moderately to revenue (typically 30% of the items). These need moderate control and forecasting.

• C-Class (5%): Low-value items with minimal impact on revenue (typically the bottom 50% of items). These require less monitoring and can be managed with bulk ordering strategies.

XYZ Classification

XYZ classification is used to analyze demand variability over time. It helps understand how stable or volatile the demand is for different items.

Default XYZ Classification Thresholds:

• X-Class (0.5 Coefficient of Variation - CV): Items with consistent and predictable demand. These can be easily forecasted with high accuracy.

• Y-Class (1 CV): Items with moderate demand variability. These require more flexible forecasting methods.

• Z-Class (>1 CV): Items with highly erratic and unpredictable demand. These are difficult to forecast and may require safety stock or alternative planning strategies.

After filling in all the mandatory fields, click the Save button to create and run the segmentation. The new segmentation will appear in the list screen with a 'running' status automatically. 

If any required fields are missing, the system will highlight them, prompting you to complete them before you can proceed.