Defining Segmentation Models Based upon Equal Sized, Randomly Selected Segments

If you go to the Segmentation Model List page, select Equal Sized Segments Randomly Selected, and click Go, the system will bring you to the Segmentation Model Definition page. This page enables you to define segments based on equally sized random selections of customers.

Using equal sized segments of randomly selected customers separates the audience into segments of relatively equal size. Customers are put into segments randomly and no customer is put into more than one segment. Note that this option is one of the few occasions that you will not have a left-overs segment. Left-overs are not necessary since all customers are assigned to random segments regardless of any filter or attribute criteria.

Do not try to use random sized segments if your audience is smaller than 100. You run the risk of the entire audience falling into the first segment and all subsequent segments being of size 0.



Define your segmentation model by:

In the lower half of the screen you will see a listing of all the segments that have been created for the chosen model. The system will automatically create two equally sized segments to start. The system will give the segments incremental names. You should rename the segments so they're easy to identify when grouped with segments from other segmentation models on reports.

From here you can: