Segmentation apps
An advanced multivariate segmentation suite, reached from the Segmentations Module in the left bar (see Explore - Top Bar and Left Bar).
The four apps
App | Purpose (one line) | Note |
|---|---|---|
Cluster Analysis | Splits respondents into mutually exclusive, like-minded segments | Explore - Cluster Analysis |
Correspondence Analysis | Maps multi-variable categorical relationships on a 2D perceptual grid | Explore - Correspondence Analysis |
Factor Analysis | Reduces large batteries of correlated questions into latent dimensions | Explore - Factor Analysis |
Audience Effects | Finds high-potential target audiences using reach / index screening plus CHAID | Explore - Audience Effects |
Common behavior
- Segmentation active notification: when a saved report with active Segmentation sessions is opened, Explore alerts the user that statistical segmentation models are attached to the file.
- Data discrepancy alerts: segmentation calculations use fixed iterations. If rows, columns or tablebases in the Crosstab are changed while a Segmentation session is active, a warning banner says the crosstab data no longer matches the segmentation inputs.
- Export: formats are PPT, Google Slides, PDF, JPEG, PNG and SVG.
How the apps connect
Links between the apps, as stated in the source:
- Correspondence Analysis is used before Cluster Analysis to identify which psychographic variables are most strongly associated with audience segments and products (sorting by % INF).
- Factor Analysis is frequently used before Cluster Analysis, so clustering runs on 5 to 10 factor scores instead of dozens of raw variables.
- Factor scores can be saved as custom audiences and used as rows, columns or clustering inputs.
- Audience Effects segments can be moved into planning tools such as the optional Audience Planner, or passed to media and channel planning tools and activation partners.