Predictive modeling can provide clarity and precision to key studies:
- Estimation of future results based on past/current indicators
- Use regression, CHAID or other multivariate analyses
- Examines likelihood of response/conversion for campaigns
- Outcomes include sales, satisfaction, loyalty, preferences
- Web click or attribute models to link searches to purchases
- Segmentation studies predicting differences by segment
- RFM models isolate best segments to target – eg. deciles
- Product adoption models for development planning
- Measures can be used for forecasting, demand metrics