- Surveys are useful for understanding WHY- quantitative data is useful for understand HOW- observing usage pattern of survivors/casualties in Week 26 after taking into account all were survivors in Week 16 will help elicit FRICTION as opposed to Motivation Gradual Stop - lost of motivation- Hard Stop - external influences
- generating of usage graph make performance of K-Clustering easier amongst groups of users that stuck around after Week 54 for reasons we know not of.