Although survival analysis (where statistical methods are applied to determine the time until a target event) is frequently used by clinical researchers, it is less often applied by data scientists. In this post, Brian Kent argues that survival analysis can be applied whenever decision-makers need to take action before observing all data. He further clarifies that survival analysis can help teams reason about the distribution of durations (e.g. to understand how long customer support reps take to close tickets); generate insights used to prioritize interventions by predicting the survival curve for each subject, and determine if treatments have a causal impact on the time-to-event. He concludes by providing several examples of real-world applications of survival analysis including churn prediction, employee turnover analysis, and inventory management.