1. Define Customer Cohorts. Group the customers in the dataset into cohorts/segments based on their first-order characteristics and the month of their first order. These cohorts should be actionable, helping lead to business insights. Make hypotheses about how churn rates will differ across cohorts, which you can reflect on after performing your analysis. 2. What are the customer retention/churn rates over time for the customer groups that you defined in Part 1? Is there a significant difference in retention/churn based on different promotion values? What insights do you find from this analysis? 3. What are the dollar value retention rates over time for each customer group? What insights can you derive from your answers to Parts 2 and 3? 4. Are there any caveats to this analysis using only the provided data? What other data would be ideal to have? Please provide comments and visualizations so we can follow along with your thinking. While you won’t be penalized for approaching the analysis in multiple ways, it’s entirely sufficient to spend a few hours on just a single, well-justified approach — a few hours won’t be enough time to create a *perfect* analysis, but it should be enough to get most of the way there. Time constraints are common, and our grading rubric accounts for this.
