In this project, I analysed customer churn data for a fictional subscription-based company offering digital services. The dataset included user activity logs, subscription histories, support ticket frequency, and demographic details. My goal was to uncover what factors led to customer drop-off and suggest actionable insights to reduce churn.

I began with data cleaning and exploratory analysis using Python and Pandas. After segmenting users by behavior and tenure, I used logistic regression and decision tree models to predict churn likelihood. I visualized key findings with interactive dashboards in Tableau, highlighting metrics such as time-to-churn, engagement frequency, and product usage depth.

The analysis revealed that users who submitted more than two support tickets in their first 30 days were 3x more likely to churn. Based on this insight, I recommended onboarding enhancements and a new feedback loop system. The final report outlined strategic actions for customer success teams to intervene earlier with at-risk users.

Skills Gained:

• Data cleaning & preprocessing

• Predictive modeling (Logistic Regression, Decision Trees)

• Dashboarding (Tableau)

• Business storytelling with data

Note - Photo credits go to RonDesignLab. The use of cover photo is for template purposes only.