<aside> 📋

Executive Summary

A SaaS company experiencing above-average subscriber churn commissioned this analysis to identify the primary drivers of cancellation before launching an enterprise-wide retention campaign. Using SQL-based exploratory and cohort analysis against a Snowflake data warehouse, this project examined user-reported cancellation reasons, workflow completion behavior, year-over-year churn trends, and cohort-level retention patterns.

<aside> 🔑

Key Findings:

• Overall Churn Rate: 37.3%
• Top Cancelation Reason: Not Useful (primary) and Expensive (consistent across all three reason slots)
• Highest Risk Cohort: First-year subscribers churn at the highest rate - churn decreases with tenure

Business Impact Model

To illustrate the stakes of not mitigating the current churn trends, the following model has been created using the numbers defined in this data set:

Given that churn risk is highest in the first year, even a modest improvement in first-year onboarding could meaningfully shift retention. If first-year churn is reduced by 20% the compounding effect on subscriber lifetime value (LTV) is significant, increasing the importance of improving usefulness of the product for first year subscriptions.

Note: the dataset is synthetic and does not include pricing; a benchmark monthly subscription price of $100/month is used for illustrative purposes.

</aside>

</aside>

<aside> 🔍

The Business Problem

Leadership identified an above-average subscriber churn rate (37.3%) resulting in significant revenue loss. Before launching an enterprise-wide retention campaign, the team needed to understand the root causes driving cancellations. Fortunately, the company collects user-reported cancellation data through a structured exit survey — giving the analytics team direct access to subscriber-stated reasons for leaving.

</aside>

<aside> ⚙️

Approach

• Data Generation & Preparation - A synthetic dataset was generated using ChatGPT to simulate user-reported cancellation data and loaded into a Snowflake data warehouse. The dataset was designed to reflect realistic SaaS cancellation behavior across multiple subscriber cohorts and exit survey response patterns.
• Exploratory Data Analysis (EDA) SQL was used to explore dataset structure, validate data integrity, and understand descriptive statistics across the subscriber base. Techniques included multi-table joins, unions, CTEs, and window functions. This EDA surfaced the distribution of cancellation reasons across all three exit survey slots — revealing which reasons subscribers cited first, second, and third during the cancellation workflow.

• Cancelation Reason and Churn Analysis - CASE statements, views, unions, and cohort segmentation were applied to understand why subscribers cancelled, how they engaged with the cancellation workflow, and what churn rates looked like in aggregate, year-over-year, and by subscriber cohort. This analysis revealed both the current state of churn and directional trends in how cancellation behavior has shifted over time.
• Data Visualization - Findings were visualized directly in HEX Notebook using pie, column, and line charts to communicate both the current state and year-over-year trends to business stakeholders in a clear, decision-ready format.

<aside> 📈

Results

• Cancelation Reasons - 'Not Useful' was the most frequently cited primary cancellation reason, with 'Expensive' as the consistent secondary driver across all three exit survey slots. When reason ranking was removed and all responses were analyzed in aggregate, 'Expensive' and 'Went to a Competitor' tied as the top overall cancellation reasons — suggesting price sensitivity and competitive alternatives are the two most strategically significant churn drivers.

• Workflow Completion Rate - While few subscribers completed all three exit survey fields, a meaningful proportion provided two or three responses — with subscribers averaging 2.18 reasons per cancellation. This level of engagement with the exit survey is higher than typical and indicates the data is a reliable signal rather than low-quality noise.

• Year over Year Cancelation Reason - Cancellation reasons as a percentage of total cancellations increased across all categories year-over-year, while null responses declined — indicating that subscribers have become more willing to report multiple reasons over time. The most notable increases were in bad customer service and expense-related complaints, suggesting these are emerging pressure points that warrant proactive attention.

• Churn risk is highest among first-year subscribers and decreases meaningfully with each year of tenure. This pattern strongly supports an onboarding-focused intervention: retaining subscribers through their first year significantly improves long-term retention probability and subscriber lifetime value.

</aside>

<aside> 💡

Business

Recommendations & Next Steps

• The top cancellation reason across all survey slots is that subscribers do not find the product useful. This is most likely an onboarding and feature adoption issue rather than a product quality problem — subscribers who never fully engage with the product's core value proposition are the most likely to cancel. Improving first-year onboarding to drive feature adoption and demonstrate value early addresses both the 'Not Useful' driver and the cohort churn finding simultaneously.

<aside> ✅

Next Step - Review training and documentation processes for new subscribers and implement structured onboarding milestones in the first 90 days — including feature adoption checkpoints, proactive check-in touchpoints, and advanced training resources for users who have not engaged with key features.

</aside>

• For subscribers who cite 'Expensive' as a cancellation reason, introducing a targeted discount or promotional offer at the point of cancellation creates an opportunity to retain price-sensitive subscribers before they fully exit. This tactic is particularly valuable given the cohort finding: subscribers who remain beyond year one churn at significantly lower rates, meaning a short-term discount investment can yield meaningful long-term LTV recovery.

  <aside> ✅

  Next Step - Consult with the product team and finance on the practicality of implementing a cancellation rescue tactic and model what level of promotional discount generates a positive LTV return at current churn rates.

  </aside>

• 'Went to a Competitor' is the top secondary cancellation reason and tied for first overall — indicating that competitive displacement is a meaningful and growing churn driver. Without visibility into what competitors are offering, the company cannot effectively respond to this threat through product, pricing, or positioning decisions.

  <aside> ✅

  Next steps - Establish a regular competitive monitoring cadence — tracking competitor pricing, features, and positioning .

  </aside>

</aside>

</aside>

💻 View SQL on GitHub | 📊 View Power BI Dashboard | 📁 View Excel Model