Monthly donors are a critical source of sustainable funding for Plan Canada. However, predicting when a donor might cancel their contributions — and preventing that churn — was a major challenge.

Key issues included:

• No early warning system to identify at-risk donors
• Limited ability to proactively intervene with personalized retention strategies
• Difficulty forecasting future revenue with confidence, impacting financial planning

Without a data-driven way to predict churn risk, retention efforts were reactive, generalized, and less effective.

CVM partnered with Plan Canada to build a set of predictive churn forecasting models using machine learning techniques.

Key components of the solution included:

• Churn Risk Prediction: Developed models that predict the likelihood that a given monthly donor will cancel within a future time window.
• Donor Profiling: Identified key behavioral and transactional signals correlated with churn, such as declining engagement, payment issues, and interaction history.
• Retention Targeting: Enabled Plan to prioritize retention efforts on donors most at risk of churning, allowing for focused interventions like special outreach or adjusted messaging.
• Revenue Forecasting: Provided finance teams with improved projections based on predicted donor survival rates, enhancing financial planning accuracy.

The models were built using Python and integrated into Plan’s existing donor management systems for ongoing, real-time use.

• Targeted Retention Efforts: Plan could now intervene earlier and more precisely, focusing efforts where they would have the greatest impact.
• Reduced Donor Churn: Targeted retention strategies helped improve donor retention rates, securing more long-term revenue.
• Improved Financial Forecasting: Enhanced visibility into future revenue trends strengthened financial stability and planning processes.
• Culture of Data-Driven Decision Making: Success with churn models helped foster broader organizational buy-in for analytics initiatives.

• Early Detection is Critical: Identifying churn risk early creates new opportunities to retain valuable donors.
• Predictive Models Enhance Strategy: Data-driven retention strategies outperform generic, "one-size-fits-all" approaches.
• Analytics Can Drive Cultural Change: Tangible success builds internal trust in data-driven methods over time.