Effective fundraising is about more than just outreach — it’s about engaging the right donor with the right message, at the right time, through the right channel.

Plan Canada faced several challenges in optimizing its campaign performance:

• When should a donor be asked for additional support to maximize responsiveness?
• What donation amount would each donor be most likely to give?
• Which outreach channel would be most effective for each segment of the donor base?

Traditional segmentation based on giving history and demographics wasn’t sufficient to personalize outreach at the scale required.

CVM collaborated with Plan Canada to develop a set of predictive models that guided campaign strategy and personalized donor engagement.

Key components of the solution included:

• Ask Timing Modeling: Predicted optimal solicitation windows to engage donors when they were most likely to respond positively.
• Donation Amount Prediction: Estimated the ideal ask amount for each donor based on prior giving patterns and engagement signals.
• Channel Preference Modeling: Identified the most effective communication channel — such as email, direct mail, or phone — for each individual donor.
• Seamless Integration: Delivered model outputs into Plan Canada’s existing CRM and marketing tools for immediate application by campaign teams.

The models were built in Python and designed to evolve with new campaign data, ensuring ongoing relevance and accuracy.

• Improved Campaign Performance: Personalization at scale led to higher conversion rates and increased revenue across channels.
• Stronger Donor Relationships: Outreach that better matched donor preferences resulted in improved engagement and retention.
• Better Use of Budget: Resources were allocated more efficiently by prioritizing high-propensity donors with tailored asks.
• Reusable AI Framework: The predictive models provided a foundation for ongoing improvements across future fundraising initiatives.

• Personalized Campaigns Drive Results: AI makes it possible to tailor timing, messaging, and channel for each donor — improving effectiveness.
• Data Integration is Key to Impact: Predictive insights must fit directly into campaign workflows to be useful.
• Modeling is a Process, Not a Project: Ongoing refinement ensures that targeting strategies stay effective over time.