How SOS Méditerranée used Predictive AI to identify its highest-potential mid-donors and cut campaign waste by 90%.
Corinne Melet leads fundraising efforts at SOS MEDITERRANEE, a European maritime-humanitarian organization dedicated to rescuing people in distress in the Mediterranean Sea. Operating across France, Germany, Italy, and Switzerland, the NGO relies heavily on private donors to sustain its life-saving missions. and amplify impact. He shares how Predictive AI helped Mercy Ships Norway sharpen its fundraising strategy, save time, and raise more funds.
Operating across several European countries and relying heavily on private donors to fund its rescue missions, the organization wanted to strengthen its middle-donor program and improve fundraising efficiency. While the fundraising team had a clear understanding of what defines a valuable middle donor, identifying those supporters within a large donor base remained difficult. Traditional segmentation methods lacked the precision needed to prioritize the right people effectively, resulting in broad campaigns with unnecessary costs and missed opportunities.
Like many nonprofit organizations, SOS MEDITERRANEE had access to years of donor data but limited ability to fully activate that data strategically.
Without a clear prioritization model, campaigns often relied on broader outreach strategies, increasing costs while reducing overall efficiency. Valuable fundraising resources risked being spent on low-potential contacts, while high-potential supporters could remain hidden within large donor lists.
The organization needed a solution capable of transforming historical donor behavior into actionable insights directly usable within campaign planning and CRM workflows — helping fundraising teams focus efforts where they would have the greatest long-term impact.
To improve donor prioritization, SOS MEDITERRANEE partnered with Allyy to implement Predictive AI focused on identifying supporters with the highest middle-donor potential.
Using eight years of historical donation data, Allyy analyzed donor behavior patterns such as recency, frequency, donation growth, and channel engagement. More than 120 behavioral variables were used to train machine learning models capable of assigning each donor a predictive score between 0 and 1.
Rather than functioning as a simple yes/no prediction tool, the model acts as a prioritization engine — helping fundraising teams rank donors according to their future likelihood to respond positively and increase engagement over time.
This allowed the fundraising team to reduce unnecessary outreach, improve campaign efficiency, and make faster, more data-driven decisions directly within their existing workflows.
By replacing guesswork with predictive prioritization, SOS MEDITERRANEE significantly improved the efficiency of its middle-donor fundraising efforts.
One of the most significant findings demonstrated the precision of the predictive model:
By narrowing focus while maintaining strong fundraising potential, the fundraising team was able to improve campaign efficiency and dedicate more attention to supporters with the greatest long-term value.
We help mission-driven organizations raise more — by focusing on the people most likely to give.
Gør dine donordata til flere donationer — med prædiktiv AI
© 2026 Allyy A/S. All Rights Reserved.
