Plan International UK and Wood for Trees, Salocin Group

With training and guidance from Wood for Trees, PIUK built a new legacy propensity model using Apteco. They explored both decision tree and profile models, testing and refining dimensions such as giving history, campaigning activity, gift aid, age and gender. The decision tree model emerged as particularly effective – delivering sharper segmentation, stronger predictive accuracy and clear guidance for campaign targeting. Crucially, Apteco’s intuitive environment enabled PIUK to test, refresh and adapt their models with speed and confidence. They could overlay results with supporter responsiveness by channel, helping them identify entirely new audience segments who had not been contacted about legacy giving for years.

The existing propensity model, built by an external agency, was based on age, wealth, gender, connection to the cause, and giving score. Leveraging the agility of Apteco in-house, the proposed design of the new model was more closely aligned with PIUK’s strategic vision of being supporter-led and insight-driven. The new model design initially focused on variables that best demonstrated a supporter’s connection to the cause. These included, but were not limited to, the number of lifetime products (split across cash giving, sponsorship, and other regular gifts), number of campaigning actions, number of complaints, gift aid status, and age and gender (where available). In addition, banded variations were applied to reflect the length of time since the last gift or payment, providing greater flexibility in how these variables were combined to identify the strongest-performing model. Due to the ease and speed with which predictive dimensions can be tested, PIUK was able to quickly identify the set of variables that delivered the highest predictive value. Over the longer term, using Predictor to create models in-house is more cost-effective, as models can be updated and refined as required at no additional cost. By contrast, changes to the external model would incur further costs if PIUK wished to amend fundamentals or request additional analysis. Furthermore, the Predictor model enables PIUK to develop additional models in-house at no extra cost.

This change gave PIUK full ownership of their modelling, reduced long-term costs, and aligned with their strategic goal of becoming more supporter-led and insight-driven. The results were impressive, and beyond the immediate impact, the project equipped PIUK with the tools and expertise to develop additional models at no extra cost, unlocking lasting value for the charity. By adopting Apteco’s Predictor bundle, PIUK has enhanced campaign performance, established a sustainable, insight-driven approach to legacy fundraising, and created new opportunities to re-engage supporters through the most effective channels.