About the client
Parkinson’s UK is a leading health charity focused on improving life for people affected by Parkinson’s. Traditional cash appeal performance had declined, primarily due to over-reliance on RFV-based segmentation.
Challenge
Cash appeal returns were deteriorating due to repetitive targeting of the same known responders. The supporter base had untapped segments being consistently excluded from mailings. Parkinson’s UK needed a scalable, data-led approach to identify new, high-propensity givers within their existing database.
Our solution
Wood for Trees developed and deployed a two-part predictive model:
- Model 1: Identified historically high-performing donors for retention
- Model 2: Surfaced unmailed supporters with strong giving potential, based on 250 behavioural and transactional predictors
Initial deployment using FastStats added 23,000 previously unmailed supporters, generating £22,000+ incremental income. Model refinement continued with each deployment, improving targeting precision.
Results
- Over £485,000 in total income generated from predictive modelling
- Net gain of £405,000, with model development costs recovered in the first campaign
- Six campaign deployments uncovered 8,500+ new givers vs. 2,400 via traditional segmentation
- Mailing efficiency improved: low-propensity segments excluded, high-propensity segments activated
- The model continues to evolve and optimise with each use, creating a compounding return on investment
Predictive modelling transformed the cash appeal strategy, generating higher returns from existing data and unlocking long-ignored donor value.
The predictive model has not only saved our declining warm appeals programme – it has pinpointed and released previously untapped sources of income from within our own database, as opposed to the alternative of a high-cost/high-volume-based acquisition programme.
James Culling, Head of Individual Giving, Legacies and Membership, Parkinson’s UK
