Predictive Analytics

Wood for Trees have extensive experience of developing predictive and prescriptive models. We have well established processes and techniques for developing propensity models and that have been successfully used by many of the charities we work with. These models have been used to boost income, improve cross sell, engage lapsed supporters and improve communication strategies.

We pride ourselves on creating models that make good business sense as well as being statistically accurate. We are able to do this because we understand the context of fundraising that we are observing the results in and we understand the language and ways of giving behaviours that are being evaluated. We will also work with you to ensure the model is fully understood, applied in the most appropriate way and monitored on a regular basis.

Not only can models used to improve individual campaign performance but they provide a framework for selections and longer-term campaign evaluation. And it is these modelling techniques which also help enable next best action models. To know what supporters are most likely to engage with it is necessary to create a series of response ranking algorithms – such as propensity models – which can then be combined and compared across your appeal programme to produce optimised journeys for supporters.

Wood for Trees use a range of analytical data mining software, including R, Python, Faststats and SAS to create our models. We also have a suite of survival-based models to look at predicting longer term supporter outcomes such as the likelihood of a donor to double their annual giving value or identifying those donors most likely to hit mid or major value giving levels.

Want to Learn More?

If you have questions about any of our services please feel free to get in touch.