Dataro uses machine learning to predict donor behaviour. The data from your organisation’s CRM paints a very rich picture of who each individual donor is and your relationship with them. That can include everything from their transactions and communications histories, through to their age, location, relationships, demographics, emails that they have opened, acquisition source and more.
Unfortunately, most NFPs have not been able to use all of that information when deciding which donors to include in each fundraising program. Instead, charities have been limited to outdated methods like RFM analysis, which only considers the recency, frequency and monetary value of gifts to generate broad ‘segments’.
Machine learning, on the other hand, is a sophisticated method used to identify patterns in the data and discover exactly how important each ‘factor’ is and how they can be combined mathematically to produce models that predict the likelihood of particular donor behaviours. Dataro uses the resulting models to generate propensity scores and ranks for common fundraising campaigns. Examples of common campaigns include warm appeals, regular giving upgrades, regular giving reactivations, regular giving attrition, conversion to regular giving, and many more.
The new building blocks for campaigns are propensity scores and ranks. Propensity scores generated using our machine learning approach range from 0-1, and represent the probability of a person doing a particular thing. The higher the score, the higher the probability of an individual taking the associated action. So if we were trying to predict direct mail giving, a person with a score of 0.76 would have a 76 percent chance of giving in the upcoming DM appeal. Ranks tell you how likely a donor is compared to everyone else in the database. So a person with Rank 1 is the most likely to take a particular action.
It you want to know more, check out our attachment: Dataro's Machine Learning Pipeline Explained.