The following is the code I used for performing a longitudinal analysis of Bread & Roses Community Fund giving from 2015 to 2022.
My goal with this analysis was to stress the importance of iterative analysis for targeted fundraising campaigns and identify churn risk and lapsed donors. A comprehensive (yearly or quarterly) review would allow BRCF to formulate a fundraising campaign.
BRCF had no process or milestones for iterative donation analysis and relied on in-house perceptions and hearsay to formulate campaign strategy. BRCF relied on a variety of donation methods (cash, checks, PayPal, stock, etc.) to make donating accessible; however, this meant that the organization relied on multiple systems for receiving donations. These systems were not in communication, and the organization's CRM (powered by Exceed) did not have capabilities for integrating any payment method. All data had to be entered manually. Therefore, a major problem (and limitation) in developing donation or segmentation analyses was the messiness of the data.
This data includes information on donors and donations from January 2015 to December 2022.
I downloaded the data manually from Exceed and cleaned the data in R. I formatted dates and limited the dataset to the pertinent columns. I broke down the data into yearly averages and medians. Likewise, I created box plots to better show distributions per year. I provided monthly descriptions of donation reasons or motives: Giving Project campaign, Tribute to Change campaign, Change Ride campaign, and fund donations. I also provided descriptions of donation methods (ATH, cash, check, PayPal, etc.), hoping to find data that could help us improve donation experiences. I also generated lists on our top 100 donors, which were cross-referenced with CRM data on lapsed donations.
A major limitation of this study is the amount of missing data in our CRM. Throughout the years, BRCF's CRM was kept by different people from different backgrounds with no real, consistent standard for data management. BRCF kept most of its donations and acknowledgments in hard copies or in analog archives. The COVID-19 pandemic forced the organization to digitize documentation and better maintain its digital data.
A major finding of this analysis was the spike in donations in 2020. The COVID-19 pandemic, the George Floyd rallies, and the 2020 election resulted in a doubling of donations between 2019 and 2020. Between 2015-2022, 2020 was the most successful year in fundraising and contained the most variety in donation sizes. Another finding was the trend towards online donations (via PayPal) and away from cash or check donations. Finally, by generating a list of BRCF's top 100 donors were able to upkeep our CRM by updating data on donors. With aid from the Development and Communications Director, I generated 3 high-risk churn groups and generated donor history dossiers for targeted Board fundraising efforts.