This week’s article is from the keyboard of my colleague, Kenny Tavares, Senior Researcher and Data Insight Consultant. I asked Kenny to share his knowledge today on one of the most straightforward fundraising analytics projects we offer (which you can also do at home!) – RFM analysis. ~Helen
As fundraisers, when we think about data analysis today, we’re likely to drift into the world of complex modeling practices that predict what donors are most likely to give based on a wide variety of factors. Go to any conference and you’re bound to hear about the wonder of Big Data. But if you work in a smaller organization with fewer resources, you may have walked away feeling like none of this was for you. Let’s face it, just in terms of time spent, these projects seem inaccessible.
However, we all need to understand our donors better. The data we keep can unlock critical new insights for our organizations. Data analytics has benefits for all of us, so we all must find ways to sort through the information we have so we can make better decisions. But how to do it? Let’s start by assessing our top donors using three important metrics: an RFM analysis.
What is RFM?
Created for the purpose of corporate direct marketing, RFM is an analytical method used to determine an organization’s best customers — or in our case, donors. It has been around for more than 40 years, but a popular method for performing RFM analysis is credited to Arthur Middleton Hughes in his 1994 book Strategic Database Marketing: The Masterplan for Starting and Managing a Profitable, Customer-based Marketing Program. In the book, Hughes describes the method of binning three attributes into five ranges, which I will briefly describe in a moment. These attributes are:
- Recency (R): In our case, how recently a donor has made a gift
- Frequency (F): How often a donor makes a gift
- Monetary value (M): How much a donor has given
For fundraisers, these three areas are important because of what we know about donor affinity:
- Donors who have given recently are more engaged than lapsed donors
- Donors who give regularly are more engaged than infrequent donors
- Donors who make larger gifts on average are more engaged than donors who make smaller ones
Because it’s easier to retain donors than to acquire them, understanding who your best donors are is a no-brainer. Having a better feel for your donor pool allows you the opportunity to solicit them is a more efficient, cost-effective manner.
At this time, when predictive analysis has become more and more accessible, you might be asking “Why should I perform an RFM analysis?
To perform an RFM analysis, all you need for each donor is a constituent ID, the number of gifts they have made, their lifetime giving and the date of their most recent gift. Compared to more complex predictive modeling, the threshold for performing this task is much lower, making this project manageable for most organizations. However, if your giving records are inaccurate or incomplete, now would be a great time to address that issue.
Once you have your data, you’ll need to score your records based on the three categories. For each category, divide your dataset into fifths, giving the best fifth a score of 5, with the subsequent fifths scored from 4 to 1. For instance, if you sort the last gift date in descending order, your top fifth – the donors who have given most recently, would receive a score of 5. Repeat the process with the total number of gifts (frequency) and lifetime giving (monetary). In the case of lifetime giving, you may prefer to divide that figure by the number of gifts and use the average gift for your scoring. Once you’ve scored each variable, combine them to create your final score. The easiest way to do this would be to adjoin the three numbers to create the final score (ex. 555, 444, 333, etc.).
Now that you’ve scored your records, sort them by their final score and check out the results. Perhaps you’ll find some surprises. It’s all good if it improves your understanding of your donor pool.
RFM analysis operates on the Pareto principle, or the 80/20 rule. For our purpose, it would suggest that 80% of our gifts come from 20% of our donors. Knowing your 20% is critical. Keeping this group happy will obviously have long term benefits for your organization, as long as you don’t give them too much attention. But don’t stop here; there are other donors to consider.
Now, you might be asking “What about the remaining 80%?” It’s a good question, because it doesn’t make sense to cast them aside. Take the time to become familiar with the other segments and develop strategies for these groups as well. For instance, how can you motivate new donors to give again? How can you encourage loyal donors to increase their giving? How can you re-engage lapsed donors? Knowing these subsets allows you to better target them based on their behavior. Ultimately, you can craft your message to these groups more efficiently.
As I mentioned, our three variables, and what they tell us about the donor, make RFM analysis an exercise in predicting which group of people are most likely to give again. That said, it does lack the flair of a more sophisticated predictive model, which would entail the use of many more variables. However, the usefulness of your RFM analysis is ongoing. The more you learn about your top donors, the more adept you’ll become at identifying similar prospects. You may not be as quick as an algorithm, but your institutional knowledge is extremely valuable. Take the time to consider the following about your best donors:
- How old are they?
- What is their marital status?
- Where do they live?
- Where do they work?
- In which industry do they work?
- What are their job titles?
- How large was their first gift?
I could go on, but you get the point. Exploring the data on your top givers will improve your ability to recognize and efficiently target promising new donors.
Your ability to analyze the information you have available is a critical component to retaining donors and discovering prospects. Regardless of resources, we all need a way to recognize what the data tells us. Applying RFM analysis is a great place to start; it might even whet your appetite for other data strategies. Once you start using analytics to discover insight in your data, you’ll find answers to questions you might not even have thought to ask!