The Old is New Again (Segmentation), and For Good Reason!

I’ve been seeing a lot written about segmentation lately. I even attended a Network for Good NonProfit 911 webinar on that very topic this week-and it was excellent. The concept of segmenting your prospect and donor data is nothing new, but the recent re-focus on the concept has me “all aflutter.” I love segmentation because it’s a beautiful example of how you can make all that data you’ve been so diligently gathering and entering into your fundraising CRM system work for you.

By segmenting your data you can:

  1. Send tailored communications to your constituents, ensuring the message they receive is geared toward them.
  2. Run reports to compare how various segments respond to your campaigns.
  3. Run queries to slice and dice your data based on any number of ways you wish to segment your data (constituency, gender, profession, highest level of education, marital status, campaigns, funds, gifts given in tribute, soft credits, and the list goes on).
  4. Create reports for your queries using the quick reporting tools in ResultsPlus. A go-to function of this tool is “sum”, but did you know you can also get averages, and perhaps even more importantly, medians? When working with donation amounts, for example, averages can be skewed high or low by your outlier donors, but the median will be the amount below which half of your donors gave and above which the other half gave. It is often a better indicator of the amount your “middle” donors give.
  5. Make data-driven decisions! Items two through four put you in the best position to make decisions and to improve how you work on your relationships with your donors, prospects, and volunteers. They can provide visibility into your past campaigns and appeals, identifying which ones worked well for which segments (and which ones didn’t work as well as you intended). This information is invaluable when working on new campaigns because it can help steer your efforts towards the actions and messages that are proven to work, and it can bring to light the areas you may want to adjust.

With all these great things you can do with segmentation, perhaps a slight distillation of the term is in order. There are, what some consider complex, forms of segmentation like RFM (also known as RFA analysis in some circles). This form of analysis generates what is called a segmentation code. The code is made up of 3 parts:

  • Recency: When the last donation was made (usually defined as a range, like 1-3 months, 3-6 months, 6 months – 1 year, etc.)
  • Frequency: How many times the constituent has made a gift (also a range like 1-3, 4-6, etc.)
  • Monetary (or Amount): This part of the code can be either the total amount of gifts ever given (the more common approach) or the single largest gift ever given.

These 3 items likely sound familiar to you, as most fundraising databases track this information. What’s unique about generating the code is that you:

  1. Don’t have to write a bunch of individual queries to put all the combinations for Recency, Frequency, and Amount values together to retrieve records.
  2. Don’t lose the historical information when the donor makes a new donation. Because a new donation changes the Recency and Frequency values (and usually the Amount, as well), so the segment changes. This is a good thing, but you may not want to lose the past segments a constituent has “passed through” as you’ve produced your appeals (because this information can be really useful when analyzing performance later).

However, there’s more to segmentation than RFM analysis. And for many organizations other ways of segmenting data are just as useful, and sometimes more useful, than RFM analysis. You may choose to review segments based on constituency, marital status, a special interest category, people you know are on twitter, Facebook, etc. Perhaps you even combine these things with one or more aspects of RFM analysis. The key is to know why you are looking at specific segments, and what you intend to do with your findings. Is it to learn how one behaves compared with another? Is it to analyze whether one segment is getting a more well-rounded view of your organization because of your approach to communications? Is it to help a specific segment to be more involved with your organization in some way? The questions are as endless as the ways you can segment your data. And this is a very beautiful thing.

Are there specific ways you segment your data that have served you well? Do you have a way you’d like to segment your data but are unsure how to go about it? Feel free to post them in the comments section.