“A Little Mining with that Model?” Part II | npENGAGE

“A Little Mining with that Model?” Part II

By on Nov 17, 2010


In Part 1 of this series we looked at the differences between data mining and predictive modeling.  We identified how they both take a look at “commonalities” or characteristics of individuals in your data base to identify patterns of behavior.  Both can be extremely useful when trying to enhance your overall fund raising effectiveness, as they allow you to take advantage of past behavior when developing new strategies.  With that in mind, a common question when exploring ways to enhance prospect research can be “which do I use?”  While predictive modeling requires a reasonably sophisticated process of statistical regression, data mining is much simpler.  I often explain to clients that data mining is more of a “2-D” process, while predictive modeling is more like a “3-D” one.   While not impossible, it would be much more difficult to effectively undertake a predictive modeling effort without outside assistance.  Data mining, on the other hand, is often a great first step as it can typically be done in house.   Many organizations find they gain valuable insights into important and useful patterns and information when they begin to analyze their own data.

So, what type of data can be “mined”?  In short, almost anything you have tracked or collected.  As fundraisers, most of our organizations keep pretty decent track of the staple of our business—gifts.  And, it can be amazingly enlightening to take a careful look at what type of patterns exist in your constituents’ giving history.  The even more amazing thing is how seldom most organizations take a look at this.  Case in point:  how much do you know about long time donors?  Take this simple test:  Run a query in your data base of individuals who have made 25 gifts or more. (You may find it useful to omit payroll deduction gifts from this exercise as this can “skew” the results.)  If your organization is relatively well established, you may find hundreds of individuals who have made this many gifts—or more.  Now, sort the list in descending order of number of gifts.  What did you find?  One organization I worked with recently had an individual on their database who had made 72 gifts.  And, they didn’t recognize the person’s name.  This is where the data mining comes in.

Take another look at your list.  Start enriching it with some additional information.  Add things like “date of first gift”, “first gift amount”, “date of largest gift” , “largest gift amount” and amount of each of the last five gifts.  See any patterns yet?  While you will undoubtedly see some individuals who have “behaved” exactly as you would have hoped—increasing their gift amount slowly but steadily, many others may have simply “plateau-ed”—or remained at the same or close to the same amount year after year.  And, you may also notice that a good number of individuals have never made what many of us would consider a “major” gift.  Take our individual above, the one with 72 gifts.  This gentleman had been giving to the organization for nearly 40 years.  His first gift was $25; his latest (and largest) was $100.  And, there were over 100 other individuals on the list who had similar patterns.  With very few exceptions most staff members didn’t recognize any of the names, simply because they hadn’t given any “big” gifts. 

What would you find if you took a careful look at this type of information?  Perhaps more importantly, what would you do once you discovered it?  This organization immediately started a “Loyalty Society”, a close cousin (and hopefully feeder pool) to their “Legacy Society”.  They saw their number of planned gift expectancies increase almost immediately, simply because once they noticed the pattern, they changed their behavior accordingly.

Loads of potential lies in other data you have at your fingertips as well.  For example, how many of your donors were involved in athletics?  Or Greek organizations?  Or have attended reunions?  As you start realizing the common denominators of your “best” donors, you can tailor your cultivation efforts accordingly—and in a much more targeted fashion.  So, get started today!  Take a careful look at your donors and see what patterns you can find.  You might be surprised what you see!


Laura Worcester, senior consultant at Target Analytics, joined Blackbaud in 2001.In her current role she advises nonprofits on utilizing screening results in identifying and evaluating best donor prospects. In 25+ years of fundraising experience, Laura has served as the chief advancement officer for numerous organizations and managed her own consulting business, providing grant writing services to arts, educational and health care organizations. She’s presented at development conferences and has been a regular contributor to Blackbaud’s blogs with selected posts being reprinted in journals such the NonProfit Times. A traveler since her study abroad days in Denmark, Laura’s committed to passing this enthusiasm on to her teenage daughters. Her family’s travel adventures were just featured in a neighborhood magazine in her suburban Milwaukee community. Contact Laura by email.

Leave a Reply

Your email address will not be published. Required fields are marked *