Getting Started in Data Mining (It's easier than you think!)

I have had a few requests for articles on simple data mining techniques and the related database maintenance necessary to make the results meaningful. Look for my upcoming companion blog post on data mining resources, too.

Before we get started, let’s talk a little bit about what might be holding us back.

  • Fear that it’s too complicated – Not much anyone can do about this one, except you. Jump in! The water is warm!
  • Assumptions that leadership will not invest and support it – Data mining and analytics are keyword candy to leadership. Leadership loves to get intelligent answers to questions like “What percent of donors rated at $100K+ gave at that level?”
  • No clear understanding of the pain/need/goal – What keeps your leadership awake at night? Is it prospect pools that don’t perform? Finding leadership donors for the upcoming campaign? If you don’t know, you can’t make a compelling case for data mining.

Donor Database Reports

Do you remember that scene in the Sound of Music where Maria is trying to teach the von Trapp children to sing? She stops singing “Do-Re-Mi” and says, “Oh, let’s see if I can make it easier”. We can do that in data mining too. (I haven’t come up with a song yet, but I’m working on it.) Here is an easy and fun way to get started in data mining – explore all the canned reports in your donor database. I’m not kidding! Even if you have no idea what deep, insightful questions you want to answer, you can begin with reports.

Consider these common reports:

  • Consecutive years giving – When donors give many years, especially consecutively, it usually means they really like us. Who are these people? Do they have high wealth ratings? Could they be good planned gift prospects?
  • Top donors – Are all of your top-giving donors getting regular attention?
  • LYBUNT, SYBUNT, & new donors – Within these reports you might find donors capable of increasing their gift, some major gift sleepers, and some new donors with wealth.
  • Lifetime giving and number of years giving – So many forgotten donors can be found in this list as well as some very good planned gift prospects.

Digging a Little Deeper

MS Excel is on most of your desktops. If you take a little time to learn to use it – I’m not talking complicated formulas, just tips and tricks – it will truly open the world of data mining to you. Imagine that you pull a report into Excel with all of the key fields in the above reports (last gift date and amount, largest gift date and amount, lifetime giving, etc.). Add in wealth ratings if you have them.

Now consider this scenario:

Custom sort:  First by largest gift amount (descending), second by lifetime giving (descending), third by last gift date (descending)

Analysis:  By scrolling down the list you can see if any donors who have made larger gifts (largest gift amount) and have lapsed (last gift date). Is there some high lifetime giving low on the list? Why?

Imagine sorting first by wealth rating and then largest gift. How about lifetime giving and wealth rating? This is fun! (I told you the water was warm.) Just be sure to watch your time. Prospect researchers have gotten lost in the data mining game.

The Secret Data Mining Trick

The secret trick to analyzing your donor information is to understand your fundraising fundamentals. Remember the fundraising pyramid?

The pyramid illustrates your areas of opportunity:

  • Occasional: Did that first-time $1,000 donor get personal attention?
  • Annual: Are there small annual gift planned giving prospects in there?
  • Annual: Can we motivate annual donors to move up a giving level?
  • Major: Do any of your major gift donors have unexplored planned gift potential?
  • Planned: Are there any planned gift donors who could make a cash gift?

Common Data Errors that Under-Mine Your Efforts (pun intended)

Now that you have the idea that you can sort on specific fields in your donor database, you will very soon realize that even sorting becomes problematic if the data is full of errors and omissions. Use your blossoming interest in data mining to clean up the database! Then when you are ready for more complicated data mining challenges, your data will be ready for you.

  • Data errors in any of the fields you pull – e.g., incorrect or missing dates or dollar values
  • Duplicate records – often happens in gift entry or multiple hands in the database
  • Deceased or bad address – if you don’t mail to your list, you probably aren’t getting your list cleaned; if you are mailing, you might not be getting a file back from the printer to update the records

What can you do about problems like these? People don’t usually like to hear this, but you need some documentation.

  • Your database probably has some maintenance reports. Set up a schedule to run them and fix the errors.
  • Do you need to run a report of all changed records daily or weekly?
  • Gift entry staff should be trained to search for the donor name first, instead of entering a new record. As in, create your own training manual for how gift entry is performed in your organization.
  • Someone should review all gifts entered, probably daily.

Robert Weiner is a consultant with some excellent free articles about keeping your database up to snuff. You can find his articles here: http://www.rlweiner.com/articles

Taking Data Mining to the Next Level

Once you have your data in order, some understanding about how the information is stored, how you can retrieve it, and what kinds of things it can tell you about your donors and prospects, I suspect you will be a lot more likely to sign up for that data mining webinar or take advantage of the APRA Analytics Symposium. It feels good to be ready, doesn’t it?

One thought on “Getting Started in Data Mining (It's easier than you think!)

Comments are closed.