November 27, 2008 | Read Time: 1 minute
NEW BOOKS
Baseball, Fundraising, and the 80/20 Rule: Studies in Data Mining
by Peter B. Wylie
Statistical modeling can help fund raisers make their appeals far more cost-effective by focusing on those donors most likely to give, writes Peter B. Wylie, an industrial psychologist and data analyst.
He writes that any fund raiser with at least a few thousand potential donors can benefit from statistical modeling, which is basically a process of identifying the most-generous donors, developing an equation to figure out who is most likely to give, and testing that approach. The analysis allows fund raisers to concentrate their efforts on those donors who are likely to give the most.
Some of his findings are surprising: For instance, a donor is more likely to give if he or she is listed in your database with the prefix Mrs., Ms., or Mr., and with a business telephone number and e-mail address, because such information usually means the donor has a closer relationship to your group than someone whose profile is less detailed. Those fields can be a part of the statistical-modeling equation that you use to identify your best prospective donors.
Mr. Wylie demonstrates how these steps can be applied to different areas of fund raising, such as the simple formula that can identify potential major donors from people who give online. He also includes sections on college phonathons, giving by former fraternity and sorority members, and how organizations can build in-house staff expertise in statistical modeling.
Publisher: Council for Advancement and Support of Education, 1307 New York Avenue, N.W., Suite 1000, Washington, D.C. 20005-4701; (202) 328-2273; fax (202) 387-4973; books@case.org; http://www.case.org; 111 pages; $19.95 for members, $24.95 for nonmembers; ISBN 0-89964-417-1.
