Philanthropy, fundraising and higher education scholar Noah Drezner, Ph.D., discusses the value of collecting data about your prospects and donors and using that data to raise funds more effectively and efficiently.

Before returning to graduate school, I served as a development officer for the University of Rochester. As a new fundraiser I continually wanted to learn how best to solicit our alumni and how and why our office determined which strategies to do so. The typical answer I would receive from my co-workers, immediate supervisor, and even the dean of advancement was “because this is what has worked at other schools.” 

When I pushed back on this answer, saying, “We are not another school. We are Rochester, why not do what is best to engage our alumni-not what is best to engage someone else’s alumni,” I was sent to a conference. At the conferences I typically got answers like: “Fundraising is an art.  We know that personal connections are important but when it comes to mass mailings it is trial and error.” This did not settle well with me. By understanding our donors, prospects, and non-donors through the information and data that we collect on them, we can make more informed decisions in our practice. 

Why use data?

Not too long ago development offices consisted of calling cards and many paper files. With access to less expensive, smaller and more powerful computers, how we manage our prospects drastically changed. Large data sets are now commonplace and segmentation within the annual fund or other mass mailings that were once nearly impossible are now rather simple.  Joshua Birkholz (2008, p. xiv) suggests that “Like the emergence of volunteer-driven fundraising and the creation of voluntary associations, data mining is the next great breakthrough in the fundraising industry. Its impact on development programs will be as great as any change in the past 50 years.”

It is common knowledge that the most successful fundraising strategies all involve personal solicitations that take into account the prospective donors interests.  However, it is simply impossible, nor is it cost-effective, to personally ask each of our prospects.  Using the data that we collect about our prospects and donors can help us understand these individuals, thereby making annual fund and other mass mailing solicitations more personal.  The use of data can add a personal solicitation aspect to traditionally less personal fundraising tactics.

This is not to say that the multiple relationships that donors have to your organization and development officers are no longer important. Rather, using analytical data can bolster your programs.  Birkholz (2008, p.xv), rightfully states that  “The new tools of analytics, when combined with centuries of insights about private giving support and volunteering, open new possibilities to build upon the current practices of fundraising and to further the important work of philanthropy.” Data collected from personal contacts and prior mailings, phone calls and event responses can tell us a great deal about how to develop future strategies and segmentations.

What can you tell from the data?

The information that you can gain from your data is only limited by the data that you collect. By looking at data such as date of gift, solicitation code and payment method, simple data mining can help identify donors’ giving patterns and behaviors that can lead to donor retention and even upgrades.

For example, you can tell when your donors typically give. Do they respond to end-of-the-fiscal-year calls for action? Do they prefer to give at the end of the tax year? Do they respond to ‘crisis’ solicitations? What are their solicitation preferences? Do they prefer direct mail, e-solicitations, or telephone calls? How do they give? Do they give online, by check, or by credit card? With this knowledge you can segment your mass solicitations by time of year and solicitation method that will be more successful and cost-effective to your budget based on the given donor.

More complicated analysis of donor behavior and knowledge of your past solicitations-both who gave and who you solicited without success-can tell you what priorities donors have for their giving to your organization.  For example, which donors are more likely to give to a special project such as library acquisitions or emergency scholarship aid, or, who responds well to requests for purely unrestricted dollars? As your organization tries new messages to communicate your case for support, understanding your data can help you realize not only how effective that message might be but with whom was it most effective and with which groups should you change your strategy.

Data mining example

Peter B. Wylie and John Sammis (2008) who have both written and worked extensively on data mining within the nonprofit world and more specifically within higher education have developed some simple models to show the power of data. In one such report they look at a model that they employed over time at five different higher education institutions, each of a different type.  Using their model and the data that each university assigned them they created a score for each alumnus and alumna that predicted their likelihood to give to their alma mater. After making their predictions, they then monitored the same alumni’s giving to see if their predications were correct. 

They found with good accuracy the scores predicted the alumni giving over multiple time intervals (some models were tested within an interval of a few months others with multiple years, showing a large power to the predication). Wylie and Sammis (2008) found that if the institutions had used the scores that the predictive models produced in order to segment their annual fund appeals, the institutions would have known where to focus their efforts, saving the institutions money and resulting in much higher participation rates as well.

A cautionary note

The power of data is extreme.  However, it can also send development officers down wrong paths. In order to reach the goals of efficiency (minimizing fundraising costs) and effectiveness (maximizing growth in giving) some might interpret the data in a way that might lead them not to want to continue to solicit and engage a certain segment of their database. Before making a decision such as that I suggest that we have an obligation to take a step back and see if there is something more there than what our data is telling us. 

For example, if data mining within a college suggests that a large segment of alumni of color should no longer be solicited, as they have not given before, I suggest that we do not simply follow that suggestion. Rather, we should ask ourselves “how are we not serving this population?”  There is a growing body of literature that looks at philanthropic giving within communities of color and how it differs in motivation and practice from the white majority (e.g.; Smith, Shue, Vest, & Villarreal, 1999; Gow Petty, 2002; Gasman & Anderson-Thompkins, 2003). Understanding these differences and the fact that the first principle of fundraising is that ‘people give because they are asked,’ before we ‘write them off’ we should look to our strategies and see if we can engage these populations in a more culturally sensitive way.


Using data in the creation and implementation of your fundraising strategy is the centerpiece of the Fundraising Effectiveness Project (FEP), a joint initiative of the Association of Fundraising Professional (AFP) and The Center on Nonprofits and Philanthropy at The Urban Institute. The FEP project looks to help nonprofits increase both their efficiency and effectiveness. By understand the power of data, collecting it, and analyzing it, all nonprofits can reach the both of these coveted ‘ideals,’ while also

allowing our organizations to raise the needed dollars to fulfill our respective missions. The use of data within fundraising allows “our strategy [to be] grounded in facts, not assumptions” (Birkholz, 2008, p. 3).

Allowing data to help guide strategy will never make the traditional development officer and their personal understanding and experiences become obsolete. In fact, the development officers knowledge of donor motivation and practice will always be need to fully interpret the analytics that will come from the statisticians. Birkholz (2008, p. 210) reminds us that “it is easier for fundraisers to be successful when they are armed with the knowledge of context and implementation than it is for statisticians armed only with technology.  [However,] when knowledge and technology come together, the potential is limitless” (p. 210). In the end, using data does not take the “art” of advancement away, it just adds the science.

Reprinted with permission from Dr. Noah Drezner and the AFP Resource Center.

Dr. Noah D. Drezner joined the University of Maryland faculty in the Fall of 2008.  He holds his Ph.D. and a masters degree in higher education from the University of Pennsylvania and a bachelors degree from the University of Rochester. In addition, Noah holds a graduate certificate in non-profit leadership from Roberts Wesleyan College. His research interests include philanthropy and fundraising as it pertains to colleges and universities, including higher education’s role in the cultivation of prosocial behaviors. Further, his research focuses philanthropic giving in “non-traditional” donor groups.


  • Birkholz, J (2008). Fundraising analytics: Using data to guide strategy. Hoboken, NJ: John Wiley & Sons, Inc.
  • Gasman, M., & Anderson-Thompkins, S. (2003). Fund raising from Black college alumni: Successful strategies for supporting alma mater. Washington, DC: Council for the Advancement and Support of Education.
  • Gow Pettey, J . (2002). Cultivating diversity in fundraising. San Francisco: Wiley & Sons.
  • Smith, B., Shue, S., Vest, J. L., and Villarreal, J. (1999). Philanthropy in communities of color. Bloomington: Indiana University Press.
  • Wylie, P. B. & Sammis, J. (2008).Does data mining really work for higher education fundraising? A study of the results of predictive models built for 5 higher education institutions” Washington, D.C.: CASE. Available here.
Share Button