Data and the Search for Big Donors

Eric Kayne, For The Chronicle

Imagine the development office of 2025. Switching on their computers, fundraisers will dig into the information that charities have long collected about donors — giving histories, event attendance, visits with nonprofit staff.

But a few keystrokes will also yield a wealth of other, more personal data. Software-produced analyses of social-media presence will identify donors’ interests and single out friends and family already in the charity’s universe. Census and real-estate data will offer insights about how they live. What magazines they read, what cars they buy, where they shop — all this will be at a fundraiser’s fingertips.

Taken together, the data will offer an unprecedented — and intimate — picture of every supporter. Fundraisers will be able to make smarter decisions about which donors to pursue and how to tailor their messages. They will micro-target appeals with ease, tailoring stories and imagery to like-minded groups of donors. Major-gifts officers will arrive at their first meetings with potential contributors equipped with the personal background to quickly build rapport and design proposals that speak directly to donors’ passions.

Such a fundraising operation is hypothetical, of course; among other challenges, there are tricky privacy issues to navigate. But it’s not science fiction. Some of the tools and ideas behind that vision of the future are already playing out in development offices across the country. At nonprofits in the vanguard of the data-analytics movement, number-crunching plays an outsize role in how they home in on potential donors and seek contributions.

Analytical techniques to boost fundraising are growing in popularity and power, says Josh Birkholz, a principal at the consulting company Bentz Whaley Flessner. A dozen years ago, when he was director of analytics at the University of Minnesota, he had few peers in similar positions. Now he estimates there are several thousand, an important factor in charities’ growing sophistication in winning support.

“Fundraising is maturing,” Mr. Birkholz says. “We’re in this generation where fundraising is industrializing.”

As the use of data to improve fundraising gains momentum, a handful of charities are pushing the boundaries. They’re mixing their information with new public data, mining donors’ social-media activity, and applying data analysis to the fundraising process itself.

Here is a sneak peak at the future of data analytics in profiles of three of the newest cutting-edge approaches.

Who’s Most Likely to Give?

When the University of South Dakota embarked on a $250 million campaign in 2012, development leaders wanted to make sure that fundraisers were talking to the people likely to give. To ensure that all those trips and outings to wine and dine potential donors led to gifts, the university’s foundation turned to predictive modeling.

Sharyl Loen

It commissioned Target Analytics, a data-consulting service run by the software company Blackbaud, to create a detailed statistical picture of contributors who had previously made very large gifts to the university. Target’s analysts used both the university’s information and outside data sets to look at myriad factors of a person’s life — whether they pay off their credit-card bills each month, how many children they have, the number of cars they own, and more.

With the resulting model of big donors, the company scours the foundation’s database to identify others who share similar characteristics and behaviors. The analysis also ranks the supporters it finds according to their likelihood to give.

The model cost roughly $15,000. Four years later, it has identified individuals who have contributed $61 million. These newfound big donors include supporters who weren’t on fundraisers’ radar. Others had been pegged as likely to make much smaller gifts, says Margaret Williams, director of prospect research at the University of South Dakota Foundation.

“We’ve been brave enough to remove the glass ceiling that we had installed there and asked them to stretch,” she says. “And sure enough, they’ve stretched.”

Predictive models such as South Dakota’s tend to show a picture in time: Fundraisers usually run them at set intervals, often once a year.

But a few organizations have embedded the models into their databases and run them more frequently. This approach is called dynamic scoring. The goal: Always work with up-to-date information and flag donors whose likelihood to give has changed.

The Alley Theatre in Houston incorporated two predictive models into its database in late 2014 — one that predicts which supporters are most likely to make a gift of $10,000 or more, and one that identifies ticket buyers most likely to become donors. The major-gifts model analyzes hundreds of factors. Some are related to the donors’ giving, like their number of gifts, the amount of their first gift, and whether they gave more with their second gift. Other factors are less directly connected to giving, such as whether the theater has a business address for the individual.

Officials say the predictive modeling brings in an additional $100,000 every six months — a handy sum for a theater that raises a little more than $7 million in contributions annually.

The fundraising department gets weekly updates on the big-donor analytics. When the theater was renovated last spring, Amy Schwab Lampi, associate director of development, used the model’s analysis to set the invitation list for three VIP hard-hat tours, which led to several large contributions. Before donor events, Ms. Lampi prints out a cheat sheet with the major-gift ranking of attendees so she can focus her attention on those most likely to give. Overall, she says, the analysis has helped the Alley shorten the time it takes to secure big gifts.

Universities and hospitals take advantage of predictive modeling more than other charities, but Ms. Lampi contends it can have a big impact with smaller nonprofits that have few staffers devoted to securing large gifts. The Alley Theatre has one full-time fundraiser for major gifts and two others who focus on them as part of their broader work.

“It’s changed who we are approaching and who we’re spending our time on,” she says. “It allows us to streamline our practices.”

The Alley uses Tessitura, a ticketing and fundraising database created by the Metropolitan Opera, which spun off an independent nonprofit, the Tessitura Network, to continue developing the software. Ms. Lampi has worked with Bentz Whaley Flessner, which creates its models, to encourage other Tessitura users to adopt the approach. Lyric Opera of Chicago, the Seattle Opera, and the Shedd Aquarium have signed on.

Dynamic scoring has the greatest potential for nonprofits such as hospitals or theaters that deliver a service or experience to donors and potential supporters, says Mr. Birkholz of Bentz Whaley Flessner. It’s important to follow up soon after that interaction, he adds.

“If you’ve got a good health-care experience, those first few months are the most important, so health care’s had to become really nimble,” he says. “The arts has that same challenge. When someone goes to a show or they just became a subscriber, that’s kind of the peak moment, not five years after.”

When Facebook ‘Likes’ Turn Into Gifts

Raising cash via social media hasn’t been the bonanza many nonprofits expected. But fundraisers are starting to glean information from social networks to better target their appeals.

The Massachusetts Institute of Technology used Facebook to identify the passion that might spark giving among donors who hadn’t made a contribution in five or more years. MIT’s fundraisers analyzed the lapsed donors’ “likes” on the university’s Facebook pages, then sent them an email newsletter that featured stories on the three topics that had generated the most interest.

The fundraisers then tallied which articles were most read. The winner? A feature on the university’s robotics program. MIT followed up with an email to those who had read the article, asking for a contribution to a robotics crowdfunding campaign. The percentage who responded with a gift was small but still roughly twice the rate of past solicitations to lapsed donors. So the university sent the email to the rest of the group.

The appeal, which went to 55,000 lapsed donors, raised nearly $30,000 — including one $20,000 gift. Such a big gift from a little-heard-from alum is unusual, says Tim Poisson, director of marketing and participation at the MIT Annual Fund. And it was a great lead to pass on to the major-gifts office.

“We’ve given the development office essentially an entry point,” he says.

Oregon State University is starting to take social-media information into consideration as it assigns donors to major-gifts officers.

The university’s fundraisers can visit and build personal relationships with only a small number of the people in its database. To determine which donors merit the attention, Oregon State’s foundation typically used wealth data and an engagement score based on factors such as whether the person contributes, volunteers, or attends events.

As the university analyzed interactions with its more than 50 Facebook pages, fundraisers wondered if the data could help gauge donor interest. So the foundation conducted an experiment.

Fundraisers identified 100 people in the database who were extremely active on the university’s social media but had an engagement score of zero. These individuals also had the capacity to give $25,000 or more, according to development-office ratings.

Fundraisers then called to see if they could set up a meeting in person or over the phone. Typically, when the university’s fundraisers reach out to set up first visits with potential major donors, 20 percent agree. Of the 100 people highly engaged with the university’s social networks, 44 percent said “yes.”

The results, though not scientific, were encouraging enough that Oregon State now rates alumni as engaged if they have liked more than 50 university Facebook posts or commented at least 10 times in a year.

The fundraising office receives weekly alerts noting posts that have garnered “likes” or comments from highly rated prospects. The information is forwarded to the donor’s assigned fundraiser. In time, however, the university hopes to capture those donor-interest tidbits in its database, says Mark Koenig, an assistant vice president at the foundation.

“That kind of insight would help us as we start to get more and more sophisticated in our analytics program,” he says.

Both Oregon State and MIT use software from EverTrue, a company that lets fundraisers pair donor-database information with data from the universities’ Facebook and LinkedIn pages and from Zillow, Google Maps, and the U.S. Census Bureau.

Bringing all that information together helps fundraisers plan events for donors who work in a particular industry or region. And it simplifies preparation for meetings with out-of-town contributors, says Anthony Barbuto, assistant vice president for leadership gifts at Boston University.

“Instead of having a full week of prep before a trip, it might be just a few days,” he says. “I can make more trips in a month than I would have otherwise.”

So far, EverTrue’s analysis of donors’ social-media activity has focused only on their interactions with the client nonprofit’s digital content, says founder Brent Grinna, but the firm hopes to expand to look at donors’ entire social-media presence.

“When you think about why Facebook, for example, is so valuable, it’s because it allows brands and organizations to target people based on specific interests,” says Mr. Grinna. “We don’t think there’s any reason that nonprofits shouldn’t be able to benefit from it in the same way.”

Chip Somodevilla/Getty Images

Studying What the Best Fundraisers Do Right

A growing number of organizations are combining examinations of donor behavior with analyses of their own work and the very act of raising money.

Memorial Sloan Kettering Cancer Center, long a leader in predictive modeling, recently hired an analyst with expertise in what’s called continuous process improvement. The goal is to break fundraising into its component parts and examine each to spot problems and potential enhancements. If changes result, the hospital will set up feedback loops to both measure the impact and identify other possible improvements.

An early project analyzed the role of the center’s prospect manager, who evaluates whether a potential donor should be assigned to a fundraiser. Information came to her from multiples sources, often in different formats. To make decisions, she had to consult various reports, switch between software programs, and manage multiple spreadsheets.

The New Frontier of Data and Fundraising

How nonprofits mine databases, sift Facebook chatter, and more to find big gifts.

• How to Rate the Potential of Prospective Women Donors
• A.I. Is Ready for Fundraising. Are Fundraisers Ready for A.I.?
• How to Identify Donors With the Greatest Potential to Give

After mapping out the process, the analytics team pinpointed inefficiencies, standardized reports, and automated much of the movement of prospects through the steps of the assessment process. The changes freed up the prospect manager to do more important work, says Kate Chamberlin, Memorial Sloan Kettering’s director of development analytics and process.

“Some of these are very nonexciting, back-end processes,” she says. “But they are what makes everything work.”

Some organizations are analyzing data on fundraisers’ activity — their phone calls, visits, etc. — to determine what sets top performers apart. And they’re uncovering fascinating patterns.

A New York University analysis found that what it considers its most successful major gifts — those closest to the donors’ capacity to give — resulted from 11 months of cultivation and five visits. By contrast, the large gifts that represented the smallest proportion of donors’ capacity to give came after six months of cultivation and two asks.

The findings make sense, says Erin Dodd, managing director of development and campaigns at the university. “If you’re going to ask somebody for a true major gift, something’s that’s meaningful to them, it takes time,” she says. “What did surprise us was how often we went too quickly.”

The university has incorporated the eye-opening analysis into its ongoing fundraiser training, but it’s important to draw the right lessons, cautions Ms. Dodd.

“The training doesn’t say, ‘Oh, you should take five visits,’” she says. “The training says, ‘Do you know what the donor’s most interested in?’ ”