Will Big Data Be the Gift That Keeps Givers Giving?
Whether or not donors realize it, non-profits are going to target them with all they’ve got
--
The heavy envelope shows up in my mailbox with astonishing regularity: a “free gift” of bird- or flower-festooned address labels from a nature non-profit, meant to entice me to donate. While I’m happy to slap the labels on bills and thank-you notes to Grandma, I always recycle the donation forms within seconds of opening the envelope. It’s not that I don’t care about the environment, but I only give to causes closest to my heart (youth mentoring programs, the local ballet company). If this non-profit limited its outreach to bona fide tree-huggers, they’d likely get more donations—and they’d certainly kill fewer trees.
A lot of for-profit companies mastered this principle years ago, using targeted data mining, or analytics, to identify customer groups most likely to respond to their appeals. Market segmentation is the watchword: If statistics reveal that middle-aged motorcycle owners like to dine at Denny’s, those bikers might get a cache of Denny’s coupons in their inbox.
But non-profits have mostly lagged behind their for-profit peers, in part because limited resources make it tough to invest in data analysis. According to a recent Nonprofit Technology Enterprise Network (NTEN) survey, only about one in four non-profits use donor data to make decisions about their program’s direction. And many charities’ outreach tactics have remained stagnant, too—witness the never-ending address label ploy.
“90 percent of giving today is still done by direct mail,” says Cleveland-based fundraising expert Dan Mansoor. “It’s the last bastion of 1840s technology.”
The work non-profits do is more crucial than ever, especially as government funding for many social programs plummets and the gap between haves and have-nots widens. But keeping such organizations afloat has also gotten challenging as budgets shrink and donor numbers dwindle. These realities have convinced some insiders that smart data is the secret sauce non-profits need to up their game. And if non-profits get savvier and more effective, donors and participants could benefit, too. When you give to a worthy cause, research shows, your brain gets happy, and committed volunteers enjoy a “helper’s high,” reporting better health and more life satisfaction than non-volunteers.
For non-profits to pull off major social transformation, says consultant George Weiner, they need to start thinking more like their data-conscious for-profit peers: “We’re not trying to sell widgets, but we are trying to sell volunteerism.” Weiner, founder of the Brooklyn-based Whole Whale agency, is one of a band of experts imparting an urgent message to non-profits: If you’re a would-be world changer, 1840s technology isn’t going to cut it much longer.
Targeted data analysis got its start in the Eisenhower era. The Fair Isaac corporation pioneered one of the first predictive retail schemes in the 1950s, assigning credit worthiness to potential customers based on individual factors such as their debt or loan-paying history. As computing power grew by leaps and bounds, early models like Fair Isaac’s gave way to data-rich portraits of consumers’ retail likes and dislikes, purchasing patterns, and degree of brand loyalty. Those portraits enabled companies to highlight groups of prospective customers who would be most receptive to their message. By the 2008 election cycle, most serious political campaigns had jumped on the analytics bandwagon as well, using demographic and consumer data to supercharge voter targeting efforts.
More recently, analytics experts have started helping non-profits enjoy these same data-driven rewards. One such veteran is Ken Strasma, founder of the Washington, DC-based consulting firm Strategic Telemetry. Strasma established his data-crunching cred by helping Barack Obama’s 2008 campaign capitalize on the interests of the voting public. With Strategic Telemetry’s input, the campaign was able to craft targeted mailings and other appeals that stressed issues likely to get certain voters out to the polls.
The approach yielded pay dirt, prompting Strasma to expand his reach. “The great thing, as we’ve been working more with non-profits, is they have a data set of past contributors,” Strasma says. “We can give them a profile of their existing donor base: Are those the people most likely to give, or are those the people who happen to have been asked?”
By scrutinizing the characteristics (age, affiliations, income level) of people who have already given, Strasma can assemble a portrait of others like them who would be apt to respond to the right appeal.
The results are not always obvious, Strasma says. “You’ll come up with things that are completely counterintuitive.” Arts organizations, for instance, tend to overlook less wealthy potential donors. “They think [donors] are the people who come to opening night in their limousine, but a lot of people might be willing to give $20 to support a local theater company.”
Strasma also asks organizations for lists of people who have been asked to give, but have declined. That can give non-profit execs an idea of what types of prospective donors are least likely to give in the future, allowing them to refocus outreach efforts on more promising prospects. (If you donated once as part of a company fundraising drive, say, you might not be a great repeat prospect over the long term.) Strasma’s projects are still mostly under the radar, but he reports that an umbrella for New York City non-profits boosted fundraising by 10 percent across the board after adopting his statistical models.
When donor records are overlaid with the kind of detailed information social networks supply, predictive models about who will give, and when and how, can become almost scarily accurate. That’s what Dan Mansoor is counting on, anyway. GiveNext.com, Mansoor’s Cleveland-based startup, allows people to manage all their giving on a single online platform — users comb a database, donate whatever they like to any of 1 million-plus non-profits, and receive a tax form in return. The company charges a small transaction fee for each donation it facilitates.
Mansoor wants GiveNext to generate a wealth of donor data — data, he says, that charitable organizations can use to fuel their own giving campaigns. Within a few months, users should be able to sign up for GiveNext through Facebook or LinkedIn, Mansoor says, and “once they do that, we’ve got a tremendous amount of data.” If you opted to enroll through Facebook, for instance, GiveNext could potentially learn your “likes” and interests (the Seattle Seahawks, volunteer tourism)and other basic information about you. With data from social networks and the donations it collects in-house, the organization plans to carry out a focused analysis of likely giving behavior.
GiveNext will be able to figure out things like whether frequent travelers are more likely to give to overseas charities, or at what age people are most likely to leave money to a charity in their will.
The organization can then share that information with non-profits in its network, eliminating some of the legwork each charity would otherwise have to do on its own.If the data show college humanities graduates between age 45 and 65 are especially likely to give to local environmental concerns, nature-focused non-profits could focus their appeals on that promising pool rather than mounting a wide-ranging campaign with a low response rate.
“To get people to dramatically increase their giving, you need to know their interests, values and circumstances,” Mansoor says. He likens the relationship between donor and non-profit to a courtship dance. “Before you’re ready to ask the ‘big question’ of someone, you have to know them well enough to be reasonably sure what their answer will be.”
Mansoor also sees potential for technology to help cement donor loyalty. “Today when you make a gift, you get a letter in the mail a few weeks later,” he says. “Why shouldn’t you get a video on your phone from the president of the college?”
If you’re a donor, that kind of instant, personalized feedback could reinforce your sense that you’re making a difference, activating the “helper’s high” and making you feel good about opening your wallet the next time around.
Platforms like Mansoor’s aim to give analytics-challenged non-profits a painless introduction to donor tracking, yet more and more charities are figuring out that data simply isn’t enough on its own. In addition to amassing it, organizations have to learn how to make sense of it. “Data always requires interpretation — all the data in the world is of no use if you don’t have people who understand the limitations of the data and come to the right decisions,” says Kentaro Toyama, a University of Michigan computer scientist and board chairman of the non-profit Digital Green. Deciding what to do with the reams of data can be a sticking point for many non-profits. “I think there’s a bit of paralysis there,” says Weiner, the Whole Whale founder. “It’s like drinking from a firehose.” In fact, a handful of firms, such as Bellevue, WA-based Delve Analytics, have sprung up to help non-profits figure out how to sift through the various data streams they’re collecting.
When data streams are channeled efficiently, though, the results are often dramatic. California-based KCET, a large non-profit public TV station, was having trouble attracting repeat donors, so station management asked the Russ Reid data-consulting agency to craft a new outreach strategy. Combing through existing donor data, Russ Reid analysts discovered that the station members most apt to become regular donors were those who’d joined or re-joined recently, confirming their commitment to the station. By directing repeat-donor communications to this group of likely givers, KCET quickly tripled donor response rates—and, most importantly, doubled the number of donors on file making automatic monthly gifts.
For a station that wants to focus less on money matters than on firing people’s imaginations with shows about poetic activism and restoring kelp forests,
that’s a serious windfall.
Despite success stories like these, many executives and boards have long shied away from investing in data-driven outreach, even if the benefits would ultimately trump the initial costs. It’s easy to understand why keeping the lights on might take precedence over building sophisticated donor targeting schemes. But cash-strapped non-profits have more data diving opportunities to consider these days. There’s DataKind, which matches interested non-profits with data scientists willing to offer their services gratis. Some data experts are willing to work on performance-incentive contracts, getting paid only when data crunching benefits a non-profit’s bottom line. And data-analysis software, once the sole province of for-profit behemoths, can now be had for a reasonable price or even for free. “As technology has gotten cheaper and cheaper,” Strasma says, “the cost of entry is much lower than it used to be.”
Once the practicality hurdle is cleared, non-profits face delicate questions about how to preserve donors’ and beneficiaries’ privacy. Companies that use personal data for their own internal, inscrutable ends tend to generate suspicion. But when non-profits use data like Facebook profile information to help promote a good cause, the moral equation gets a bit more complicated: You might be super-enthusiastic about a charity, but you also might not realize the charity is studying your personal details and interests.
In general, Weiner says, non-profits need to be completely transparent about how they use donor and participant information — which includes posting clear privacy policies on official websites. “Err dramatically on the side of privacy,” Strasma says. “If you ask for someone’s email, be very clear about what you’re going to use it for.” Mansoor stresses that when GiveNext shares information with organizations that are part of its network, it will do so only collectively, based on statistics it compiles about the aggregate pool of donors. So while it might tell Harvard, for example, that another of its donors’ favorite causes is youth enrichment programs, it wouldn’t tell the university which specific donors gave to youth non-profits or how much.
For non-profits willing to brave data diving’s initial challenges, considerable reward awaits: a flood of enthusiastic participants who’ve been looking for
an organization just like them.
Take The Lamp, which teaches New Yorkers digital literacy skills and hired Whole Whale to help maximize its impact. After Weiner and his colleagues studied data on how people interacted with The Lamp’s website and constructed site mock-ups for testing, they were able to get a better idea of which features most engaged users: Teachers were intrigued when the site clearly showed where they could get a free lesson plan for their students. Knowledge like this helped the team make site design tweaks, like a lesson sign-up box at the top of the page, that doubled the rate of e-mail registrations. The upshot? Students looking to build their computer skills will be more apt to benefit from the curriculum.
It’s in understanding just what makes people tick while respecting their autonomy, Weiner believes, that non-profits can best realize their potential as forces for positive change. In the process, they could help more of us tap into the deep well of our own generosity. (Fewer superfluous address labels would be a nice side bonus.) The key, Weiner and other experts say, is data. “The challenges domestically, globally, are so daunting. To reach the levels we need, there needs to be a multiplier.”
