With its name recognition and high volume of Web traffic, Wikipedia is uniquely positioned to raise money online.
People around the world use the free, crowdsourced encyclopedia, which has become an almost ubiquitous resource—sometimes to the chagrin of schoolteachers.
There’s no secret tactic that helps the nonprofit Wikimedia Foundation raise millions of dollars each year, says Lisa Gruwell, its chief revenue officer. The organization just asks for feedback from its global readers, and uses their thoughts to craft simple messages asking for donations.
“It’s a straightforward approach,” she says. “We don’t use any gimmicks. We just tell it to them straight and they respond, and that’s true in every country in the world.”
But that straightforward approach is honed through extensive A/B testing, as explained by a recent Wikimedia Foundation report and online fundraising presentation at the annual Wikimedia Foundation conference.
User Feedback
Wikipedia hosts pages in 288 languages, and according to Ms. Gruwell, readers around the world like Wikipedia for the same reasons as do readers in the United States. That’s not a guess; it’s one of the findings from the organization’s user testing.
The foundation solicits feedback in three ways: user surveys, talking to focus groups around the world, and A/B testing on its pages.
The nonprofit’s website asks all donors for their comments about the donation experience, and the foundation takes that qualitative feedback into account when planning future drives. The organization uses cookies (personal information stored by users’ Web browsers) to track who has contributed to Wikipedia, so that donors are not prompted to give again.
The in-person focus groups are informal monthly gatherings of people who have donated. Invitees are drawn from the foundation’s list of about 7 million donors who live in nearly every country, and they gather in places like coffee shops to discuss Wikipedia in their native language.
“We have a highly multilingual staff. We don’t have every language covered, but pretty close to it,” Ms. Gruwell says. “When we don’t, we work with a consultant on the ground when we’re running that campaign.”
A/B Testing
Each year, Wikipedia conducts hundreds of A/B tests, which pit different banners against each other to assess their performance, live on its pages for about an hour at a time. The tests provide quantitative data—about payment options, text, visuals, and preset donation amounts—that is complemented by survey comments.
By changing one variable at a time during A/B tests, the foundation incrementally crafts more effective banner appeals. In 2009, it tested text-only banners and discovered that the most effective linked to a personal appeal from Wikipedia founder Jimmy Wales, who outlined in a letter what distinguishes Wikipedia. It performed 20 times better than the next most effective message, so the foundation adopted it for future use. In 2010, the foundation found that adding a photo of Mr. Wales increased donations significantly, but that it didn’t matter much which photo was used.
Realizing that most people don’t click on the banner to reach Mr. Wales’ letter, in 2012 the foundation experimented with putting key phrases from the letter directly into the banner text, highlighting the facts that Wikipedia is a nonprofit, accepts no government funding, and doesn’t run advertisements. It proved three times more effective than previous banners.
During testing in 2013, some A/B testers commented that they had intended to donate but forgot to do so by the time they scrolled through a Wikipedia entry. The fundraising team added a side tab that moves with the user as he or she scrolled through articles, and it increased the number of donations by 15 percent. When they added language suggesting a donation amount—“If we all gave $3, the fundraiser would be over in an hour”—to the scrolling tab, it brought in 40 percent more individual donations and 20 percent more money. This tactic reduced the average donation size, but brought in more money over all.
When the team tested banners against each other, the best-performing banner of 2013—which highlighted key parts of the text, suggested donation amounts, and equated Wikipedia to a public service provider like a library or a public park—garnered 92 percent more donations than the best-performing banner of 2012.
With mobile donations on the rise, the foundation also runs A/B tests for its mobile fundraising campaigns. The small size of cellphone screens challenges the fundraising team to distill its appeals down to their essential components in order to fit the most relevant information. The team found that adding the phrase that suggests people donate $3 is very effective in encouraging mobile donations. So is adding a high preset donation amount—$100—which the team attributes to the anchoring effect, a psychological phenomenon that alters individuals’ perceptions of what seems reasonable by making them consider a higher suggestion.
The team found that using a single appearance of a mobile banner that fills the entire phone screen increased donations by 250 percent compared with showing a small mobile banner every time a user visits a Wikipedia page.
Most of the feedback from surveys, in-person focus groups, and A/B testing suggests that donors appreciate campaigns that are clear, transparent, and simple.
“What compels them to give is knowing we’re a nonprofit and knowing how the money is spent and how it benefits the world,” says Ms. Gruwell. “The best approach is one that’s really straightforward.”
By the numbers
Amount raised in the December 2012 campaign: $25-million from approximately 1.2 million donors
Amount raised in the December 2014 campaign: more than $20-million from more than 2.5 million donors
Average gift: roughly $15
Online fundraising total for the 2013-14 fiscal year: $37-million from more than 2.5 million donors
Portion of online gifts donated via mobile in 2013: 5 percent
Estimated portion of online gifts donated via mobile in 2014: 25 percent