In June 2017, Time, themagazine known for its person of the year, released a list of the 25 most influential people on the planet. The newly-crowned Donald Trump was there, as were Matt Drudge, J.K. Rowling, and K-Pop band BTS. Also on the list, though, was Ser Amantio di Nicolao—not the minor character in the Puccini opera Gianni Schicchi but the pseudonym of Virginian Steven Pruitt, a records and information officer for the US Customs and Border Protection. Pruitt is also the world’s most prolific contributor to English Wikipedia. According to the site itself, Pruitt has created more than 35,000 articles and made more than three million edits. He is said to spend more than three hours every day poring over books and reading through academic journals to research, write and edit the Internet’s encyclopedia.
He doesn’t get paid for that work. Pruitt is a volunteer, as are almost all of the people who create Wikipedia. The Wikimedia Foundation, a non-profit that manages the platform, employs some 354 staff and contractors, but the site is created by the nearly 40 million people who have registered a Wikipedia account. Of those, a little under 150,000 edit and create new content every month for free. And of those near-150,000 volunteers, just a tiny fraction contribute the bulk of the new content. A 2017 study found that just 1 percent of Wikipedia’s editors have created about 77 percent of the site’s content. It’s people like Steven Pruitt who keep Wikipedia alive.
And yet despite that gap between the theory of a platform created by millions of members of the public, each an expert in their own field, and the reality of about 1,300 people creating about three-quarters of the new articles posted daily, Wikipedia has been remarkably successful. It’s one of world’s most popular websites, beating Netflix, Office.com, and even Pornhub. It’s the first stop for children and students the world over forced to write homework assignments. It’s living proof that when you create the right infrastructure, the right projects, and the right incentives, it is possible to hand over responsibility to a community and leave it to users themselves to create the product.
Wikipedia relies on the principle that while mass, open contributions will produce poor content, editing and peer review will soon weed out false or misleading information, ensuring that the quality of content overall continues to rise. Editors have their own watchlists, a list of articles that they’ve created or that they monitor. Some use automated software to alert them when one of those pages has been changed. It takes “only a few minutes” for an editor to notice and act on a bad or harmful edit. There is no hierarchy of editors, as there is in a publishing company, but rather an assumption of community goodness to ensure the quality of the content.
For that assumption to be true though, editors and contributors like Steven Pruitt need a reason to dedicate their unpaid hours. Researchers have identified six key motivations: altruistic values; social engagement; the desire to increase their own understanding; career progress; a wish to protect themselves such as by allaying guilt for their personal privilege; and a desire to show off their own knowledge. They may also contribute to Wikipedia if they believe in the ideas underpinning the project, and if they enjoy it. Of those motivations, “fun,” “ideology,” and altruistic “values” were the most common for Wikipedians.
But are the practices and motivations pioneered at Wikipedia unique or can they be adopted by other platforms? Most importantly, can they be used by for-profit businesses?
How Duolingo’s Users Gave Investors a $700 Million Business
Duolingo was always meant to do two things at the same time. On the one hand, it would teach students a foreign language. On the other hand, it would charge content platforms like CNN and Buzzfeed a fee to ask its users crowd-translate their content. The translation feature was soon dropped but Duolingo has since morphed into one of the world’s leading online education platforms. It’s used by more than 300 million people who complete over seven billion exercises each month in 38 different languages, including Arabic, Spanish and English, as well as Scottish Gaelic, Navajo, and High Valyrian. Although it’s possible to learn for free, the company has more than a million subscribers who pay to remove ads and download lessons. Co-founded by Luis von Ahn, the inventor of CAPTCHAs, it’s raised more $138 million in venture capital from investors including Ashton Kutcher and Tim Ferriss. In December 2019, the company was valued at $1.5 billion.
With that kind of return on investment, Duolingo’s users might have been wary of contributing to the growth of the product. After all, the financial value of their efforts will go to Duolingo’s investors, not themselves. Wikipedia, at least, is entirely free. It’s funded by a foundation which is dependent on donations. Duolingo is a private company that plans, one day, to generate profits.
And yet, without the donation of time and expertise by volunteers, Duolingo would not have been able to scale at the pace it has done. An interview on CMXHub, a site for community developers, with Kristine Michelsen-Correa, Duolingo’s Head of Community, explained how the company did it.
Duolingo, says Michelsen-Correa, was finding that using only the company’s internal team to develop courses was laborious and took a long time. The process wasn’t scalable, and it was becoming clear that the size of the team meant that there would be languages the platform would never be able to offer.
So Duolingo turned to its users. It created a course-creation engine called the Incubator where “volunteers give life to new Duolingo courses.” Anyone can apply to be a contributor but unlike Wikipedia, the platform isn’t open to everyone. Contributors must be committed and passionate about languages “and helping the world” but they also have to be bilingual. Volunteers don’t need to be a native speaker of both languages but they must be fluent and they should be able to write at the level of a native speaker.
Each new Duolingo course goes through the three phases: creation; beta testing; and launch. In the first phase volunteers translate sentences. They can expect to spend about five hours a week at this stage, participating in the community, and each new course takes about three months before it’s ready for the beta phase. Like Wikipedia, Duolingo tries to take care of as much of the technical aspects of the work as possible so that the volunteers can focus on the content creation. The volunteers then review the translation in the beta phase before accepting feedback and implementing changes once the course is live.
Not all of that work requires the same amount of commitment. On application, Duolingo selects leaders. Those leaders can then choose their own volunteers to help them build a new course. That’s a high level of involvement—the Steven Pruitt level of volunteer content creation. People with less time to offer can review the sentences in the beta phase in order to catch errors. Members of staff mentor the teams of volunteers to ensure quality, and a chat system enables team members to communicate with each other and stay motivated.
Like the volunteers at Wikipedia, Duolingo contributors also have their own areas of interest. For Duolingo, that’s particular language pairs. In addition to a main Duolingo discussion forum, each language group has its own forum on the platform, another way to ensure that volunteers stay in touch and build a supportive community.
Incentivizing Your Community
The most important aspects of keeping volunteers engaged through the three months needed to bring a course to beta—and the additional time needed to move it from beta to launch and re-iteration—are engagement and incentives. To maintain engagement, Duolingo uses cooperation inside the team so that everyone feels part of a group, but the company also encourages competition between the language groups to ensure that volunteers continue to contribute and raise the quality of their courses.
In general though, Duolingo has identified four kinds of incentives that can keep an unpaid army of volunteers building content that the company can sell for a profit.
Recommendations for jobs or college gives the volunteer real value for their efforts without costing the company. Those recommendations will remain valuable as long as the company appears prestigious and maintains a reputation for quality.
Duolingo also recognizes its volunteers’ contributions. Wikipedia users have to register then dig into the site to see who created or edited a Wikipedia page. Choose a language on Duolingo’s desktop version, however, and you can see who contributed to the creation of that course. The courses tend to have up to nine contributors, with one volunteer typically doing about 30 percent of the work. (Language constructor David J. Peterson has contributed 73 percent of the High Valyrian course, but he also created the language for the Game of Thrones television series, as well as Dothraki, and the language used by the Dark Elves in Thor: The Dark World.) That kind of public recognition is a useful way to thank volunteers for their efforts.
Other incentives include the opportunity for volunteers to give back by teaching their own language to others, and the chance to learn about language itself by constructing a course.
CMXHub summarizes the ingredients of a successful user-generated content program as flexible working so that people can contribute at a level that suits them; creating specialized groups that both work together and can compete against other groups; strong guidance; and intrinsic incentives.
From SETI to Social Media
That combination of community-building, flexible working, and non-monetary rewards has clearly worked for Duolingo, just as it’s worked for Wikipedia. It’s not too different to the motivations that have allowed social media sites to reach multi-billion dollar valuations based on the content that its users have created. Facebook, Twitter, and YouTube all depend on the site’s users creating a product that they can then offer to others. Duolingo differs by telling people exactly what to create and by helping them to make it. It also sells a subscription service that asks people to pay for a product that those other people built for nothing. But when users feel that they can create valuable content and that creating that content delivers rewards for them—whether that’s through public recognition or a chance to promote a cause or an interest that they support—they will create it.
Nor is this new. It’s the same incentive that once gave SETI, the Search for Extraterrestrial Intelligence, access to thousands of people’s home computers in return for a sense that volunteers were contributing to something important. (That project has now gone into hibernation but there’s no shortage of other projects to which you can volunteer the services of your computer.)
What all of these ventures have in common is a sense that the contributors are helping to build something greater than themselves. Whether they’re telling the world about some esoteric topic on Wikipedia, helping people to learn their language, or searching for alien life, they’re doing something that they’re willing to do without being paid.
The question, though, is at what point do volunteers start to feel exploited? Like Wikipedia and Duolingo, Huffington Post was also built on the contributions of volunteers. Those writers were happy to provide free content when they felt that they were receiving an audience, a byline, a platform for their views, or a chance to land a paying gig, to raise their consulting fees, or win a publishing contract. But when Arianna Huffington sold the site to AOL for $315 million, a group of writers sued. They lost but the lawsuit showed that a willingness to volunteer has its limits. Wikipedia, a non-profit, won’t reach that limit. Duolingo hasn’t reached it yet but if the company is sold, they might find that the course creators want a cut of the check.
Businesses hoping to recruit a community to create their product should make sure that the product supports a popular cause. They should provide leadership and flexible contribution levels so that everyone can chip in. But if they make serious money, they should also expect a backlash.
Recent Comments