Abstract
Crowdsourcing is a practice employed by many types of organizations that makes use of the internet to harness the time, energy, and talents of individuals who are otherwise unconnected to the organization. Hana Shepherd explores some of the many creative uses of crowdsourcing and discuss the issues the practice raises in terms of the nature and future of work.
When was the last time that you thought to yourself, “Why, I think I’ll categorize some galaxies, design a logo, or create a campaign video today?” Unless you are employed as an astrophysicist/graphic designer/political strategist, chances are slight that such a thought crossed your mind. We often assume that certain tasks require a degree of training, skill, or experience for successful execution. The now-prevalent phenomenon of crowdsourcing upends that belief.
In crowdsourcing, a community or “crowd”—often organized online—is invited to participate in tasks typically completed by organizations or by specialists. A call to a crowd might seek to generate a large volume of content, or to lessen the burden of a task by distributing it among many people. Those using crowdsourcing techniques assume that putting a problem to a large group increases the likelihood of generating an important innovation.
Putting a problem to a large group may increase the likelihood of generating an important innovation.
The World Wide Web is a crucial component of crowdsourcing. It is the most common way organizations make contact with the crowd—people who could feasibly contribute an idea, content, or time. It also provides the tools through which the crowd can perform its tasks. For example, scientists distribute pictures of galaxies to users who will then categorize them. Sometimes, the web provides the task itself, as with websites that ask users to label content (often called “tagging”) to create metadata—information about online content.
Jeff Howe, now a journalism professor at Northeastern University, coined the term in his 2006 article for Wired magazine. Howe wrote about organizations that were adopting a new model for accomplishing tasks through what he termed crowdsourcing. Howe’s pun on the word outsourcing—the practice of moving jobs abroad to reduce labor costs—implied that crowdsourcing provided a new way for organizations to reduce their costs by harnessing cheap labor. Instead of formally hiring and employing workers at a much-reduced rate, as in outsourcing, crowdsourcing often requires no formal relationship between an organization and those who generate content or complete tasks, though some organizations may provide compensation or incentives for participation.
Howe documented how people generate photographs, television content, and scientific ideas outside of organizations that were employing them to undertake such tasks. Today, crowdsourcing is commonplace. In apparel, the company Threadless collects t-shirt designs from whomever wishes to submit a design, and then produces and sells the t-shirts. Since 2007, the snack company Frito-Lay has solicited Super Bowl commercials from consumers, and has asked online users to vote on their favorite commercial. The movie provider Netflix held an open competition to develop an improved algorithm for predicting users’ ratings of movies.
The website Galaxy Zoo provides users with Hubble Space Telescope pictures and asks them to categorize the galaxies they see in the pictures.
Crowdsourcing has been creatively used for big projects as well. Computer scientists at Carnegie Mellon University created reCAPTCHA, which uses humans to identify distorted words on websites. The distorted words have been culled from scanned books and digital archives that computers, using optical character recognition, cannot read. Thus, as we type two distorted words into a website field, identifying ourselves as humans and not bots, we contribute to digitizing books and media. Google Books and the New York Times, among many others, use this service for digitizing text.
Commerce is not the only arena that has enthusiastically adopted crowdsourcing. The non-profit investigative newsroom Pro-Publica developed a “Distributed Reporting Project” that organized non-journalists—“citizen journalists”—to collect information, follow news story leads, and develop sources. A number of websites including Wikinews, All Voices, and Now Public feature news stories collected and aggregated by users. In 2002, the Cornell Ornithology lab and the Audubon Society launched the website eBird, which collects and records bird sightings from amateur participants to provide data to scientists and amateurs about the geographical distribution of different bird populations. The website Galaxy Zoo provides users with Hubble Space Telescope pictures and asks them to categorize the galaxies they see in the pictures. Finally, organizations have adapted crowdsourcing techniques to further the democratic process. All Our Ideas software helps governance organizations solicit suggestions and preferences, asking users to compare the importance of different pairs of issues. It then aggregates the results for the community of users.
Corey Fields
Organizations often benefit, financially or otherwise, from these user-generated contributions; some users who submit ideas or spend time on tasks are compensated while others are not. The website Mechanical Turk pays users piece-meal for their work. Sometimes contributors are compensated only if their ideas or suggestions win or reach some threshold of success, for example, Netflix provided an award of 1 million dollars for the improved algorithm. Some of the scientific papers published using the categorizations provided by amateurs on Galaxy Zoo include these amateur users as authors (indeed, the founder of Galaxy Zoo, astrophysicist Kevin Schawinski, prefers the term “citizen science” to “crowdsourcing” because the former implies more active engagement from amateurs). Other content created through crowdsourcing remains unpaid and unrecognized.
Crowdsourcing is an innovative way of structuring certain tasks that harnesses the power of many people towards multiple possible ends. Organizations and researchers benefit from crowdsourcing because it cuts down on organizational costs by capturing the talent and time of others. Contributors can also benefit, as they exercise their amateur skills, or aid and interact with professionals while practicing their hobby.
Should crowdsourced content be considered labor? What compensation or intellectual credit should a participant expect? Will crowdsourcing challenge the boundaries professionals draw between themselves and practitioners who lack recognized credentials? Can we consider crowdsourced content (about, for example, tags on a website, or suggestions for political action) an ideal reflection of the preferences of the group? Or do crowdsourcing systems systematically favor some results over others? Given its potential benefits and drawbacks, crowdsourcing is a practice to watch.