Crowdsourcing Project – Summary of Best Practices


A few weeks ago a group of 15 students from all around the world embarked on a mission to create a new program around Mozilla Labs’ Design Challenges.

The group split up into three teams with each team focussing on one particular aspect of analysis: Past Design Challenges, best practices and existing research.

The write-up and summary below is a guest post from Jimmy Chion for group 2 – Best Practices:

The following are several main aspects that are characteristic of successful crowdsourcing sites/businesses and examples of businesses that employ the characteristic well. Included are a few other metrics that our group discussed.

Low Barrier to Entry

Easy access to contributing to the crowdsourced content is almost necessary for increasing participation among first-time users. Although a registered account gives more commitment and community for an individual to come back, it is a small hurdle that prevents many from using a site. One solution to this that we’ve seen in sign-in through accounts many people already/probably have, such as Facebook.

Wikipedia – anyone can edit almost anything. In fact, the biggest barrier to entry for Wikipedia is its editing tools which many have criticized as un-user-friendly.
Kickstarter – Facebook sign-in and easy donation process encouraging small amounts of money.
Hunch – once you go onto the site, you are already contributing by answering 20 questions and developing your profile. It asks you to register to save the information. Tricky.

Business Incentive: Competition for top submission

Some of the best-known crowdsourcing businesses use many professionals to produce many solutions for a challenge. The submitter of the challenge picks the best one and pays only for that one. The business putting out a call for submissions benefits from the intense competition between the contestants. The best known sites of this kind are Innocentive and Crowdspring.

User Incentive: Intrinsic motivation

Apart from the external motivators (such as money, points, gifts), the crowdsourcer must implicitly appeal to the internal motivations of the participants to participate. This can range from recognition within a community, desire to share knowledge and opinions, to wanting to develop personal talents. The most successful sites create intrinsic motivation for the user to contribute to the crowd. People come back for the points, rewards, reputation, but the biggest motivator is people’s own motivations of desire to share (possibly show off).

Wikipedia – seems like pure altruism, but is successful because it empowers the user to contribute something that they may be an expert on to many many people.
Ushahidi – crisis mapping. Aggregates knowledge (via sms, web) of where violence was occuring. Made into open source and used in Haiti and Washington, DC for snow fall. There is no other value to the contributers other than aggregating knowledge for the greater good.

Give people the ability to shine

This is usually done by diversifying the different tasks to be crowdsourced. By letting people contribute parts of the whole, the task is less daunting, and it also allows people to shine in what they’re good at. This point is also related to the previous one.

OpenIDEO – breaks up the creative process and awards points at different phases in the project.
Quirky – similar to OpenIDEO, rewards proportionately to whatever parts of your contributions are successful
Wikipedia – probably the best example where people are given their ability to share whatever unique knowledge they have.

The following are aspects that are great but not entirely necessary for a successful crowdsourcing project


The smaller the niche and the bigger the community, the more identified the user will feel. Wikipedia, Yahoo Answers, and InnoCentive are great counterexamples though. The former two aim to be all-encompassing and general as can be, and the latter pretty much has no community (though the company has talked about nurturing it).

Threadless – has a tight community of artists and tshirt consumers. The entire community helps each other refine designs and choose the best ones.
Quora – similar to Yahoo Answers, but for young silicon valley startup entrepreneurs. As a result, the community respects each other and the quality of answers are extremely high unlike Yahoo Answers and Facebook Questions.
OpenIDEO – collaboration is two of the four phases in its process.

Some component of fun

Adding a playful or fun aspect to the whole experience gives more reason for users to participate and come back.

Google Image Labeler – Users earn points for trying to label an image the same label as what other people would label it as. Turned something entirely boring into a game.
Hunch – A decision engine driven by user surveys. Who doesn’t love filling out surveys about themselves in order to find out more information about themselves they supposedly don’t already know?

Our compiled list of crowdsourcing initiatives can be found at here