This post seeks to expand on a recent successful review of the speed at which the Guardian was able to sort through the 2,000,000 pages of documents surrounding the recent UK expense scandal using crowdsourcing. While this project was made easy for the Guardian with a circulation of around 340,000 to call for volunteers, it was still a large task for any media company and to get it completed in time before their competition.
The Guardian operation in brief took a week from a friendly software developer, one day of help from his department of 15 staff, a few other contacts such as designers, system admins, and $AUD100 for some temporary web servers. This is not typically for many commercial companies who try to utilise crowd sourcing.
What made the Guardian crowd sourcing project successful was an understanding that they had to make the process fun, and more like a game otherwise they may struggle to complete it on time. This game interface was accomplished through a simple 4 panel interface, and a community progress bar that helped shared the end goal of 457,153 pages reviewed. They also understood that competitive nature would be increased with a scoring system that showed top users by line items reviewed. The last point discovered during the process that the addition of MP mug shots would increase the overall participation rates and increase the personal element.
The point was known that the project would require stability so the platform was built around an open-source and scalable solution Django. The biggest issue that the Guardian was working to a tight deadline and with the interest in this story they had to do a lot of the process without testing which is not usually best case for crowd sourced solutions. If items break or are not reliable they can lead to dissatisfaction and increase drop out rates and the project might not be completed on time.
So what other solutions use variations of crowd sourcing to successfully grow their business, brand and increase potential profits? One new example is Breaking News which is distributed to 620,000 twitter followers who seed niche stories quickly to other followers. While this seems to be more wisdom of crowds, rather than crowd sourcing? According to Wikipedia it may qualify as crowd sourcing as the task or distributing PR using their twitter followers who then distribute important/relevant stories via RT to their followers is a open call.
There are dozens of online examples using crowd sourcing such as Seti@Home, NetFlix Rating System, Wikipedia, Mechanical Turk. Other more niche projects using for Crowd sourcing have included a Question/Answer forum for Famous People, these
Where Crowdsourcing fails in IT made headlines in venture capital circles last year when Cambrian House failed after $7.75 million was poured into the idea. Dell has made good use of Crowd Sourcing with its “IdeaStorm” center built on a SalesForce.com platform, but did take a steep learning curve with niche groups hijacking the platform around certain products but gave great insight into their target audience.
There has been a successful online charity Kiva typically focused on Africa, Asia & Eastern Europe has expanded its microloan products to the working poor into the USA. Who allow lenders to join teams such as The Lost Agency Trust to increase their potential lending and make a difference for only $25.
There are an increasing number of FMCG companies who are struggling to continue to generate high profits with existing products and need new fresh ideas but may not have the time/budget to risk a failure and have turn to the crowd. A recent competition in Australia asks for you the crowd to do them a flavour and come up with the next Smith’s chip great flavour and get 1% of all its sales plus $30,000. On a much larger scale Kraft Foods has moved closer towards mass customisation of their products with their new website help innovate with Kraft.
Crowd sourcing maybe here to stay, but it can require a lot of hard work as shown by the successful Guardian experiement or large resources and advertising budgets as demonstrated by FMCG groups. Would crowd sourcing work for every project, most likely not but it can be useful if you have hit a roadblock or have limitations on budgets or time.
Good thoughts, David. I’d also argue that crowdsourcing works best when many people are pre-motivated to contribute, even before they hit the game-like interface. Sort of a “many eyes” phenomenon.
Picking such moments is probably difficult.
Michael, thank you and yes picking such a moment can be very difficult, i think crowd sourcing can often be also called panic sourcing.
Hey David,
Great post. To clarify – Cambrian House isn’t in the deadpool – we have a technology platform, called Chaordix, that enables companies to integrate crowdsourcing into their businesses.
There’s other really cool examples as well, such as Challenge Post (just launched yesterday), Crowd Flower, Kluster and Tongal.
The most important thing to remember about crowdsourcing is that it is only 50% about the technology, if you are involving any crowd, you need to give them the right purpose, the right call to action and the right incentives.
Keep on with the great blogging,
Blue
Good post,
I think crowdsourcing can have a huge impact on how all sorts of businesses can be conducted. There of course has to be proper incentives in place for the “crowd” for most implementations to be successful. I recently took a couple looks at how crowdsourcing could be applied to the venture capital model:
Crowdsourcing Venture
Another Take on Crowdsourcing Venture
There’s definitley some promise there, save for regulatory hurdles (which can be overcome I think).
Could anyone tell what did the experiment of The Guardian produced in terms of concrete data/information/revelation?
Good question, Roberto. We’re working on a follow-up at NiemanLab — look for it around Oct. 1 or so.