News

By Emma Ferneyhough, November 5, 2013

Trends in Crowd Labor | CrowdConf Recap

This year at CrowdConf there were a number of scholars presenting both from the crowd contributor as well as work provider points of view. Vili Lehdonvirta, Ph.D., from the Oxford Internet Institute at University of Oxford, and Tim Olsen, Ph.D., from W.P. Carey School of Business at Arizona State University kicked off the Research & Trends series of talks at the conference with their findings on the people who do online tasks for a living as well as the companies who provide the tasks.

What is Life Like in a Freelance Society?
Lehdonvirta’s research was spurred by his observation that while economies may benefit from an increase in more fluid and temporary work, the overall impact of this shift in work on society has yet to be determined.

His research focused on three groups of workers in three different countries; Mechanical Turk in the U.S., MobileWorks in the Philippines and CloudFactory in Nepal. He interviewed workers, observed workers at their work sites, and on online forums. He found that of the three channels, Mechanical Turk workers expressed having the most challenges. Specifically, they are more isolated and the tasks they work on, as well as their wages, are more arbitrary. In contrast, MobileWorks and CloudFactory workers perform more project-based work, there is a greater sense of community via worker forums and teams, and wages are more consistent because they are set by the platform, which does all wage negotiation with clients on their behalf.

More Choice, More Challenge
What I found most interesting about Lehdonvirta’s talk was his finding that as workers have more freedom of choice, they experience more challenges to their self-identity. In other words, the more they are able to select what work they want to do and how they want to do it, they have to develop their own structures and self-control techniques in order to have a similar sense of purpose, continuity and stability that people have working at larger institutions.

These structures can be both social and technical. On the social side, workers develop online peer groups where progress and earnings goals are shared, and where they alert others about the availability of work. On the technical side, they have developed solutions for self-control, such as custom browser plugins that reduce distraction and track earnings.

Overall, channels such as MobileWorks and CloudFactory provide more of this structure and advocate for their workers, whereas self-described “Turkers” have developed similar techniques on their own.

Managing the Human Cloud
Olsen’s presentation provided a birds-eye view of four different platform models that each have a different way to build trust and provide governance of crowd workers.

  • Facilitator – It connects suppliers and buyer directly through a bidding process, e.g., Upwork and Elance.
  • Arbitrator – It engages multiple suppliers through competitions, e.g., 99Designs and crowdSPRING.
  • Aggregator – It aggregates hundreds or thousands of microtasks performed by multiple suppliers, e.g., Mechanical Turk and CrowdFlower.
  • Governor – It provides project governance and certifies supplier quality, e.g., TopCoder and uTest.

Each of these platform models have different typical use cases and benefits. They also vary on a number of key metrics such as quality control, project terms and pricing, and supplier expertise among many others. Needless to say, there is likely a crowdsourcing platform model among these that can do work for you more efficiently than going a traditional route!

Crowdsourcing Industry Report 2013

In addition to the platform model overview, Olsen gave a rundown of the crowdsourcing industry report for 2013. Compared to last year, the number of crowdsourcing workers has been growing fast across three main categories: expertise-based, ideation, and microtasks. The proportion of industries using crowdsourcing in the technology services sector has taken the lead over media and entertainment. In terms of worker education, those in the expertise-based category are more likely to have a college or professional degree than those in the microtask category, whereas those in the ideation category are much more likely to have a scientific degree, e.g., a Ph.D., than those in the expertise-based category. Overall, the top five workers in the microtasking category make the least amount of money at an average of $7,700 per year whereas the top five in the expertise-based category make the most at an average of $210,000 per year.

Further Reading

What Cloud Labor Teaches Us About the Future of Work and Society by Vili Lehdonvirta, Ph.D.

Managing the Human Cloud by Tim Olsen