Milestone 3 TestSet TrustIdea 1: PhotoAndBio
Everyone knows crowdworkers are not automated robots, ready to spam requestors and perform tasks with little cognitive thought. But it's difficult to feel crowdworkers' humanness when all you see is a interface with text, buttons, links, and no human face or story. This anonmity makes it easier for requestors to disregard workers as people who are willing to offer their skills to them, and harder for them to recognize that rejecting work for no reason, for example, can result in actual human feelings (like dejectedness), and even be unethical.
One of the main points of "Design notes for a future crowd work market" was that it's vital to present "workers as people, not computation." One way we thought we could communicate this to both workers and requestors is to create bio profiles for both workers and requestors that present workers as contract employees and requestors as managers. The end-goal is to replicate the feeling of a real office -- a real workplace -- that can bring human-to-human relationships to this crowdsourcing platform.
Rather than a standard bio profile with a name and photo, however, these bio profiles will include ratings of 1 to 5 that will appear under their name, much like Yelp. Below that will be the amount of how many ratings the worker/requestor received. Ratings will be mandatory for both parties.
"Stanford Crowdsource, of course, is a fake name of the crowdsourcing platform. Below a search menu and primary menu are profiles of a worker or requestor. This visual is an example of a worker profile, but requestor profiles will follow the same visual format. In the performance box, however, requestor profiles will include rejection rate and response rate in place of "jobs done" and "requestor satisfaction rate."
Worker Profiles will include:
1) A photo 2) Uploaded credentials (degrees, certificates) that will show as verified after a moderator reviews them 3) A list of their chosen skills, much like how you can list your skills on LinkedIn and people can endorse them. Some examples are copywriting, content writing, and language fluency for translations 4) A big, visible message button for easy communication 5) A box full of statistics: how many jobs this worker has done; requestor/customer satisfaction rate 6) A classification label identifying them as either a part-time crowdworker who uses this platform for extra cash or a full-time crowdworker trying to make ends meet 7) And of course, a bio section with a 100-150 word limit where the crowdworker can tell their story: who are they and what their specialties are
Requestor Profiles will include:
1) A photo 2) A short description of the kind of work they offer 3) A big, visible message button 4) A box full of statistics: the percentage of their response rate; the percentage of their rejection; average amount paid per project 5) Their job listings (this will only list a couple, and there will be a "View More" link that will show the rest when clicked
Designing a crowdsourcing platform that replicates a real workplace requires ease of communication and the ability to develop professional relationships. In a physical workplace, for example, employees build relationships with their managers and both parties earn trust from each other. If someone needs to ask a question, they can easily knock on their office door and do so. The problem with AMT is that worker-requestor relationships are more like one-night stands: they come and go and never last. Lack of a messaging system and, even, an incentive to message the other party can build walls instead of bridges -- that is, the trust we need to establish in order to create a truly sustainable crowdsourcing economy. The ability for requestors to see who workers are (and vice versa) -- at least with a thoughtful profile page -- can make building professional relationships through cyberspace more possible. Eventually, workers may become regulars for certain requestors. Furthermore, reputations for both parties can develop this way.