When you think about crowdsourced work, you probably think about Uber or equivalent. On the one hand, an anonymous platform where people feel like mechanical parts. On the other hand, a face-to-face platform where people are rapidly obsolescing and doomed to be replaced by robots. We call this the “Turk Myth,” that crowdsourced work is only done by people because we haven’t built good enough robots yet. The Turk Myth is unfortunate because it creates a fatalistic spiral: if people think of crowdsourced work as exclusively low-skill, they demand less of crowdsourcing platforms and the workers they employ; this in turn leads to an underinvestment in the training and management of those workers. Thus, the Turk Myth becomes the commodified reality for many.
Like many myths, this one has a basis in fact. There’s a lot of clickwork in the gig economy that’s neither well-paid nor fulfilling. But there’s also a crowd economy that rewards human agency and creativity. If you’re thinking of using crowdsourcing to move your business forward, make sure the Turk Myth doesn’t “crowd out” the more complex (and more hopeful) reality. You can bend the curve toward more cognitively and financially rewarding work, and get great results in the bargain. Here are three things to think about if you want to maximize your return on crowd intelligence.
Look for faces in the crowd
If, as the Turk Myth suggests, people are anonymous parts performing low-skill tasks, it doesn’t pay to spend time helping them improve. But just as with other kinds of work, in crowdwork training, pride and positive reinforcement improve results. Because crowdwork is more ad hoc and distributed than traditional labor, much of the management happens through a software platform, but it’s still management, and it demands a specific set of skills and practices. Most importantly, a continuous feedback loop connecting the end-customer with the crowdworker fosters both quality and job satisfaction. The customer is the ultimate judge of quality, so your input matters. Make sure you communicate expectations directly to the workers, and engage with them directly when necessary.
A crowdsourcing platform should relieve you of the burden of convening the crowd, parceling out their tasks and assigning work efficiently, but if it makes the workers invisible, you can’t solicit additional insights that create value for you. In software testing, for example, crowdtesters who discover defects in the software can also collaborate with you to help diagnose them. Crowdworkers take pride in their work, and crowd platforms that capitalize on this encourage collaboration.
Understand the economy of the crowd
Crowdworkers need to earn money. They also seek projects that are interesting and advance their careers, and prefer conditions that are fair and transparent. If you’re employing a crowd workforce, this interplay of financial and personal incentives determines whether you’re getting the best possible work for your money. You’ll get the best results if you understand the crowd economy. Passengers detest Uber’s surge price, but it transparently reinforces an economic lesson: financial incentives matter. When the number of people who can perform a task is limited by geography, skill requirements or interest, a crowd platform must create liquidity, typically by changing the incentives. If you engage with a crowdwork platform, make sure you understand how these incentives work — both for the people doing the work and the people hiring them. A crowdsourcing platform that doesn’t flexibly create incentives for people to meet demand is unlikely to command the loyalty of the best workers, and headline numbers of workers at your disposal are meaningless if the task requires skill and the incentives are not effective.
It’s time for a new discipline – crowd work management. Despite a buzzwordy obsession with ninjas and gurus, we also see this happening in social-media management. The same thing will happen with crowdwork management. Managing crowdworkers effectively requires a combination of empathy, technology and business sense. If your business — or your crowdwork platform — thinks of them as purely interchangeable and their labor as a commodity, you’ll get commodity results. That’s the danger of the Mechanical Turk myth: It’s a mutual race to the bottom for employers and crowdworkers. We can all do better. And we should.