Update: As of April 5, this story has been updated with criticism about Uber’s use of behavioral science.
Companies are tracking and changing workers’ behavior in increasingly subtle ways. With big data and algorithms, workplaces are finding new ways to judge and change behavior, and employees may not even see it happening.
On Sunday, The New York Times revealed how Uber employs hundreds of data and social scientists to dream up new ways to keep drivers working longer hours for Uber.
It’s a case study of what happens when social science enters the workplace. For the ride-hailing giant, treating Uber drivers like they were players in a never-ending video game was lucrative for the company, but perhaps less so for drivers who worked longer hours. Here are some big takeaways.
Algorithms are exploiting our love of goals
Uber scientists designed noncash rewards for drivers that cost the company very little, but would have profound psychological effects on the drivers.
For instance, Uber drivers in 2016 were quitting before they reached 25 rides and became eligible for a signing bonus. The data scientists found that once drivers reached 25 rides, they were much less likely to quit.
To keep new drivers from leaving, the app would encourage drivers with: “You’re almost halfway there, congratulations!” The message outlined a goal that drivers didn’t ask for, but with Uber’s steady prodding, it became a goal drivers would be driven to meet.
We really like money goals
A 1997 study on New York City’s cab drivers found that taxi drivers work “one day at a time.” They set a daily income target and will quit driving when they reach that goal. That’s a behavior that Uber scientists tapped into for the Uber app.
Uber drivers who would try to log off would receive pleading messages like, “You’re $10 away from making $330 in net earnings. Are you sure you want to go offline?” The accompanying graphic would show an engine gauge’s needle that was almost, but not quite hitting a dollar sign. Drivers would read this, be reminded of their income target, and stay logged on just a bit longer.
Daily income targeting is great for Uber. Its effect on drivers is more complicated. Driving requires attention, and long hours lead to fatigue. The income targeting also encourages drivers to drive longer hours to meet a daily goal, even when it’s more efficient to only drive at busier hours. This is how a Florida Uber driver featured in the article could earn a dozen excellent-service and great-conversation badges and still make less than $20,000 in a year: more driving, but fewer fares.
The default is you can always be working
Uber’s app automatically queues up the next ride during a current ride in a practice known as “forward dispatch.” It’s overrides drivers’ decisions with the company’s.
Drivers don’t get to see where their next ride is going, so they can’t estimate how profitable the next ride will be. After drivers complained that the feature made it impossible for them to even use the bathroom, opting out of “forward dispatch” became possible, but it’s still inconvenient for drivers to set up that option.
That’s all part of Uber’s bottom line. “The optimal default we set is that we want you to do as much work as there is to do,” Jonathan Hall said of Uber’s algorithm for drivers.
And yet, a March 2017 working paper found something positive about Uber’s work culture. The researchers found that Uber drivers, who are independent contractors that set their own hours, economically benefited from the hour-to-hour flexibility “to adapt work schedules to unpredictable shocks to reservation wages.” This flexibility is something that “lower-wage lower-skill workers typically have limited ability” to get from traditional workplaces.
Uber is not alone in experimenting on employees
Uber is not the only ride-hailing company to do this. As long as there is data, companies will look to analyze it.
Lyft tested out money incentives with new drivers in a focus group. What the company found: telling new drivers how much they were losing by working on a Tuesday was more effective at getting drivers to change than telling them how much they were winning by switching to Fridays.
The consultants concluded that we’re sore losers and we hate losing more than we like winning.
Unlike Uber, Lyft chose not to go with this loss-aversion message to its drivers.
Tracking contractors and employees is moving beyond the realm of apps, and into our bodies.
Swedish startup Epicenter gives employees the option to become cyborgs, implanting a microchip in their fingers, so employees can wave their hands to open doors and buy food.
On the upside, you’ll never have to worry about leaving your badge at home again. On the downside, you can always be tracked. Your boss can now see when and where you’re at work under the philosophy that it makes life “easier.”
Is Uber a sinister outlier or just one more employer ‘gamifying’ employment?
Reception to the Times findings has not been all positive. Quartz reporter Alison Griswold criticized the Times for making the use of behavioral science more sinister than it was: “Alternatively, what if there was no ‘whiff’ of coercion because there was no coercion? Uber signed up a bunch of people to drive on its platform and then it gave them tips on ways to earn more money. We might consider that a ‘trick’ if Uber’s advice didn’t actually translate into higher earnings, or if Uber promised incentives that it never paid, but Scheiber has no evidence of anything like that happening.What he does have is a lot of adjective-laden insinuations that Uber is doing something dark and untoward.”
Tech ethnographer Alex Rosenblat, whose research gets cited in the Times article, says these sinister things do happen, but that examples weren’t included in the article.
it's not in NYT story, but drivers do experience unpaid cancellation fees, unpaid hourly guarantees, phantom requests, etc. https://t.co/h3TEDLoMoI
— Alex Rosenblat (@mawnikr) April 4, 2017
A different reporter pointed out that employers ‘gamifying’ employment in obscure ways has already been popularized by advertising and marketing.
6. From Madison Avenue's tricks over half a century ago, to Silicon Valley's obsession today with "dark patterns" https://t.co/gJ7JpBPlHI
— Jeff Guo (@_jeffguo) April 3, 2017