Marissa Mayer may extoll the virtues of having employees under one roof. But interface designer Andrew Wilkinson found that after moving his company, Metalab, into office space, he was "less productive than I used to be when I worked from home," he says. The problem, he decided, was "shoulder taps"—the brief but distracting check-ins that managers foist on their employees to see "how things are going." Wilkinson found these interruptions so damaging to his concentration that he cooked up an entire application, called Peak, in order to banish them from his workplace.
Peak basically works like a timesheet that fills itself out for you and reports back to the boss without your input. It hooks into cloud-based productivity applications like Github, Google Apps, and Dropbox and tracks your activities in them. Then it sends Tweet-like status updates to your boss, which might include how many emails you sent and received or how many Github commits you pushed. It also aggregates these datapoints and translates them into more humanoid terms, such as "Tim is drowning in email" or "Your team is most active on Monday afternoons."
Managers feel safer when they have these little updates, and Peak is designed to deliver them without disrupting employees' actual work. Still, I had to ask: Could Peak just be micromanagement by other (automated) means? Sure, it's nice not to have the boss interrupting me five times a day. But what does having a digital Dwight Schrute surveilling my workday do to my morale? With all of my "activity" quantified and reported by Peak, would some part of me become self-conscious about how many emails I've sent or Dropbox file-modifications I've initiated?
"We understand the micromanaging concerns, but the fact remains that if a manager or colleague needs to know what’s happening with an item, they’re going to interrupt someone to find out," says Marshall Haas, director of Peak at Metalab. "The idea isn’t to encourage micromanagement, it’s to encourage passive management. We want to enable managers to simply take a look at Peak to see what someone is up to on a given day, or what they accomplished yesterday, and only intervene as needed." He also adds that Peak isn't just for bosses. "The same benefits are there for colleagues working together," he says. "At MetaLab all of our employees can see what everyone else is up to via Peak."
As for how Peak tracks and measures worker activity, Haas says that Metalab took inspiration from the quantified-self movement. But the most interesting feature in Peak isn't its robotic reports on how many emails someone has sent, but rather in those aggregated "translations" like "Brandon has been working longer days lately." It reminded me of the Anil Dash's exhortation earlier this year to replace all dashboards with feeds—that is, design analytics software to not just mine the data and display it (however prettily), but also to digest it in verbal form. After all, if you delegated the task of "checking in" with your workers to a human assistant manager, that person wouldn't take a census of how many emails everyone sent. Data without context and interpretation is not insight. Haas agrees: "The challenge is in mining [data] and pulling out what's important. We don’t want to say that a certain employee is better than another. The idea is simply to provide you with data that you can draw conclusions from yourself and help inform your decisions."
But just as I was skeptical of an app for quantifying one's own emotions, I'm not sure I buy into the idea that reducing an employee's productivity into machine-readable statistics is the best way to observe, much less evaluate, said productivity. Then again, as a freelancer, I have no idea what it's like to actually manage an office full of employees in the first place. Peak seems like a clever and humanely intended solution to a specific problem: "My manager's check-ins distract me, so how can we remove those check-ins?" Whether or not this solution generates other unforeseen problems of its own remains to be seen. To its credit, Metalab seems to be aware of this possibility: The team is implementing Peak only with certain hand-picked partner companies "to ensure a fit, and learn about their needs as we go."
[Images: Courtesy of Peak]