In early 2009, a small team in Google’s people analytics group brought a list of research questions to its boss, Prasad Setty. One of them had come up again and again since the company was founded: do managers matter?
By Harvard Business Review’s account, most Google engineers saw management as a distraction from real work. In 2002 the founders had gone as far as removing engineering managers altogether, an experiment HBR says lasted only a few months.
The new study was code-named Project Oxygen. Its co-lead, Neal Patel, knew his audience. “Google has high standards of proof, even for what, at other places, might be considered obvious truths,” he told HBR. “Simple correlations weren’t going to be enough.” So the team set out to prove the opposite: that managers made no difference.
For a while the data seemed to agree. Exit interviews showed some link between people leaving and low satisfaction with their managers, but turnover at Google was low overall, and nothing proved that managers caused it. Patel then compared the top and bottom quarters of managers on survey scores and performance reviews. “At first,” he said, “the numbers were not encouraging. Even the low-scoring managers were doing pretty well.”
Sharper statistics broke the tie. Multivariate analysis showed that even “the smallest incremental increases in manager quality were quite powerful.” In 2008, teams under high-scoring managers had less turnover, and whether people stayed tracked the quality of their manager more closely than their own seniority, performance, tenure or promotions.
“Luckily, we failed,” Patel said.
The harder question was what the good ones did. The analysts gathered more than 10,000 observations about managers, from reviews, surveys and award nominations, and coded the comments by hand. Double-blind interviews with high- and low-scoring managers filled more than 400 pages of notes.
Out came eight behaviors, and most sounded like common sense: be a good coach, don’t micromanage, take an interest in your people’s well-being and careers, have a clear vision. Laszlo Bock, who ran Google’s people operations, told the New York Times his first reaction was, “that’s it?”
Then Bock’s team ranked the eight. For much of its history Google had left engineers alone, on the theory that if they got stuck they could ask a boss whose deep technical expertise had put them in charge. Technical expertise came last.
“In the Google context, we’d always believed that to be a manager, particularly on the engineering side, you need to be as deep or deeper a technical expert than the people who work for you,” Bock said. It was important, he added, but it paled next to making a connection and being accessible.
Google turned the eight into twice-yearly surveys in which employees rated their own managers, starting in June 2010, and into training. From 2010 through 2012, median favorable scores rose from 83 to 88 percent, and the lowest-scoring managers improved most. HBR noted the limit: nobody could yet show how much those scores moved sales, productivity or profit. Google later stretched the list to ten, adding collaboration across the company and strong decision making.
Eric Clayberg, a software engineering manager, had led teams for 18 years before Google bought his latest startup. Six months of Oxygen surveys and classes, he said, taught him more about managing than the two decades before. He began spending a third to half of his time helping his people grow, and to his surprise they welcomed it.
“Engineers hate being micromanaged on the technical side,” he told HBR, “but they love being closely managed on the career side.”