Employee morale: A Googly eye

Tech titan uses algorithm to reveal who is an unhappy employee.

Google Inc. obviously knows a thing or two about helping people find digital information. But when it comes to stopping employees from finding other jobs, management experts question the search-engine company’s people skills.

Once a trailblazing startup that could keep employees happy with golden perks (such as free meals and on-site oil changes) and lucrative stock options, Google is now one of the many mature tech companies cutting labour costs and workplace gravy to fight a downturn that has taken the shine off stock-based compensation. As a result, Google, based in Mountain View, Calif., faces a brain-drain issue due to rising employee resignations. So the firm did what it does best: it developed a mathematical algorithm to search out unhappy employees.

Using data to drive management decisions is one of Google’s 10 Golden Rules. In this case, Laszlo Bock, the company’s head of HR, recently told Dow Jones that the plan is to “get inside people’s heads even before they know they might leave.” The company declined to comment on development costs (which sources say could be relatively low) or elaborate on the data inputs being used to try to read employee minds. But the algorithm is billed as non-intrusive, which would seem to rule out using personal data such as employee sick days or e-mail and Internet usage. Whatever the case, critics say company managers appear to be computing too hard.

“It strikes me as pretty bizarre,” says James Smither, a professor of management at La Salle University in Philadelphia. He calls it a classic example of doing something because it can be done, not because it makes business sense. “Why not just ask people whether they are happy employees?”

And Rick Brenner, principal of Chaco Canyon Consulting based in Cambridge, Mass., doesn’t think computer-based “predictive attrition” will become a significant HR trend any time soon. “As an ex-software engineer,” he says, “I would love for this story to be true. For right now, though, it probably is science fiction.” Pointing to the old computer science saw “garbage in, garbage out,” he insists “no algorithm is any better than the data you feed it. And most data about behaviour has questionable reliability.”

Brenner adds that “a capable manager can always outperform a computer program” when it comes to “motivational assessment” — a thought echoed by other HR experts.

Andrew O’Keeffe, an Australia-based management consultant and author of The Boss (a just-released novel about the impact leadership has on employee moral and output), notes Google is not the first company to mine data to try to identify unengaged employees. That said, he notes, “you don’t need a lot of complex analysis to discover the obvious.”

And the obvious, O’Keeffe says, is that companies with retention problems have too many managers who are “not great to work for.”