Google uses algorithm to identify workers likely to leave

Internet giant Google is working on a database that will determine which of its 20,000 employees are the most likely to leave the company, after it has seen some of its high-end staff head out the door.

The database is based on employee reviews, pay and promotion histories and a mathematical formula that the company claims can accurately determine which workers will quit ahead of others.

The company says that the database has already identified workers who feel their skills are being under-used, a key complaint of workers contemplating leaving.

Former Google employees, such as designer Doug Bowman, engineer Steve Horowitz and search-engine worker Santosh Jayaram, have all left the company to work on internet start-ups.

The company has even lost some of its executive team to other companies, with head of display advertising David Rosenblatt and advertising sales chief Tim Armstrong both departing.

Valerie Frederickson, a consultant in the Silicon Valley area who has worked with Google employees, told the Wall Street Journal that Google is losing employees by not implementing personable human-resources programs and training.

“They need to come up with ways to keep people engaged. If Google was doing this enough, they wouldn’t be losing all these people,” she said. 

But Google spokesperson Matt Furman responded that “we haven’t see the most critical people leave”.

Laszlo Black, Google human resources director, said that the algorithm helps the firm “get inside people’s heads even before they know they might leave”.

Related stories:

•  How to lead in lean times
•  10 ways to boost morale and productivity