Microsoft, GE, and the futility of ranking employees
FORTUNE — In his heyday, General Electric CEO Jack Welch was such a renowned manager that other corporate giants like Microsoft, Ford, and Conoco rushed to mimic his policies, including Survivor-esque evaluations that guaranteed some workers would be graded as failures.
The “rank and yank” system that he popularized results in workers being pitted against their peers to avoid being labeled as losers. Those workers who ended up on the wrong side of the ranking curve were penalized, usually by a denial of merit raises or bonuses, and sometimes by losing their job.
“An employee could only get high ratings if everyone else fails,” says Cindy Parker, a management professor at George Mason University School of Management in Virginia.
While company performance experts frown on such “stacking” evaluations, plenty of organizations, including legal, consulting, and even banking — not to mention college professors who use grading curves — employ informal stacking systems, for example, by measuring employee value according to the dollar amounts they bill clients.
But forcing managers to segregate their workers “does far more harm than good,” says Bob Rogers, president of Development Dimensions International, the management development firm, who called out the practice in his book, Realizing the Promise of Performance Management.
“It causes damage by filtering employees from the bottom, and causes changes in people’s behavior, and not to the good,” he concludes.
As far back as the early 2000s, companies ran into trouble with the forced ratings scale, adapted from the military’s up-and-out promotions model. Even so, a sizeable chunk of companies continue to use it in all or part of its structure. The most notable company was Microsoft (MSFT), which hung onto it until last week, finally bowing to long-standing criticism that such rigid employee ratings can cripple collaboration and creativity.
Even as Microsoft was jettisoning the controversial appraisal practice, the beleaguered tech giant Yahoo (YHOO) was adopting it, according to a report from AllThingsD, a technology news site. Yahoo did not confirm its use, but reports indicate that its high-profile leader, Marissa Mayer, is relying on the stack ratings model to winnow out employees and fire them — the “yank” part of the equation.
Questions were raised about the Welch approach as far back as the early 2000s, when employees of Goodyear and Ford (F) challenged the rankings as discriminatory. Employees at both companies claimed they were singled out because of their age and, in 2002, Ford paid $10.5 million to settle two class actions suits. Both companies later dropped the evaluation system.
Microsoft also settled lawsuits with employees who claimed the forced ratings led to racial discrimination by “predominantly white male” managers, and Conoco (COP) settled a lawsuit brought by the Justice Department that accused the Houston-based company of using the appraisals to favor cheaper foreign workers over U.S. citizens.
Although GE (GE) faced its own lawsuits alleging age and race discrimination, Welch has defended the grading, claiming it’s necessary to weed out bottom performers and that it’s kinder to eliminate employees before they hit the middle of their careers, when it is much tougher to change jobs. Critics feel otherwise and companies are steadily dropping the system, instead allowing managers to measure employees against broader company standards, not against their peers in a small section or division.
Clifford Stevenson, lead management researcher for the Institute for Corporate Productivity, a Seattle research firm, said his organization’s 2011 survey found a decline in the number of companies, especially those that are high-performing, using the approach.
“The percentage of companies reporting that they used a forced-ranking system declined from 42% to 14%,” he says.
“The majority of organizations surveyed, about 57%, calibrate pay and performance outcomes for the entire organization, while smaller percentages calibrate at the division, department, team, or geographic level instead,” he says.
That flexibility allows managers, he says, to exercise more discretion in assigning ratings, but it also places “an increased burden on managers to make subjective decisions on employee performance.”
There are variations by company, but forced rankings typically require managers to divide their employees, sorting a fixed percentage of workers into the bottom 10% or so as underperforming, the middle 50-60% as passing, and the rest as superior or top-performing. The stars receive raises, promotions, training, and education opportunities.
When managers are forced to rate their employees, personal factors come into play — like favorites and personalities — and managers and employees spend more time networking and currying favor to highlight their accomplishments than actually achieving them. But the unhealthiest result, experts say, is the fact that someone on the work team will be pigeonholed as a failure.
Microsoft, whose competitive momentum, critics argue, has been hindered by lack of idea-sharing, is going to focus on teamwork and collaboration and developing its employees, according to an announcement by its human rights director Lisa Brummel.
She has said there will be “no more curve” at Microsoft, and that managers will have the discretion to reward individual employees.
While forced ratings systems have considerable downsides, there can be a place for them at companies, argues Brooks Holtom, a professor at Georgetown University’s McDonough School of Business.
“The ratings can be arbitrary, and they can have serious effects on creativity and team-sharing, but they can have the benefit of helping cut the deadwood,” he says. “But, typically speaking, forced rankings can work in the first year or two, so it’s only a short-term strategy.”