The Struggle To Measure Performance

Rigid rankings hinder the teamwork and risk-taking necessary for innovation. But what combination of methods works best?

Holiday shopping, yearend deadlines, and emotional family dramas aren't the only stresses in December. 'Tis the season for companies to embark on that dreaded annual rite, the often bureaucratic and always time-consuming performance review. The process can be brutal: As many as one-third of U.S. corporations evaluate employees based on systems that pit them against their colleagues, and some even lead to the firing of low performers.

Fans say such "forced ranking" systems ensure that managers take a cold look at performance. But the practice increasingly is coming under fire. Following a string of discrimination lawsuits from employees who feel they were ranked and yanked based on age and not merely their performance, fewer companies are adopting the controversial management tool. Critics charge that it unfairly penalizes groups made up of stars and hinders collaboration and risk-taking, a growing concern for companies that are trying to innovate their way to growth. And a new study calls into question the long-term value of forced rankings. "It creates a zero-sum game, and so it tends to discourage cooperation," says Steve Kerr, a managing director at Goldman Sachs Group Inc. (GS ), who heads the firm's leadership training program.


Even General Electric Co. (GE ), the most famous proponent of the practice, is trying to inject more flexibility into its system. Former Chief Executive Jack Welch required managers to divide talent into three groups -- a top 20%, a middle 70%, and a bottom 10%, many of whom were shown the door. Eighteen months ago, GE launched a proactive campaign to remind managers to use more common sense in assigning rankings. "People in some locations take [distributions] so literally that judgment comes out of the practice," says Susan P. Peters, GE's vice-president for executive development.

Striking that balance between strict yardsticks and managerial judgment is something every company, from GE to Yahoo! (YHOO ) to American Airlines , is grappling with today. But finding a substitute for a rigid grading system is not an easy task. It drives truth into a process frequently eroded by grade inflation and helps leaders identify managers who are good at finding top talent.

That's one reason GE isn't abandoning its system. But it has removed all references to the 20/70/10 split from its online performance management tool and now presents the curve as a set of guidelines. The company's 200,000 professional employees tend to fall into Welch's categories anyway, but individual groups are freer to have a somewhat higher number of "A" players or even, says Peters, no "bottom 10s." Even those low achievers are getting some kinder treatment, from a new appellation -- the "less effectives" -- to more specific coaching and intervention than in the past.

The changes are key for a company trying to evolve its culture from a Six Sigma powerhouse to one that also values innovation. Tempering such rigid performance metrics, says Peters, "enables individuals and organizations to be more comfortable with risk-taking and with failure." To drive that point home, the company's top 5,000 managers were evaluated for the first time this year on five traits, such as imagination and external focus, that represent the company's strategic goals.


Separating stars from slackers remains a long-standing part of GE's performance-driven culture. But for most companies, especially those without such cultures, the benefits of adopting a forced ranking system are likely to dissipate over the long term.

A recent study lends hard data to that theory. Steve Scullen, an associate professor of management at Drake University in Des Moines, found that forced ranking, including the firing of the bottom 5% or 10%, results in an impressive 16% productivity improvement -- but only over the first couple of years. After that, Scullen says, the gains drop off, from 6% climbs in the third and fourth years to basically zero by year 10. "It's a terrific idea for companies in trouble, done over one or two years, but to do it as a long-term solution is not going to work," says Dave Ulrich, a business professor at the University of Michigan at Ann Arbor. "Over time it gets people focused on competing with each other rather than collaborating."

One company that recently decided to dump forced rankings altogether is Chemtura (CEM ), a $3 billion specialty chemicals company formed by the July merger of Crompton in Middlebury, Conn., and Great Lakes Chemical in Indianapolis.

"The system forced me to turn people who were excellent performers into people who were getting mediocre ratings," says Eric Wisnefsky, Chemtura's vice-president for corporate finance. "That demotivates them, and they'd follow up with asking: 'What could I do differently next year?' That's a very difficult question to answer when you feel that people actually met all your expectations." Chemtura's new process still assigns grades. But to better motivate employees in the middle, labels such as "satisfactory" have been upgraded to phrases such as "successful performance."

Yahoo, too, was looking for better dialogue and less demoralizing labels when it made substantial changes this year to its rating system, which compared employees' performance to an absolute standard rather than to each other. Libby Sartain, Yahoo's senior vice-president for human resources, knew that review discussions at the Sunnyvale (Calif.) tech leader frequently included the wink-wink "I wanted to put you here, but I was forced by human resources to do something different" comment that discredits so many appraisals. This year, Yahoo stripped away its performance labels, partly in hopes that reviews would center more on substance and less on explaining away a grade.

But that doesn't mean Yahoo went all Pollyanna on its employees. To do a better job of finding and showering top performers with the rewards necessary to keep them from jumping ship in talent-tight Silicon Valley, the company also instituted a "stack-ranking" system this year to determine how compensation increases are distributed. It asks managers to rank employees within each unit -- a group of 20 people would be ranked 1 through 20, for example -- with raises and bonuses distributed accordingly. During reviews, employees are told how their increases generally compare to those of others.

Some Yahoo managers are livid about the new system. "It's going to kill morale," laments one senior engineering manager who says he's getting a stronger message to cull his bottom performers. Yahoo says its new program doesn't automatically weed out a bottom group and was designed specifically to reward its stars.

Indeed, what Yahoo has introduced in place of its old system shows how hard it is for companies to find ways to foster merit-driven cultures that coddle standouts while staying tough on low performers. Whether a company calls it stack ranking, forced ranking, or differentiation, "there's no magic process," says Sartain. "We just want to make sure we're making our bets and that we're investing in the people we most want to keep. That's what this is all about."

By Jena McGregor

    Before it's here, it's on the Bloomberg Terminal.