Grading on a Curve Undermines Performance

Microsoft offers a lesson on how not to conduct performance reviews. Its evaluation process, called “stack ranking” — essentially grading on a curve — has had a disastrous effect on morale, performance, and innovation, reports Kurt Eichenwald in “How Microsoft Lost Its Mojo.” His article in the August Vanity Fair describes a system in which managers are permitted to give only a few employees top reviews, while the majority receive mediocre reviews, and a few receive poor ratings. This structure sets up a competition among employees and fosters an environment in which it is not enough to do outstanding work, but also to make sure that you outshine your co-workers. Said one former Microsoft engineer, “…people do everything they can to stay out of the bottom bucket, [including] openly sabotaging other people’s efforts…. I learned [] to give the appearance of being courteous while withholding just enough information from colleagues to ensure they didn’t get ahead of me in rankings.” This zero-sum thinking undermines collaboration at all levels. Particularly in an industry in which engineers must collaborate to create innovative products and teams must pull together to meet aggressive timelines, fostering competition between teammates is toxic.

In addition to back-stabbing within teams, there arose a highly politicized and inefficient environment of competition among teams in the same group, because the curve applied to the whole group, not just individual teams. So at every six-month review, managers would engage in horse-trading among themselves to achieve the bell curve. This meant that it was not enough to look good to your own manager, but also to make sure you were visible to other supervisors within your group. Ultimately, employees said that their reviews often focused more on playing the political game than on their performance.

According to Forbes, stack-ranking was a more effective tool at GE under Jack Welch, who called the system”Rank and Yank.” When he took as CEO, the company was bloated, and Welch was looking to make cuts, so a rating system in which 20% received top ratings, 70% were mediocre, and 10% got fired served his goal. (Yank and Rank has since been phased out). But if Microsoft is to compete with Google, Apple, and others, Microsoft’s goal should  be attracting and retaining top talent. Steve Jobs, for all his flaws (and he had many) insisted that he wanted all “A” players on his team. You can’t expect to grade such people on a curve and have them stick around. Maybe if Microsoft employees spent less time competing with one another and spent more time pulling together to innovate, they might stand a chance.

Clearly, the evaluation process is not the only thing that has driven Microsoft’s less than stellar performance since the departure of Bill Gates and ascendance of Steve Ballmer, but the review process is emblematic of a lack of attention to what kind of culture produces strong teams and great innovation.

What would be a preferable system? Well, if you want to attract and retain great people and want to foster team collaboration, accountability, and innovation, then your review process must reflect these values. High performance must always be recognized, and not just in relative terms. It also needs to be honest with under-performing employees and constructively evaluate their performance. And while interpersonal interactions and relationships are rightly to be included in reviews, they should emphasize and recognize collaboration, teamwork, and constructive communication rather than on “playing politics.”