It was just before Thanksgiving last November when the always-lively debate about performance management received a new jolt.
First, Microsoft Corp. announced it would ditch its stacked ranking system, a performance management process instituted by General Electric Co. in the 1980s that effectively ranks employees on a bell curve, looking to weed lower performers out and improve the organization’s overall performance.
The use of the performance measurement system, known as “rank and yank,” would become pervasive at large companies in the decades to follow, before many started to realize its unpopularity with employees and, in some instances, its penchant to actually hurt performance.
Then, in the same week, Yahoo Inc. CEO Marissa Mayer entangled herself in controversy when news reports picked up employee chatter on message boards revealing that the technology firm had instituted a similar stack ranking performance management system, called “quantitative performance ranking.”
The development ignited a firestorm in the human resources industry, just as when Mayer reined in Yahoo’s work-from-home program shortly after taking over as part of a massive turnaround effort.
“In this case, it looks like Microsoft is on the right side of history, and Yahoo is still in the Stone Age,” Max Nisen, a strategy reporter at Business Insider, wrote that week.
A System That Pleases No One
But on a broader level, the reignited debate over how to measure employee performance — and how to formalize the process to the benefit of both the employee and the organization — has always been tricky. For starters, the annual performance review has never been popular with either employees or managers. What’s more, attaining objectivity and fairness in the process is almost always threatened by the hidden biases.
With so much ambiguity, it’s no surprise that some leaders avoid controversy by rating everyone’s performance the same. In an August 2013 survey by human resources advisory and research firm Towers Watson & Co., fewer than half of respondents agreed that high performers in their organization are adequately rewarded, likely due to managers’ tendency to clump performance rankings toward the middle.
This is supported by research in the Journal of Economic Literature by University of Chicago economics professor Canice Prendergast. According to his research, when managers are given free rein on ratings, they tend to fall into two categories: centrality bias, where scores are clustered around the midpoint, or leniency bias, where scores bunch nearer the top.
In many cases, the research notes, neither provides incentive to work harder and little consequence to slacking off, so performance plateaus.
Psychologists call this “social loafing,” when hard workers or high performers notice their lazier colleagues getting away with less while receiving the same rewards, so they reduce their effort. In turn, slackers ease off even more.
This is where the forced ranking system was supposed to come to the rescue. At its most draconian, the system sees the bottom 10 percent exit the business. “Rank and yank” certainly provides a wake-up call for coasters, but downsides abound. Most significant: After several cycles the free riders have all left and managers are forced to rate the worst of a very good bunch. As a result, the system creates a gladiatorial culture, where employees are competing against one another.
At Microsoft, according to a lengthy Vanity Fair feature in 2012, it led to employees “competing with each other rather than competing with other companies.” In some instances, according to the story, talented employees would go to great lengths to avoid working with other high performers so as not to harm their position in the rankings. The company abandoned the system in 2013 in favor of more frequent and qualitative employee evaluations.
It appears talent leaders are caught between a rock and a hard place. Is it best to opt for the forced distribution approach with the resulting competitiveness, perceived injustice and distraction of providing a rating? Or let managers rate performance haphazardly, leading to social loafing and a drop in performance?
Perhaps the solution is a healthy mix of the two. Either way, talent leaders and line managers alike are poised to change their mindset.
Guided distribution, where managers are encouraged to differentiate performance but with blurred edges to allow for individual discretion, has been suggested to lead to greater performance. The focus is on performance conversations rather than the overall score, so manager capability is the key to success.
But to rate fairly and consistently and to have constructive conversations about performance, managers with a guided distribution mindset need high levels of emotional self-regulation and the ability to keep unconscious, innate biases in check.
Avoiding Common Traps
The first step toward this end is to recognize and avoid the common rating traps.
Manager-employee relationships can hinder objectivity. Not only do managers give higher, less accurate scores to those they like, but also research shows that a poor rater-employee relationship can lead to inflated ratings, as managers tend to shy away from conflict. Also, our innate “like-me” bias means that managers unwittingly prefer team members who are from a similar background or who share similar interests, further muddying the waters.
A second likely trap is rating potential, not performance. Afraid of demotivating budding high performers, it’s tempting to give them a higher rating than they objectively deserve to make sure that they stick around to realize that potential.
The third trap is failing to consider the entire year. Our brains are primed to pay more attention to recent events, meaning that fresh successes or failures unduly influence scores. Similarly, the “halo or horns” effect happens when a single high-profile achievement or mistake affects the way we interpret people’s entire performance — that is, we’ve unconsciously made up our mind about whether they’re a strong or a weak performer and only see evidence which confirms this view, thus making it even stronger.
The fourth bias managers must avoid is being anchored by an individual’s previous score. Research by German psychologists Fritz Strack and Thomas Mussweiler shows that our judgments are influenced by information we’re given previously.
Managers often assume that individuals’ performance is consistent from one year to the next, especially if they lack concrete evidence. People crave internal consistency, so managers will unconsciously justify last year’s rating by awarding the individual the same this time around.
Another trap involves falling foul of cultural rating norms, such as scoring everyone around the midpoint or being lenient and inflating everyone’s scores, which often goes hand in hand with wanting to be liked.
A final trap is using the performance score as a means to rectify other issues — for example, reasoning that a higher rating will give the individual more chance of receiving a long-overdue raise.
Educating managers about these six traps helps to move toward objectivity. With every rating, managers need to ask themselves whether they’re truly being impartial. They need to make sure they’re rating individuals regardless of their relationship, based on what they’ve actually achieved rather than what they’re capable of. They need to consider the full year’s performance, making a decision independently of last year’s score. And they need to avoid any cultural bias, taking into account performance against goals only.
Gathering concrete evidence throughout the year from a variety of sources about what employees have achieved — and how — helps to increase objectivity, as do calibration meetings, where managers discuss, justify and corroborate provisional ratings.
Calibration allows managers to compare and rank individuals across departments and adjust any scores that are unduly lenient or severe. These meetings also build a shared understanding of what “excellent” looks like vs. “very good” and, as a result, help managers give consistent feedback throughout the year.
Making Conversation Work
Once managers have arrived at an objective, unbiased rating comes the hardest part — delivering the verdict.
In general, people tend to have an overinflated view of themselves. According to research published in 1993 in the European Review of Social Psychology, most people see themselves as more attractive, intelligent, funny and better drivers than they actually are. It’s no wonder that managers shy away from having honest conversations about performance — how do you tell half your team that they’re worse than they think they are without demotivating them?
The answer, in many instances, lies in equipping managers with the mindset, skills and confidence to tackle these tricky conversations. Bosses must accept that they’re not always going to be liked, and nor should they strive to be. By delivering messages in a rational, empathetic and authentic way, they will be respected.
But they also need to take responsibility for the rating by avoiding get-out clauses such as “I wasn’t allowed to give you the rating I think you deserve.” This only increases bitterness and a sense of injustice.
Managers who get performance rating conversations right share their decision with conviction, backed up by concrete examples of achievements and behaviors. They put emotions to one side, share facts, not interpretations and avoid making generalizations. They invite the other person to share his or her thoughts and feelings about the rating and prepare for likely reactions. Moreover, they focus on the individual’s achievements, development areas and career path, avoiding explicit comparisons with other people.
Most important, managers who leave team members feeling motivated succeed because they don’t just have these types of conversations once a year. They give high-quality feedback on a daily basis and have regular conversations about how their team members are performing against their goals.
In short, managers should aim to yank the “rank and yank” way of thinking and the biases that come with it, and foster a performance measurement mindset that values continual, qualitative measures that leave employees motivated and the C-suite informed of how their talent stacks up.
Click here to read more about MetLife Inc.'s efforts to build manager capability.
Sebastian Bailey is president of Mind Gym Inc., a performance management firm. He can be reached at firstname.lastname@example.org.