Job Description and Job Evaluation- Scholarly Activity

Lori Goler is the head of People at Facebook. Janelle Gale is the head of HR Business Partners at Facebook. Adam Grant is a professor at Wharton, a Facebook consultant, and the author of Originals and Give and Take.

ZS U

ZS A

N N

A IL

IJ IN

HBR.ORG

Let’s Not Kill Performance Evaluations Yet Facebook’s experience shows why they can still be valuable. BY LORI GOLER, JANELLE GALE, AND ADAM GRANT

November 2016 Harvard Business Review 91

LET’S NOT KILL PERFORMANCE EVALUATIONS YET

tThe reality is, even when companies get rid of performance evaluations, ratings still exist. Employees just can’t see them. Ratings are done sub-jectively, behind the scenes, and without input from the people being evaluated. Performance is the value of employees’ contribu-

tions to the organization over time. And that value needs to be assessed in some way. Decisions about pay and promotions have to be made. As research- ers pointed out in a recent debate in Industrial and Organizational Psychology, “Performance is always rated in some manner.” If you don’t have formal evaluations, the ratings will be hidden in a black box.

At Facebook we analyzed our performance man- agement system a few years ago. We conducted fo- cus groups and a follow-up survey with more than 300 people. The feedback was clear: 87% of people wanted to keep performance ratings.

Yes, performance evaluations have costs—but they have benefits, too. We decided to hang on to them for three reasons: fairness, transparency, and development.

Making Things Fair We all want performance evaluations to be fair. That isn’t always the outcome, but as more than 9,000 managers and employees reported in a global sur- vey by CEB, not having evaluations is worse. Every organization has people who are unhappy with their bonuses or disappointed that they weren’t pro- moted. But research has long shown that when the process is fair, employees are more willing to accept undesirable outcomes. A fair process exists when

evaluators are credible and motivated to get it right, and employees have a voice. Without evaluations, people are left in the dark about who is gauging their contributions and how.

At Facebook, to mitigate bias and do things sys- tematically, we start by having peers write evalua- tions. They share them not just with managers but also, in most cases, with one another—which reflects the company’s core values of openness and transpar- ency. Then decisions are made about performance: Managers sit together and discuss their reports face-to-face, defending and championing, debating and deliberating, and incorporating peer feedback. Here the goal is to minimize the “idiosyncratic rater effect”—also known as personal opinion. People aren’t unduly punished when individual managers are hard graders or unfairly rewarded when they’re easy graders.

Next managers write the performance reviews. We have a team of analysts who examine evalua- tions for bias (after the managers’ names have been stripped away). For instance, are words like “abra- sive” used more often to describe women—and how might that be affecting their assessments, their promotions, their pay?

And last, we translate our ratings directly into compensation. Notably, this process involves a for- mula, so managers have no discretion in compen- sation decisions. It’s fair: If you excel, your bonus multiplier rises according to a predetermined equa- tion, not someone’s opinion. This focuses managers on what they can accurately assess and allows the company to manage pay using compensation

he long march to the boss’s office to get evaluated—it’s a moment we all dread. Performance reviews are awkward. They’re biased. They stick us in boxes and leave us waiting far too long for feedback. It’s no surprise that by the end of 2015, at least 30 of the Fortune 500 companies had ditched performance evaluations altogether. But let’s not throw the baby out with the bathwater.

92  Harvard Business Review November 2016

expertise. It’s also a huge time-saver. When other companies eliminate performance evaluations, they still spend many hours agonizing over compensa- tion decisions. For us, time invested in performance reviews is time saved on compensation.

Being Transparent People want to know where they stand, and perfor- mance evaluations offer transparency. They help employees understand how their contributions are seen in the organization, and they make it easier for the organization to effectively recognize and reward top performance.

Many companies that are abandoning perfor- mance evaluations are moving to real-time feedback systems. That is an excellent way to help people re- peat their successes and learn from their failures. But it doesn’t help them—or the organization—gauge how they’re doing overall.

Long before he won the Nobel Prize in econom- ics, psychologist Daniel Kahneman worked with the Israeli army to evaluate hundreds of cadets. He found that even after they’d been rated on many specific dimensions, a global rating of overall per- formance added information. “A global rating is very good,” Kahneman says, “provided that you have gone through the process of systematically evaluating.” For example, when managers keep a journal of key perfor- mance episodes, the quality of their feedback tends to improve, and their employees often react more posi- tively to the evaluations. At Facebook we have experi- mented with various approaches to translating micro assessments into a macro performance rating, using categories such as “technical contributions,” “team contributions,” and “planning and execution” as key dimensions that contribute to the overall score.

In the CEB survey mentioned earlier, people whose companies had eliminated evaluations judged their

performance conversations 14% more negatively than those whose organizations still used them. (The people who had fared better under the old system were, understandably, the most miffed.) At Facebook a focus group participant said that ratings serve as a punctuation mark, because they’re clear. Another participant said she likes “having the chance to be a unicorn.” We value outsize contributions—and those who deliver them should feel appreciated.

Developing People Finally, there’s development, the third advantage of performance evaluations. In a recent HBR article, Wharton’s Peter Cappelli and HR expert Anna Tavis argue that annual reviews favor accountability over development—and that can certainly be true. But when conversations about professional growth are near-constant and untethered by ratings, people get overwhelmed. Facing a barrage of feedback, employees often struggle to figure out which infor- mation matters most and what to ignore. A compre- hensive analysis of 607 studies showed that more than a third of all feedback interventions backfired, decreasing performance instead of increasing it.

Idea in Brief THE TREND Because performance evaluations are often biased and their annual cycles leave employees waiting too long for feedback, many companies are dropping these reviews.

THE PROBLEM But getting rid of evaluations doesn’t eradicate bias, since managers still rate people subjectively, without formal input. And it doesn’t ensure that employees will receive feedback when it’s most useful.

ANOTHER PERSPECTIVE Companies needn’t throw out reviews altogether. They can still reap the benefits— fairness, transparency, and development—while managing the costs. This article explains how Facebook is doing that.

HBR.ORG

November 2016 Harvard Business Review 93

LET’S NOT KILL PERFORMANCE EVALUATIONS YET

feedback. If you’re nervous, you’d rather find out you’re failing than not know how you’re doing at all.

If we can make one general statement from neu- roscience, it’s that the brain is remarkably flexible and adaptable. Some people may react strongly to ratings, but they can learn to respond differently. At Facebook we are trying to build a culture in which people ap- proach ratings with curiosity and a learning orienta- tion. When our senior leaders receive performance evaluations, they often share the feedback with their teams, normalizing the fact that even people who con- sistently deliver strong results sometimes have lapses.

Another common critique is that ratings cre- ate fixed rather than growth mindsets. It’s true that when managers have fixed mindsets, they’re less likely to notice improvements or declines in perfor- mance (they’ve already locked people into catego- ries) and less likely to coach people. But when com- panies eliminate reviews, managers actually devote less time to performance management.

The solution here is not to throw out performance ratings but to build a culture that recognizes and re- wards growth. At Facebook we don’t believe in A, B, or C players—we’re assessing a period of time, not a person. Even David Bowie released a bad album once in a while. In fact, new evidence from a large retail company reveals that performance evaluations are surprisingly variable: People have just a 33% chance of getting the same rating from one year to the next. At Facebook we’ve found that people who receive as- sessments in the bottom 10% have a 36% chance of making it into the top half within a year.

We set stretch goals—superstretches, actually— which we call 50-50 goals. These are so ambitious that there’s an equal chance people will or won’t achieve them. Those who do meet them want—and deserve—to know who they are. So do those who fall short of their goals. We aim for clarity at both ends of the spectrum. Classic research suggests that people are often highly motivated when success is a coin toss. With lower odds they’re more prone to give up, and with higher odds they don’t marshal enough effort or creativity.

“Democracy is the worst form of government,” Winston Churchill reminded us in 1947, “except for all those other forms that have been tried from time to time.” The same is true for performance evalua- tions: They’re far from perfect, but they’re also far better than the alternatives. HBR Reprint R1611G

When people receive negative feedback, they often fixate on small points. Without ratings, they can spend weeks pruning a few trees while the for- est is on fire. If a manager receives multiple pieces of feedback about being late to meetings but misses the larger issue of prioritization, she might become the timeliest person to deliver mediocre results. Performance evaluations allow for an overall assess- ment that helps people prioritize. Employees learn what their key strengths are and where they should focus their development efforts. Evaluations also serve as a forcing function to make sure that tough feedback is delivered rather than swept under the rug.

Trade-Offs All Around We’re not saying that performance evaluations are uniformly beneficial. We’re saying that they in- volve trade-offs, and we’ve decided to keep ours to achieve the goals of fairness, transparency, and de- velopment. Performance reviews were put in place for good reasons; discarding them entirely might be an overreaction to how they’re often executed.

Critics of performance evaluations have sug- gested that ratings automatically produce a fight-or- flight response. Actually, many people have stronger reactions to not being rated. Neuroscientists have found that highly anxious people have more-intense neural reactions to uncertainty than to negative

94  Harvard Business Review November 2016

HBR.ORG

Copyright 2016 Harvard Business Publishing. All Rights Reserved. Additional restrictions may apply including the use of this content as assigned course material. Please consult your institution's librarian about any restrictions that might apply under the license with your institution. For more information and teaching resources from Harvard Business Publishing including Harvard Business School Cases, eLearning products, and business simulations please visit hbsp.harvard.edu.