Companies are increasingly ditching their performance reviews.
It’s no wonder, with headlines like “Performance Reviews Are Pointless and Insulting – So Why Do They Still Exist?” Not to mention, managers hate doing them: in one informal poll, described in the aforementioned article, 35 of the 37 managers at one company were in favor of getting rid of performance reviews.
But this trend is raising red flags for researchers of diversity & inclusion. Research has consistently demonstrated that the lack of a formalized evaluation process can lead to increased bias against women and minorities. Without a transparent process and clear evaluation criteria in place, managers are likely to rely on pre-existing biases when making pay and promotion decisions. Ambiguity can increase the tendency for managers to use cognitive short-cuts, such as gender and racial biases, that frame their decision-making in subtle ways and lead to systematic disadvantages for women and minorities.
This trend away from performance reviews has its upside. Some companies, like GE, Deloitte, Adobe, and others, have opted for an emphasis on “one-on-ones”: encouraging and structuring more frequent touch-points between employees and their managers. Providing a process for managers to coach employees throughout the year can help identify problems early and recognize wins.
But one-on-ones alone cannot replace a clear, consistent, formalized evaluation process. When managers sit down to decide on pay and promotion decisions, if all they have are notes from various one-on-ones conducted over the past year, they may have trouble consolidating those into a fair decision. Too much unstructured information may be as bad as no information at all – both open the door to cognitive shortcuts and bias.
Instead, here are four tips for companies seeking to design a fair, bias-free evaluation process.
Establish clear criteria that link to organizational impact.
Rather than evaluating employees on nebulous – or nonexistent – criteria, companies should specify clear expectations for employees. Research by the Stanford VMware Women’s Leadership Innovation Lab has shown that providing clear prompts and rubrics for managers can help prevent bias from influencing their decisions.
Also vital: make sure your criteria measure the right factors for organizational success. When companies rely on poor proxies – such as long working hours – you not only inhibit your own success as an organization, but you open the door to bias. Limit feedback about employees’ personality and communication styles, as these are known areas where gender and racial biases tend to flourish. Instead, tie feedback to business outcomes.
Choose a reasonable number of criteria and require examples of behaviors that illustrate the criteria.
One reason managers hate the existing performance review process: too many competencies. Traditional competency models can be cumbersome to implement, with an unreasonable number of competencies managers must keep in mind when evaluating employees. Instead, stick to no more than 3-5 competencies.
That said, make sure each criterion or competency is clearly defined, detailing specific behaviors expected of employees. How will managers know if an employee is demonstrating that competency at the appropriate level? How will success be measured? Encouraging managers to be as specific as possible is critical, since research by the Stanford VMware Women’s Leadership Innovation Lab suggests vagueness in performance reviews can lead to bias. An “open box” is not as effective as specific prompts in structuring managers’ feedback for employees. Managers should carefully document specific examples of employees’ behaviors within each competency – something regular one-on-ones can help identify.
Also consider what the rating scales will look like. Companies often struggle to connect qualitative and quantitative feedback. When assigning numerical ratings, ensure those ratings are fairly applied and match the content of written performance evaluations. Research shows women and underrepresented minorities are often held to a higher standard for receiving the top ratings, so make sure to stamp out bias in the connection between performance feedback and numerical rating.
Apply criteria consistently across employees.
Research suggests managers can shift criteria to justify their preexisting biases. Ensure the criteria on the books are consistently applied to all employees. If certain criteria are more important to the organizational mission than others, be sure to communicate that expectation openly to employees and hold managers accountable for applying those criteria properly. Obviously specific departments or teams may need to customize their criteria to accurately reflect the work they do, but make sure employees are held to a consistent standard. Calibration or talent review meetings represent one stage at which biases can particularly sway employee evaluations. Companies can help combat this tendency by assigning a “criteria monitor,” whose role includes flagging if criteria are applied inconsistently across employees.
Make pay and promotion decisions based on performance reviews.
If pay and promotion decisions are disconnected from performance evaluations, this can lead to what Emilio Castilla calls “performance-reward bias,” where the reward doesn’t match the performance assessment. Make sure rewards are based on data rather than gut feelings. Employees may lose faith in the evaluation process if their performance evaluations don’t translate into commensurate rewards. Furthermore, making pay and promotion decisions without a firm basis in employees’ demonstrated behaviors can leave an organization vulnerable to legal challenges.
Historically, performance reviews have caused a lot of strife. They are cumbersome to complete, biased in their application, and disconnected from organizational success. But that doesn’t mean companies should toss them entirely. By establishing clear, consistent criteria linked to business impact, companies can provide employees with actionable feedback and reduce the effect of bias on employees’ career outcomes.