Bias at Work: Three Steps To Minimizing Bias In Performance Reviews

Bias creeps into almost every area of our lives and decision-making, whether we know it or not.

It’s not that we’re bad people, determined to pre-judge others. Rather, it’s our brains taking shortcuts that have been created over centuries. That makes it pretty tricky to undo effectively, but there are ways.

When I last talked about performance reviews at Buffer here on the blog, I shared that one practice we’ve tried during the review process is to share an information sheet on minimizing bias with teammates before they write their reviews.

I’d love to dive deeper into the topic of minimizing bias in performance reviews in this post, and pass those resources – and all of the lessons we’ve learned so far – on to you.

Why it’s important to address bias in performance reviews

But first, a quick look at why this matters.

A study by Kieran Snyder looked at 248 reviews across 28 companies, and found that while 59% of the reviews received by men contained critical feedback, 88% of reviews received by women did. Many were critical of women’s personalities and tone in particular as compared to me.

Data also demonstrates that black workers receive extra scrutiny from bosses, which can lead to worse performance reviews, lower wages, and even job loss.

Disabled people, gender non-conforming people, LGBTQ folks and more are also vulnerable to harmful bias.

According to the Level Playing Field Institute’s incredible Corporate Leavers Survey:

• People of color are three times more likely to leave a job because of workplace unfairness
• LGBTQ professionals leave almost twice as often because of workplace unfairness.
• 27% of those who experienced unfairness at work said their experience strongly discouraged them from recommending their employer to other potential employees.
• Unfairness at work costs U.S. employers $64 billion on an annual basis.

Unfairness erodes trust and company culture. It damages your company and brand. It leaves vulnerable your very best resource – your people.

Fixing performance reviews won’t magically change these dismal numbers overnight. But it’s an important step toward building a culture of mutual respect, inclusivity, and psychological safety for all. So how to take that step?

Step 1: Build awareness of bias

The first step for us at Buffer was to help everyone on the team understand that bias exists, and to build a common vocabulary to talk about it.

On one of our in-person team retreats, we took part in an afternoon of unconscious bias training with the consulting firm Paradigm (thanks, Natalie!) This introduction to bias opened up a lot of great conversations that have been ongoing ever since.

If your company doesn’t have the resources to bring in an external facilitator, you can try a “train-the-trainers” program that will allow someone on your team to take back the knowledge and skills to present to your whole company.

After your team is familiar with these broad concepts, keep the conversation going. At Buffer, we have a very active diversity and inclusion Slack channel. It’s a place where we share resources and ask questions about everything from the term “housekeeping” to what the composition of a speaker panel should look like in order for us to feel comfortable taking part.

We also created important team-wide reference documents, including a Buffer code of conduct and a workplace harassment policy, so there’s a common understanding of what types of behavior are expected and what types won’t be tolerated.

Step 2: Evaluate your review tools and process

Buffer took a bit of an unorthodox path from no reviews to some, sporadic reviews to 360 reviews which involved full-team training and understanding of these concepts.

We’ve learned that a regular, consistent review cadence is best, as feedback-on-the-fly creates conditions for bias to flourish.

If you are already on a regular review cadence or are considering beginning one, an important step is to evaluate your review process, questions and tools to unearth bias opportunities.

Processes

Does your process:

• Begin with clear, specific performance criteria directly related to job requirements?
• Use a consistent review system across your team (or minimally, across an entire area)?
• Ask for specific evidence and examples from the evaluation period?
• Separate personality from skill set?
• Evaluate performance evaluations across the team for consistency?
• Work with managers whose reviews don’t provide specific examples or show signs of bias?

Questions

Do your review questions ask about observable, specific, objective, measurable behaviors? Trakstar shares that a good example of a top rating might be, “Rarely or never misses deadlines.” This is clearer and more objective than a simple, “Meets Expectations” or “Exceeds Expectations.”

We’ve learned to beware of and push back on subjective review observations, which can often delve into personality, style, or attitude, and lead reviewers to “go with their gut” – exactly where unconscious biases rise to the surface. We’re also working on questioning too vague observations, like “they’re rocking it!” to ask for more specific, measurable examples.

As you seek to hone in on specific cases and reviews, a great resource is the site Bias Interruptors. They have many toolkits and resources, but I especially wanted to share their Seven Powerful Bias Interrupters:

1. Give evidence (from the evaluation period) to explain and back up your rating.
2. Make sure to give everyone (or no one) the benefit of the doubt.
3. If you waive objective rules, do so consistently.
4. Don’t insist on likability, modesty, or deference from some but not others.
5. Don’t make assumptions about what mothers—or fathers—want or are able to do.
6. If you comment on “culture fit,” “executive presence,” or other vague concepts, start with a clear definition and keep track to ensure such concepts are applied consistently.
7. Give honest feedback to everyone who is evaluated—otherwise some groups won’t get notice of problems in time to correct them.

Tools

As you look for a tool to help automate parts of the process, or evaluate an existing tool, make sure you research how the tool approaches bias. Is it something they’ve considered and built into their process?

Select software that makes it easy to see company-wide trends, while offering the ability to zoom in on individual employee. As you evaluate, keep an eye out for outliers, like a single rater with much higher or much lower ratings.

At Buffer, we chose CultureAmp for our performance review software because we felt they had a strong focus on eliminating biases, backed by data from psychologists and data scientists. It also made a difference that we could see what a huge focus they put on this topic in their own content and studies.

Step 3: Interrupt bias at its origin

A final step is to build additional checks, resources and awareness into your performance review process to interrupt bias at the points where it’s most likely to occur.

Self-reviews
Self-evaluations are a great reflection tool, but they can be rife with bias when used incorrectly as part of a performance appraisal.

People can differ greatly in terms of confidence and comfort with touting their own accomplishments. Women, for example, are more likely to underestimate their capabilities. And the anchoring effect means that managers can’t help unconsciously adjust their own reviews — for better or worse — based on what they see in a self review.

To interrupt:

• Keep self-reviews and manager reviews separate until both are fully complete.
• Share this resource on writing effective self-reviews with your full team to help level the playing field.

Manager reviews
It’s helpful for your whole team to be aware of bias and the particular types of bias and errors most common in performance appraisals, but it’s especially crucial for team leads and managers who complete performance reviews more frequently.

To interrupt:

• Share a resource like Bias Interruptor’s Identifying Bias in Performance Evaluations Worksheet.
• Share some of the most common biases likely to effect reviews. (There’s a great list below!)
• Encourage reviewers to rely less on their memory (a very unreliable source!) and instead focus on notes and documents.

At Buffer, we encourage anyone who coaches another teammate regular in one-on-one chats to keep an ongoing document where they record achievements and challenges. During review periods, they can look back at these notes to remind themselves of performance highlights, ultimately improving the accuracy of their reviews.

The most common biases likely to affect reviews:

Try peer reviews or 360 feedback
Many of the above techniques and resources listed for manager reviews can also benefit the full team in the event of 360 reviews.

The good news in this area is that a fair amount of research shows that 360 reviews are preferable for mitigating bias because they employ a variety of sources rather than relying on a single data point.

Performance feedback from multiple sources gives you a more complete picture while minimizing the impact of any particular rater’s bias. We’re now doing 360 reviews at Buffer at least once a year for the whole team, and teammates have the option to get more feedback during any review period to round out their overall feedback.

More Reading on Bias at Work

• Clayman Institute for Gender Research’s “See Bias, Block Bias” Toolkit (1, 2)
• Nextions Consulting’s series exploring bias when evaluating candidates of color (1, 2)
• Bias Interruptor’s Identifying Bias in Performance Evaluations Worksheet
• Bias Interruptor’s Interrupting Bias in Performance Evaluations
• Bias Interruptor’s Writing an Effective Self-Evaluation: Worksheet
• Tools for Managers and HR: How to Interrupt Bias in Performance Evaluations

Over to you!

How does your company do performance reviews, and how are you working on the challenge of bias? Has this post sparked any thoughts on bias at work? Any resources to add to this post? I’d love to hear from you in the comments!

Photo by Christin Hume