Textio's Analysis of 23,000 Performance Feedback Reviews Exposed Gender Bias, Racial Bias, and Hyperbole. Plus: My Interview With Textio's Co-Founder

Textio 's latest feedback research, Language Bias in Performance Feedback 2024 , was released a few hours ago. It not only confirms existing research but also offers significant new and more nuanced insights, enlightening us on how we can all improve in the challenging but vitally important dynamic of giving feedback.

Before my interview with Dr. Kieran Snyder of Textio, here are some of the report’s findings.

Previous Textio research has uncovered many of the feedback communication dynamics I’ve shared here and elsewhere, including:

1. Employees who don’t get clear feedback quit

Yes, among survey respondents planning to leave their jobs, an incredible 17% specifically named “insufficient feedback” as the primary reason they were doing so.

2. Black employees get 26% more unactionable feedback than non-Black employees

This is despite receiving only 79% as much feedback overall.

Today’s report takes it all to the next level. Let's start with gender bias.

Gender Bias in Feedback

Among other insights, the report highlights how language bias in performance feedback impacts women in terms of what they are told and what they recall being told:

Women also receive feedback about their personalities rather than their work 22% more often than men. In this regard, women are more likely referred to as “helpful” and “nice,” whereas men are more likely to be called “confident” and “ambitious.” As the report states:

“Not even being the highest performer exempts women from paying the personality tax.”

Lastly, as it relates to gender, women are 2x more likely to be referred to as an “overachiever.” Black women were described as overachievers 4x as often as white men.

Why does this matter? As the report states:

“Among instances of problematic feedback, overachiever is royalty. The label is often applied to strong performers transcending low expectations. Overachiever communicates an insult masquerading as a begrudging compliment.”

Racial Bias in Feedback

While previous Textio reports highlighted the distinct differences between, for example, the kind of feedback received by Black vs. non-Black employees, this report focuses on how different racial groups internalized feedback that aligns with the social stereotypes about their demographic identity.

In this sense, the research serves to confirm the role of stereotype threat in feedback. In other words, as the research found, people of color are more likely to be referred to with negative characteristics that align with their demographic group stereotypes. This takes a cognitive toll and means “these patterns have an especially destructive impact on these groups.”

Here’s a look at how critical feedback reinforces negative stereotypes across racial groups:

Additionally, feedback can also reinforce positive stereotypes. Note below how over half of East Asian people and nearly half of all white people can recall being described as intelligent, a dramatically higher percentage than Black and Hispanic/Latino people:

High Performers and Performance Feedback

The highest performers in a company surely get the best feedback, right?

Not quite.

Textio’s research found that while high performers get 1.5x more feedback than everyone else, they tend to get exaggerated feedback using words like “never” and “always” rather than the truer and more nuanced positive and negative feedback they need to improve.

Additionally, high-performers get 2.6x more "fixed-mindset feedback" than low performers.

Why is this problematic?

美国斯坦福大学 psychologist Dr. Carol Dweck, who popularized the fixed and growth mindset terms, states that a fixed mindset believes qualities are innate and "fixed" and, therefore, can't be further developed. So, high performers constantly receiving compliments about how they are "naturally smart" could begin to feel that their hard work isn't recognized or doesn't matter.

You can see the cross-section of fixed-mindset feedback and positive stereotyping play out here:

The report states:

“Given the research on stereotype threat, this contributes to a fixed-mindset mentality that makes it harder for people to escape social stereotypes.”

My Interview with Dr. Kieran Snyder

With a PhD in Linguistics and Cognitive Science from the 美国宾夕法尼亚大学 , and now over a decade of researching and publishing about how language bias shows up in feedback, Kieran Snyder is at the forefront of the field. Beyond researching and publishing on feedback communication challenges, she's the co-founder and Chief Scientist Emeritus of Textio , a company that provides AI-powered solutions that help over 100,000 people managers give less biased and more effective feedback.

Cameron Conaway (CC): Thank you for taking time out for us here at Feedback Tips Weekly , Kieran. Congratulations on your latest feedback report . I’ve been reading your research for years, and it’s an honor to interview you here. Let’s dive in.

CC: There are many different workplace communication dynamics worth studying. What initially led you to focus on feedback, and what has sustained your interest in this area over the years?

Kieran Snyder (KS): I first started studying feedback because I observed anecdotally that the high-performing women I worked with were more likely to be criticized for their personalities. I got curious about whether that was actually true. My entire academic career focused on using statistical techniques on large language corpora to find socially significant patterns, so I got to work collecting data.

When I first published insights about this in 2014, I found that women were in fact much more likely to get feedback about their personalities than men, who were more likely to receive feedback about their work. This was true even if the manager providing feedback was a woman. Over the years, numerous other researchers replicated these insights. Eventually, Textio found bias by race and age in addition to gender.

I continue to be interested in feedback disparity because it’s a leading indicator of problems to come. If we can catch bias in day-to-day feedback, then maybe we can get ahead of some of the really problematic stuff like pay and promotion disparities. Most organizations run pay equity analyses these days, which is great. But if you’re only catching disparities at that stage, a whole lot of things have already gone wrong. Fixing the feedback gap is a chance to get ahead of the bigger issues.

CC: Despite decades of academic research highlighting the critical role feedback plays in development and numerous popular business articles discussing the intense anxiety many managers feel when giving feedback, many talent development leaders who understand feedback's importance have told me their organizations still aren’t investing in comprehensive feedback training and tools. Why do you think this is the case, and are you seeing any changes in this regard?

KS: Last year, I met with an HR exec at a Fortune 100 company. They were complaining about how poor their managers' feedback skills were, regretting all the trainings that hadn't worked.

I asked a question:

"If a manager hits all their deliverables, but they never give feedback to the people on their team, does the manager get fired or promoted?"

The CHRO answered without hesitation:

"If they hit all their deliverables? They get promoted."

I said:

"Seems like we've found your feedback problem."

By and large, I am seeing two kinds of organizations. The first type is more traditional. Best case, they run annual feedback trainings for managers a month before performance reviews and hope that some of it sticks. By and large, these tend to be organizations that represent slow-growing or declining businesses.

But there’s another type of organization that has emerged in the last few years. The most rapidly growing organizations in most industries are reinventing the way they think about talent and people processes. On-demand coaching for managers is extremely common. The number of HR pros is limited, so increasingly, this means the organization is using software tools to scale.

CC: Your ongoing and uniquely accessible feedback research has been shared widely since your 2014 article on high achievers at Fortune . First, I’m struck by what a beautiful celebration it is for your career’s work that Textio’s study today returns to this topic 10 years later. But I also want to hear from you: How, specifically, can Textio help managers address some of the feedback challenges you’ve been bringing to our attention over the years?

KS: I have been studying feedback for a decade and I still sometimes catch myself getting it wrong. I still default into using always or never statements when I give feedback. I sometimes forget to provide specific examples. When I was managing ten people and moving fast, it was especially hard to get it right consistently.

We quickly realized that even at a small company like Textio, it was impossible for our HR team to coach managers on every piece of written feedback. Maybe HR will jump in if a manager explicitly asks for help, or in especially sensitive situations. But they just can’t be everywhere. We built software to help.

As you’re writing your feedback, Textio flags cases of personality feedback and exaggerations, instances where the feedback is unclear, and places where you really need to add an example. It also helps you translate your rough performance notes into clear, well-written feedback automatically.

Finally, and this is really, really important in the era of every HR Tech app pushing the APIs from ChatGPT into their app: Textio explicitly validates all its output for bias, and shows you the Textio Verified seal when feedback is trustworthy and safe to use. It’s not enough to help people write feedback. You actually have to make sure it’s high-quality and free from bias.

CC: Okay, although I highlighted many aspects of the report earlier, I want to get your thoughts on a few parts of it. The report opens with a focus on the feedback language used in the performance reviews of high-performing employees. As a feedback nerd with hundreds of academic papers saved in my Zotero, it’s clear to me that you are addressing a research gap here. But, beyond that, why else did you find this topic compelling enough to study?

KS: Funny enough, my very first paper on this in 2014 focused on high performers too! So it is a nice full circle moment. But this year, we focused on high performers simply because organizations find it so valuable to retain them. After all, independent research suggests that, across industries, high performers are up to 400% more productive than their peers.

In the 2023 Textio report , we saw that people who receive low-quality performance reviews are 63% more likely to quit within the next year, even when the feedback is positive. Given the importance of retaining high performers, this made us wonder how their feedback stacked up. And the 2014 data showed big differences between the feedback received by high-performing men and women, and we wondered if that was still true.

Unfortunately, we found that high performers get some of the lowest quality feedback. High-performing women get the lowest quality feedback of all.

CC: What did you find the most surprising about this report’s findings on high performers? Why?

KS: Is it bad that I was not very surprised?

I did find it interesting how much more feedback high performers receive — almost 50% more than everyone else. To me, that suggests that managers are investing more time in their high performers. But then when you look at the quality of the feedback that is actually being provided, it’s so much lower than what middle-of-the-road performers get.

CC: Your research shines a light on the many inequities that show up in how feedback is given, especially as it relates to feedback given to women and people of color. In your experience, what steps can individual managers take to address and mitigate these biases in their feedback processes?

KS: There are tricks managers can use, like writing feedback, letting it sit for a few days, and coming back to it with a fresh eye a week later. But in real life, busy managers are usually writing feedback right before it’s due, and few people have the space for this kind of edit cycle.

Similarly, managers can ask for a second opinion from someone they trust, HR or otherwise. People often ask for outside perspective, especially when the feedback is sensitive or difficult to provide. But again, this is not the typical case.

The top reason I have spent my career building software is that I deeply believe software is the best way to enable change at scale. Textio has a free trial you can use in your own upcoming performance feedback cycle. It’s fascinating to use it to spot your own biases in action.

CC: As Emily Peck at Axios and others have reported , backlash against DEI (diversity, equity, inclusion) efforts continue to intensify. Some major brands that appeared all-in on it are now divesting from it through layoffs and language (it’s now talked about less often on earnings calls). As the co-founder of an evidence-based company doing important work to address bias, including the types of biases many folks deny exist, I’m curious about your thoughts on what’s happening and where Textio can play a role in building a more inclusive future for all.

KS: One reason I love publishing data is that data can often speak for itself. Especially in tech, the field I come from, quantitative data is extremely valuable currency.

It’s one thing to say, “Women are experiencing bias in your organization.” It’s a very different thing to say, “In your last 5 years of performance reviews, women in your organization have received only 30% as much written feedback as men.” It’s just a lot harder to argue with the facts.

Textio was founded based on this principle and continues to operate on this principle. We show the real data. Whether you call it DEI or something else, few organizations want to be demonstrably unfair.

CC: Lastly, our subscribers here at Feedback Tips Weekly are a diverse group including CEOs, military leaders, talent development professionals, professors, scientists, marketers, artists, and athletes. What brings us all together here is a passion for improving our feedback literacy. What advice might you have for how we can keep improving?

KS: Especially for people with leadership scope, you just get to decide how important feedback is within your organization. Are managers truly accountable for providing it? What happens if they don’t? Do they get promoted anyway? Do they get fired? There’s nothing that increases feedback capability like having real, organizational accountability for doing so.

But even once those accountabilities are in place, we can all still improve our core feedback capability. Real-time practice sessions where you get feedback from someone you trust are a great idea, especially before difficult conversations.

Textio also has a free hour-long course you can take to get certified in how to provide equitable and effective feedback. This gives you great fundamentals to make sure that the feedback you give is fair and high-quality.

***END***

Team: I highly recommend taking Textio's Equitable Performance Feedback Certification Course . You can see that I've added it under the Licenses & Certifications section of my profile. It will take about an hour of your time but is well worth it. Click the link above or the image below.