Here’s what happens when AI takes over performance reviews

There are a few workplace topics that consistently bring out strong feelings, and performance reviews is at the top of that list.

While most people would agree that it’s a good thing to have a tool to measure how employees are doing at their jobs, and a time for managers to discuss career advancement, very few seem to think that the way performance reviews are currently set up is working.

In fact, a Gallup survey last year found that only 2% of human resource officers at major companies think their performance management system is working and just 22% of workers felt their review process was “fair and transparent.”

One of the biggest complaints employees have about performance reviews is that they are so subjective. What it takes to be considered good at your job or eligible for a raise or promotion is often down to the opinions of just a couple of people. That means it’s a fertile ground for bias.

So, if both employees and leadership think performance reviews are broken, could artificial intelligence be the magic bullet that fixes it? The new crop of startups selling AI-powered performance management tools certainly thinks so. But is it just swapping human bias for AI bias? Is something as nebulous as being “good at your job” quantifiable? And are humans ready to be evaluated by a robot?

On the latest episode of The New Way We Work, I spoke to Bryan Ackerman, head of AI strategy and transformation at the management consulting firm Korn Ferry.

He explained the benefits and drawbacks of using technology in both performance reviews and layoffs. 

Ackerman says that when considering introducing AI into performance reviews it’s important to start by asking the right questions: What is the fundamental thing we’re trying to change? There are lots of pain points with performance reviews, but AI isn’t suited to fix them all. Here’s what it can do.

How AI can help the performance review process

Efficiency: One of the most straightforward ways that AI can help improve the performance review process is efficiency. Ackerman notes that managers can use generative AI to help draft reviews from their notes, but it’s only good as a starting point. Since the manager still needs to edit the draft to make it meaningful, that can end up not really saving any time.

On both the employees and manager side, AI is good at putting in data and quantifying it (for example sales numbers) but again, it’s subject to the quality and accessibility of that kind of data. 

Making reviews more understandable: One of the biggest issues with performance reviews is how subjective and arbitrary ranking systems feel. Ackerman says there’s potential for AI to help employers to standardize and be more transparent about rankings and use AI note taking apps to help serve as a jumping-off point for conversions. 

Assistance with career development: Ackerman thinks the most useful way that AI could help in performance reviews is by using it as a way to get reviews back to their original intention: as a career development tool. AI has the potential to help managers deliver better and more effective feedback, he says.

What AI can’t help with

There are pitfalls to using AI in performance reviews, Ackerman says, especially if you’re relying on it too heavily. It can give you too much data to work with, or “create more of a problem than it solves, if the data quality isn’t great,“ he says. “The question [is] are we making a manager’s life easier or harder?”

Relying blindly on AI also has the potential to exacerbate the problems with performance reviews. “Is adding more data into the mix, making this process easier and more efficient and more consistent and safer and [with] less bias? Or is it just adding complexity that then the manager is somehow supposed to still make sense around?” he says.

The bottom line is to not remove the human conversation and nuance from the process and instead use AI as a tool to help where it can.

Listen to the full episode for more on the pros and cons of introducing AI into the review process, where he thinks things are going in the next few years, and if AI will be used in layoff decisions.

You can listen and subscribe to The New Way We Work on Apple Podcasts, Google Podcasts, Stitcher, Spotify, RadioPublic, or wherever you get your podcasts.

View the full article