Feature
Well-rounded, fair employee performance reviews involve troves of data from communication platforms, collaboration tools and emails from clients, co-workers and managers. Compiling this information can be cumbersome, so many leaders are turning to generative AI to help.
When it’s time to start preparing performance evaluations, too often, people are “starting from scratch and scrambling for information,” said Jeff Smith, head of product at 15Five, a performance management platform.
AI can help avoid this. We looked into how and why you may want to use Gen AI for performance reviews.
Streamlining Data Collection for Performance Reviews
AI technology helps HR leaders and team managers to continuously collect information to highlight employees’ work and achievements, whether that’s via emails, Slack messages, data from Salesforce or other customer relationship management software, Microsoft Teams or Zoom meetings, Smith said.
Pulling together data with AI can also prevent human error and create efficiencies, allowing organizations to seamlessly reward high-achieving employees and identify areas where improvements are needed, added Patrick Flynn, assistant professor of human resource management at North Carolina State University.
A constructive review is crucial for maintaining valuable relationships with employees. Research from Gallup shows performance reviews fail to inspire or motivate most workers to improve.
AI also enables the continuous collection of data in real-time over longer periods of time, which can promote fairness and prevent recency bias, or giving preference to recent events over historic ones, Flynn explained.
“Then, we can have quicker, more impactful, more regular performance evaluation conversations,” he said. This creates opportunities for organizations to identify problems and make adjustments in real-time, rather than waiting until the next review — which might create a “performance deficit” for the worker.
Related Article: What Defines a Modern Performance Management System
Driving Efficiencies … and Engagement
Velocity Capital Group, a funding platform for small businesses based in New York, uses AI to track the progress and efficiency of its underwriting teams, including how many transactions employees handle compared to others, said Jay Avigdor, president and CEO.
It also uses AI to collect data on how its 100-plus employees adhere to security and privacy protocols, and to compare employee productivity levels to the hours they log.
All of these details are incorporated into an employee’s review, which is conducted about every six months, Avigdor said. “It’s a click of a button, and it saves time. It saves money. It saves headaches.”
Efficiency is the key benefit of using AI for performance data collection, Flynn said. It frees up managers and HR teams to build connections directly with employees, such as via coaching, mentoring or goal-setting.
Automation facilitates “better performance conversations,” Smith said. Employees will feel like they’re part of a fair process that rewards and recognizes them and offers opportunities to help them grow.
Related Article: Using Generative AI to Write Performance Reviews? Not So Fast
7 Tips to Incorporate AI in Your Performance Evaluation Process
There are use cases for AI that remain relatively new and may not be widely understood, Flynn said. For those seeking to use the technology for performance review information collection, he shared some tips for getting started:
1. Set goals and parameters
Setting specific goals and metrics for performance evaluations is vital for identifying what data is relevant, Flynn said. Have firm metrics in place guiding what leads to raises, promotions or disciplinary actions, and automate the collection of the data to support these metrics.
For instance, if you tie performance to employees participating in regular training, develop an AI system to gather data from learning management systems or communication tools, like email, showing that training was completed.
2. Choose an AI tool
While handling this process alone comes with a learning curve, organizations can set up their own data collection using ChatGPT or another generative AI platform. Some small HR teams may prefer to partner with outside vendors or consultants to help them through the process.
Whatever path or tool you use, make sure it’s helping you “drive the outcomes that you’re looking for to drive in your business,” Smith said.
Related Article: Internal AI Platforms May Be Desirable, But They're Not for Everyone
3. Get educated
There's an AI skills gap, Flynn said. Not understanding how to use AI or how to analyze the insights it generates could inadvertently lead to bias, so HR teams may want to consider an AI certification program or other related training.
Conference
Sep
16
HR Tech Conference Las Vegas 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
One of the biggest challenges, according to Smith, is understanding the nuance of “prompt engineering,” or what information to input into the AI system to generate the data you need. “You need to ensure everyone’s using the same prompts because the quality of the prompt could influence the quality of the data,” he said.
Organizations also need guidelines for interpreting data consistently and equitably.
4. Keep employees in the loop
Be as transparent as possible about how you’re using AI to compile information, what data is being collected and what goes into the decision-making process for performance reviews, Smith said.
And be available to answer employees’ questions and hear their concerns. Transparency and openness build trust and may give employees a more positive impression of performance reviews, he added.
Related Article: Why It's Time to Overhaul Our Performance Management Processes
5. Know the legalities
Along with telling employees about how you’ll gather data using AI, Flynn suggests also getting their consent prior to doing so. States like New York and California have laws requiring consent and transparency around what employee data is collected and how it’s used.
The European Union’s General Data Protection Regulation (GDPR) also offers guidance on fairness, transparency, integrity, confidentiality and accountability for employee data.
Staying up-to-date on these rules and regulations — and complying with them — is imperative, particularly as more laws and mandates are likely to follow as AI use becomes more widespread, Flynn said.
6. Constantly re-evaluate the AI process
Automating performance data collection isn’t a one-and-done thing, Flynn said. The process should be consistently reviewed. Check how the system is working, how it could be improved and how it’s changed your business and employee performance overall.
Be flexible and keep re-evaluating what’s truly driving performance and what data you need to support it. “You might come to the conclusion that factor X is actually driving performance more than we thought so we should devote more attention to that,” Flynn said.
7. Don’t lose the human touch
Despite the efficiencies created by automating data collection, Smith emphasizes that you still need a human element.
“Human plus AI is the most powerful combination,” he said. “You want to be capitalizing on what AI does well, which is looking at a wide range of signals, and then, creating insights, providing guidance and coaching throughout a review process.”
Smith added that performance feedback is best delivered in an individualized way that recognizes achievements, praises positive behaviors, addresses problems constructively and suggests improvements. This promotes fairness across the organization, something Flynn believes will better motivate employees than giving them a strictly automated performance review.
“No matter how much AI and automation you have, no successful business will ever be as successful if there’s no human integration,” Avigdor said.
