Transforming the conversation: How AI is reshaping performance feedback

In the previous blog, we covered how to ensure fairer outcomes with tools like Ratings Preview. In this final part of our series on getting performance management right, we’ll explore the most critical step: empowering your managers to turn that fair process into a great coaching moment with an AI performance management coach.

Great leaders are memorable because their impact is so profound. The people they helped grow remember how they delivered insights with care, providing feedback that was personal, tailored, and unique. They didn’t just evaluate our work; they showed us a path to becoming better. Day-to-day work fades over time, but a thoughtful and well-delivered performance review can build trust, unlock potential, and have a lasting impact.

Ensuring every manager can consistently deliver meaningful reviews is one of the hardest parts of building a thriving, high-performing culture. The challenge isn’t a lack of effort. It’s the inherent difficulty of managing the bias, subjectivity, and inconsistency that can creep into any review process. We know from our own research at Culture Amp that effective feedback improves employee performance. Shockingly, our research also revealed that the inverse is also true: In fact, ineffective feedback has a worse effect than no feedback at all.

Ensuring every review meets a high bar has, until now, been hard work. New managers may let the fear of the blank page hold them back, or they may feel unsure how to phrase constructive feedback. For tenured leaders with large teams, it’s a battle against time and recency bias, as they try to synthesize an entire year of contributions for each person. This cognitive load sometimes manifests in different ways; their reviews may be rushed, lower quality, or offer ineffective feedback. Sometimes leaders are overwhelmed by not knowing where to start, so they put off a task that is critical for their team’s success.

The unseen risk of generic AI

Faced with these challenges, it’s no secret that some of your leaders are likely already turning to tools like ChatGPT for help. But this introduces serious risks. First, the result is often generic, robotic feedback that sounds nothing like your manager and isn’t grounded in any best-practice coaching principles.

More importantly, it poses a significant data security risk to your company. When managers input confidential details about an employee's performance and experiences into unapproved tools, that sensitive information is no longer protected by enterprise-grade privacy controls, leaving your people and your company vulnerable.

Helping good managers to become great coaches

That’s why we built AI Coach in Perform. It’s not just another AI writer. It’s an AI performance coach and partner, designed by our team of experts in organizational psychology and machine learning, trained on insights drawn from 1.3 billion data points gathered over a decade of research into the world’s top-performing teams. AI Coach in Perform is crafted specifically for the art of performance feedback with your company's context and security at its core. Think of it as giving every one of your managers a Culture Amp people scientist in their pocket.

A partner, not a robot

Unlike other AI tools that simply write for you, AI Coach in Perform is designed to co-create with you.

Writing a performance review has traditionally been a solo task, often done in isolation – but it doesn't have to be. We've designed AI Coach in Perform to be a confidential thought partner for your managers. It’s a safe space where they can bounce ideas around, crystallize their thoughts, and get objective input, all without the pressure of having to be vulnerable with another person before they're ready.

Here's how that partnership comes to life:

1. Solving for recency bias

Our AI Highlights & Opportunities feature consolidates feedback sources from the entire review period, surfacing key themes and achievements. It tackles the time-consuming prep work of gathering data, giving the manager a balanced picture to start from.

2. Conquering the blank page paralysis

From there, AI Coach helps conquer the blank page by co-creating a first draft that is rich with detail. As a manager writes and refines their thoughts, our AI Suggest Improvements feature acts as a real-time guide, offering ways to make feedback more specific, impactful, and fair.

3. Adapting to varying manager needs

Throughout this co-creation process, AI Coach’s conversational interface learns a manager’s unique tone and style from their past writing, so the final review feels authentic and personal. It also tailors the level of support it offers, providing more hands-on drafting help for managers who want it, while letting experienced ones take the lead. It reduces the time spent pulling together data sources and crystallizing thoughts, and always keeps the manager as the author.

Smart tools, human decisions

We believe the most important decisions should always be made by people. That's why we've built in controls to ensure human oversight is at the heart of the process.

AI Coach never submits a review on a manager's behalf. It can help draft the narrative, but the manager is always the one to press the final button. Crucially, it will never suggest an answer to a rating question, like a performance rating or readiness for promotion. Instead, it will act as a confidential sounding board, coaching the manager to reflect on the available evidence so they can make a fair and confident decision themselves. Those critical judgment calls remain firmly in human hands.

This co-creation process ensures managers are reflecting and providing their own crucial context, giving you the efficiency of an AI performance management coach without sacrificing the human element that makes feedback meaningful. The result is a fairer, more objective set of performance insights that gives your organization greater confidence to shape high-stakes talent and remuneration decisions.

How to get started

Introducing a new AI tool can feel like a big step, but it doesn’t have to be. Perform’s AI features are toggled off by default, putting you in control. We recommend creating a test performance cycle with a pilot group of managers to build confidence and familiarity, clearly communicating the purpose: this is a tool to support their judgment, not replace it.

Depending on the AI readiness of your company, you may choose to enable AI Suggest Improvements and Highlights & Opportunities for your next performance cycle, and then assess the results before introducing Coach in Perform. Our team is available to support you with guidance on communicating any upcoming feature changes to your organization, from initial rollout through to implementation.

If you’re ready, please reach out to your Customer Success Manager today to schedule a planning call.

The next chapter for performance management

AI Coach in Perform is more than just a feature update. It’s the next chapter in performance management, shifting from a process to a partnership that helps your people grow.

Ready to experience the future?

Early access to AI Coach in Perform is now available. You can register to join our Early Access Program (EAP) and be among the first to put a Culture Amp People Scientist in every manager’s pocket.

Register now

New to Culture Amp? See how our performance management platform can help you turn an outdated performance process into a high-impact practice. Request a demo today.