What if the real opportunity of AI in performance management isn’t automation, but redesign?
In a recent meetup we hosted, Sophie Rustichelli and James Hayward from Experian shared how they approached one of the most persistent challenges in HR: making performance management actually matter.
Not in theory — but in day-to-day work.
They started with a simple but powerful observation: many employees didn’t have clear goals. And even when they did, those goals weren’t always connected to what the business was trying to achieve. A familiar story for most HR leaders.
So instead of asking…
They asked a better question:
“Where is there a real business problem that AI could help solve?”
“There’s a real temptation to get caught up in shiny new AI tools. We found it more effective to start with the problem first, and then decide on the solution which, in our case, happened to be AI!”
What does an agile approach to AI development involve?
They developed an AI-powered performance coach, a tool designed to help employees translate organisational priorities into meaningful, actionable goals, and stay on track with them over time.
A great tool, but that wasn’t necessarily the most impressive part.
What really sets this apart is how they built this tool with an agile and human-centric user approach.
Important factors in the build
They worked in short, focused sprints.
They used real employee input from the start.
They iterated quickly, often showing unfinished, messy versions of prompts to leaders and stakeholders.
They built the solution in collaboration, not in isolation, with employees, leaders, and even the chosen AI coach vendor.
There was also a moment that felt particularly telling. A leader challenged them: why not just use tools like ChatGPT or Claude? Why invest in building something more specific?
“I have always struggled with the "messy prototype" part of Agile, but with some simple prompting you can set up an AI bot however you like and test with your users. Then refine your prompt and test again!”
Instead of debating, they demonstrated.
They ran a live comparison, same question, different tools.
The difference in output was clear. The tailored AI coach, designed with context and purpose, produced significantly more relevant and actionable guidance.
It was a simple but powerful reminder: AI on its own is not the solution. Design and context are.
TOP TIP: Don’t be afraid to iterate in front of leaders. It’s a powerful way to demonstrate your ability to deliver value quickly, while drawing them in to engage with and buy into your approach and project.
What was the impact of this approach?
Within four months, the percentage of employees with defined goals rose from 60% to 90%. But even that doesn’t fully capture what was happening.
This wasn’t just a technology rollout. It was a change process.
Employees tested the tool. Some tried to break it. Others pushed its boundaries. There were mistakes along the way. But because the team worked in an agile way — iterating continuously, improving through continuous loops — they were able to respond and evolve quickly.
What’s more, this work didn’t sit within one team. It brought together HR expertise, analytics, and AI capabilities.
As they delivered, teams were building cross-functional ways of working, opening up new avenues and opportunities for future collaboration.
“This really reinforced something I see time and time again: AI in HR is not a technology project. It’s a behaviour change project. And that requires a very different mindset from the start.”
Another important decision they made was to stay focused.
Whilst people quickly saw other potential use cases for the AI coach, they intentionally ring-fenced it to performance management. That clarity helped drive both quality and adoption.
What does this mean for HR leaders?
What Sophie and James vividly brought to life was a shift that is already underway.
Performance management is moving away from static processes and annual cycles, towards something more dynamic, continuous, and embedded in everyday work.
AI can enable that but only if it’s implemented in a way that connects directly to real business outcomes.
This story is just one example, but it raises important questions that are important to reflect on…
Questions to reflect on...
- Are we focusing on the right problems?
- Are we moving fast enough to learn?
- Are we designing solutions with — not just for — our users?
The full conversation went much deeper into how this was built, challenged, and evolved in practice.
👉 Click here to learn more from Sophie and James at our next meetup.
Sophie Rustichelli
Senior Talent Manager at Experian
Sophie Rustichelli is a Senior Talent Manager at Experian, where she combines a passion for developing people with a love of innovation. With a career spanning global talent development, executive coaching and performance management, Sophie is energised by helping teams and leaders grow in fast pace environments (ideally the kind with coffee). When not working, she’s usually planning her next adventure, whether that’s scuba diving, skiing, or exploring a new city.
James Hayward
Global Learning Technology Manager at Experian
James has spent 20 years in L&D making it easier for people to learn the skills they need for their current and future roles. As Experian's Global Learning Technology Manager he oversees all learning platforms and the overall learner experience. He loves learning new things and has recently become one of Experian's volunteer beekeepers! Outside of work he is a dad of two, a Scout and Cub Leader and has been a primary school governor for 16 years.
Kate Rand
Agile HR Coach & Trainer and CEO at Serve Legal
Kate is Chief Executive Officer at Serve Legal and an award-winning HR executive. Kate is an experienced Agile HR Trainer and Coach and has worked with the likes of Netflix, BMW and GSK. Kate was also one of Agile HR Community's very first certified practitioners!
