If the thought of role-play exercises makes your hair stand on end, you’re not alone. But for a long time, it was the best way to give learners practical experience and build their confidence.
As Commercial Director in a previous business, I used role-plays all the time as a way to assess potential recruits. It made sense – I wanted to see how people would handle different scenarios in their roles. But the role-play exercises didn’t work for everyone, even those who, on paper, were perfect.
The pressure of ‘acting out’ a scene was too much for some people, and that pressure often got in the way of allowing potential gems to shine. The role-plays weren’t helping us to see people performing at their best.
On top of that was the fact that there was a team of us carrying out those exercises, adding extra variation and inconsistency into the mix depending on our own feelings on how or what to assess on that particular day.
It’s a truth universally acknowledged that role-plays are daunting at best and toe-curling at worst, especially for younger learners like those in Gen Z. So why do we keep using them, particularly within learning and development? And could there be a way to get all of the benefits of traditional role-play exercises without the cringe factor? AI could be the answer that many L&D managers are looking for.
First, let’s take a closer look at the issues with traditional, face-to-face role-play. Because it’s not all about the awkwardness – there are definitely wider issues at play.
But we’ll start with the obvious one – discomfort. Most of us have endured a role-play exercise at one point or another, and can agree that it can feel awkward and embarrassing. And while it can be tempting to tell people they just need to ‘get over it’, we know that the feeling of being watched in and of itself can actually change behavior.
We know when we’re being ‘observed’ and tend to act and speak differently when it’s happening.
So the discomfort of feeling watched isn’t necessarily something you can just wave away. It means that, when you’re running role-play training, you’re not necessarily getting an accurate picture of how each person will behave during the real thing, which somewhat defeats the point.
There are two potential sides of inconsistency in role-play training; inconsistency in inputs and of outputs.
Inconsistent inputs are all about how the role-play is conducted. In my experience, it’s unlikely that everyone asks the same questions in the same way every time. It’s not just the trainees thinking on their feet in these situations; the people running role-plays can have a huge influence on the outcomes depending on how they react each time. It’s just part of the human condition – but it’s not ideal when you want to give everyone the same level of training.
Inconsistent inputs are bound to lead to inconsistent outputs, meaning subjectivity when assessing the outcome of the training, varying levels of skills and knowledge across your team, and – inevitably – variation in customer experience, which is exactly what you want to avoid.
Then, of course, there’s the cost of role-play training, either in terms of the actual monetary value of hiring an external trainer to come in and run sessions with your team or in the lost hours of taking senior and junior staff members off-task at the same time to complete the training.
And remember, this kind of training doesn’t have a lasting impact if it’s only completed once. It needs to be repeated regularly to keep the information fresh, meaning we’re not talking about a one-off cost but a regular investment.
Even if you were willing to pay for regular role-play training for your teams, how do you know it’s working? You might know how many people attended the session, but how many actually boosted their knowledge and skills as a result? Can you quantify how much they improved? Having these kinds of impact stats is invaluable to justify continuous investment in high-quality learning and development for your teams, but it’s not easy to get from traditional role-play training.
If role-play’s so bad, why do we keep using it?
I want to be clear here – the idea behind role-play training is absolutely solid. We should be giving people opportunities to practice using their skills in a risk-free setting before they have to apply them on the job.
Plus, in an independent survey of over 2,000 employees in the US and UK, 75% said they’ve experienced anxiety around difficult conversations in the workplace, and 72% would welcome training on how to deal with them (rising to 87% among Gen Z).
So there’s a clear need for this kind of practical learning to develop confidence and skills around interpersonal interactions at work. And we know that role-play can produce tangible results when done right. But there needs to be a more scalable and impactful way of doing it.
Advancements in AI technology and simulation-based training mean that organizations can now allow their teams to practice their skills anytime, anywhere – no role-play trainer required.
Realistic simulated environments and AI-powered avatars give an immersive experience that engages learners with active participation while providing a safe place to learn from their mistakes as they move through their learning and come back to it as and when they need to.
Built-in procedures for instant, personalized feedback and data capture can provide individualized support for each user as well as an overall view of impact for learning and development managers.
Better yet, digital learning like this is fully scalable to give every participant a consistent experience no matter when or where they engage with their training.
In 2024, Attensi launched REALTALK, the first role-play training solution to incorporate Agentic AI and AI free-form dialogue with realistic virtual humans. REALTALK takes AI role-play training to a whole new level, with conversations that feel as natural as talking to an actual human.
It gives learners an even more realistic environment to practice interpersonal communication skills (with real-time practical feedback) without the drawbacks of traditional role-play exercises.
The great news is that people really don’t seem to mind where their training comes from as long as it’s relevant and challenging. When asked whether they’d prefer training from their manager or an AI role-play trainer, employees were evenly split either way, and almost a quarter said they had no preference. Over half even said they’d welcome AI training to help them improve their human skills. It’s time for leaders to act on that and implement AI-powered training that works.
Gone are the days when we needed to rely on expensive, inconsistent and downright awkward role-play exercises to give learners the chance to practice difficult conversations and interpersonal scenarios.
The range of AI training solutions now makes it easier than ever for trainees to get the practical experience they need to succeed in their role – no cringe required.
Author Rich Angus Business Development Director | Attensi
