When frontline employees are empowered to go off script in their interactions with customers, the result is not only happier customers. It’s also more satisfied employees. That’s one of the findings of “Frontline Employee Improvisation: Uncovering Its Meaning, Practice and Impact in Service Industries.” The paper, which was published in the Journal of Service Research, examines how and when customer-facing employees decide to improvise.  

“Customers are requiring more bespoke interactions,” said paper co-author Mayoor “Max” Mohan, associate dean for strategic affairs and research initiatives, and professor of marketing at the VCU School of Business. “It can be off-putting when a frontline employee is just sticking to the script and there's no wiggle room. That's really what motivated us to look at, what would it mean for employees to go off script and have more autonomy?”

Mohan sat down with us recently to talk more about it.

How are you defining frontline employees? Is it accurate to say their primary duty is customer service?

I think the more appropriate way to describe it is employees within an organization who have one-on-one interactions with customers. It could be a salesperson, but it also could be someone who’s there to resolve an issue. Also, it doesn't need to be in a physical retail setting. It could also be over the phone, via messaging tools, or through other avenues like that.

Why would employers care about this? It’s a tight job market – you know, do it my way or I can find someone else.

As long as organizations are concerned about the bottom line, they must recognize that it isn't simply made up of realizing efficiencies by having employees stick to the script. Longer-term strategies all advocate a customer-centric approach. When companies recognize that, they'll quickly realize that providing employees the autonomy to improvise as and when needed, especially if they're qualified to do so, will result in better outcomes in the long run.

Can employees be trained in how to go off script?

That's difficult because the way to figure out how to improvise, or train employees to improvise, very much depends on some core traits and abilities of the employees themselves as well as the organizational settings in which they work. One of the things we uncovered in our research is that someone who's new to a frontline career may not have had the body of customer-facing interactions to have built up an accessible archive or repository, if you will, of when and how they should pivot. The more experienced you are, typically you're going to be more effective at improvising.

Also, there are other dimensions to it. Improvising requires a level of confidence and faith, feeling like you're empowered to do something and not be afraid of repercussions if things don’t pan out because you improvised. It's not so much about an employer or an organization figuring out what that off-script scenario looks like as much as it is empowering employees by saying, you have what it takes to navigate these situations. You're allowed to do it.

Is sticking to the script ever effective?

You know, it's interesting. As I think through some of this, why were we sticking with scripts in the first place? What was the big advantage there? Some of it was a shortcut for training. It was easier to hand workers a script and now, you don't have to train them to effectively engage with your customers. Stick to the script and we solve 90% of our problems.

You know, it's interesting. As I think through some of this, why were we sticking with scripts in the first place? What was the big advantage there? Some of it was a shortcut for training. It was easier to hand workers a script and now, you don't have to train them to effectively engage with your customers. Stick to the script and we solve 90% of our problems.

What’s next in this line of inquiry?

Actually, I’m about to submit a paper to a journal that addresses the extent to which AI and agentic tools, as we call them today, drive service captivity.

Service captivity? That’s an interesting term.

So captive service experiences are those where customers feel like they don't have a voice, they don't have power, they don't have choice. With AI, sure, it works in many settings, but there are certainly situations where customers want a little more warmth—a human touch, if you will. They might want to explain their situation or explore more options and solutions. As of 2026, AI is probably not at the point where it can handle the myriad possibilities a customer might present in terms of challenges and nuances in relational exchanges. One of the things we talk about is AI's inability to improvise as effectively as human beings can.

Sounds like good news! Frontline employees don’t need to worry about being replaced by AI.

For me, the more interesting scenario to consider is not AI replacing frontline employees. It's how AI and humans can work together, think collaboratively, to better resolve customer issues. I think that might represent not only what’s realistically possible but also what’s better.

Maybe there’s a human frontline employee who’s augmented with an AI agent. They’re tag-teaming to resolve a customer’s issue. I can see a scenario where the AI agent is prompting the human to improvise in certain ways the human may not have thought about on their own. It’s like you're providing a tool to employees to make them more effective at their job.