For years, retail training has followed a familiar sequence: teach the standard, show the behavior, ask a few questions, and mark the course complete.
That works well for knowledge. It works far less well for conversations.
Anyone who has spent time in stores knows the difference. An associate can understand the steps for handling an objection and still freeze when a customer pushes back. A manager can complete a course on performance coaching and still struggle to find the right words when an employee is sitting across from them.
The issue is not always knowledge. Often, it is practice.
That is why AI role-play is getting so much attention right now. There are more tools entering the market that let employees rehearse conversations with an AI coach. That is a useful development, but we believe the bigger opportunity is not to create another coaching destination.
The bigger opportunity is to make practice part of the training itself.
The gap between learning and doing
Training transfer has been studied for decades. At its simplest, it asks whether people actually apply what they learned when they return to work. A recent systematic review of workplace learning research describes transfer as the effective application of knowledge, skills, and attitudes from training into the job itself.[1]
That distinction is especially important in retail.
Completion tells you someone opened the training. A quiz tells you they recognized the right answer. Neither tells you whether they can perform the behavior when the customer is impatient, the line is long, or the manager has to deliver feedback that may not be well received.
Research on deliberate practice has consistently emphasized focused repetition, feedback, evaluation, and opportunities to try again as important ingredients in skill development.[2] Research on professional training also points to the value of learner-centered approaches that include practice situations, feedback, and role-play.[3]
Retail has always known this intuitively. We used to do live role-plays in classrooms, regional meetings, and manager workshops because there was value in saying the words out loud. The problem was scale. You needed a facilitator, another participant, enough time, and some consistency in how the exercise was evaluated.
AI changes that equation.
Practice should happen while the learning is still fresh
Our view is that role-play is most valuable when it is connected directly to the training that introduced the behavior.
If you just taught a new service model, let the associate practice it. If you just trained a manager on performance feedback, put that manager into the conversation. If you just introduced a new selling behavior, do not stop at asking whether the learner knows the right response. Let them try it.
This is where INCITE Coach fits into the learning experience. It is an activity inside INCITE, not a separate coaching platform employees have to remember to visit later.
A training path can move naturally from instruction to practice and then into reinforcement:
Learn → Practice → Feedback → Apply → Reinforce
That last step matters. One role-play is useful. Repeated practice over time is more powerful. Simulation research has suggested that repeated practice can strengthen communication and decision-making skills, particularly when learners have opportunities to receive feedback and try again.[4]
That means the same behavior can come back later in a new scenario. The associate may encounter a tougher customer. The manager may face a more defensive employee. The skill stays alive after the course is complete.
The character matters more than people think
One of the most important lessons we have learned over decades of producing retail training is that a scenario is only as good as the person inside it.
A customer cannot just be “the difficult customer.” A manager cannot simply be “the unhappy manager.” Good training characters need a reason for behaving the way they do.
Maybe the customer is not really objecting to price. Maybe they are worried about making the wrong choice.
Maybe the employee whose performance has slipped is not disengaged. Maybe they feel they have stopped receiving useful feedback.
Maybe the associate pushing back on a schedule change is dealing with something the manager does not know yet.
That depth is what gives the learner something to discover.
It also creates a very different kind of practice. If the learner jumps too quickly to the answer, the character may shut down. If they ask the right question and actually listen, the real issue can surface.
That is much closer to what happens in a store.
AI voices have changed dramatically
This is another area where the technology has moved much faster than many people realize.
A few years ago, synthetic voices were easy to spot. They could read a script, but they often sounded flat. Emotion, hesitation, tension, and personality were difficult to create convincingly.
That is changing quickly.
Recent research in speech synthesis has moved beyond basic natural-sounding speech toward increasingly fine-grained control over emotion, timbre, style, and prosody.[5] Other work is exploring more nuanced emotional expression, including intensity and combinations of emotional states.[6]
That matters for training because conversation is not just words.
A customer can sound skeptical. An employee can sound defensive. A manager can sound frustrated. Someone can hesitate before answering because there is something they are not saying.
Those cues create context. They make the learner listen differently.
For role-play, that is a major step forward.
From AI presenter to AI character
This is where MMP AI Studio and INCITE Coach come together in a way we find especially interesting.
We have spent a large part of our history creating characters for training. In the past, that meant casting talent, finding locations, coordinating wardrobe, shooting scenes, and then building the learning around the footage.
MMP AI Studio changes how quickly those characters can be created.
A company can choose the talent, wardrobe, environment, voice, and scenario. The character can appear in a branded video that sets up the situation.
Then the video ends, but the character does not.
Instead of answering a multiple-choice question about what they just watched, the learner talks to the character.
The customer they just saw can become the customer they now have to help. The employee from the manager-training video can become the person the manager now has to coach.
That continuity gives the character a life beyond the video.
The AI is no longer simply presenting training content. It becomes part of the training experience.
This is bigger than sales coaching
Sales is an obvious use case for AI role-play, and it is an important one. Associates can practice discovery, objections, recommendations, and closing behaviors without waiting for a real customer.
But retail has plenty of conversations that have nothing to do with closing a sale.
Some of the most important ones happen between managers and employees.
A manager needs to address declining performance without putting the employee on the defensive. A high performer is losing interest. A new manager needs to reinforce a standard with someone who has been with the company longer than they have. Two associates are creating tension on the team.
These conversations affect turnover, engagement, execution, and customer experience.
We train managers on them all the time. Historically, we have had far fewer ways to let managers rehearse them repeatedly.
INCITE Coach is designed for sales, service, peer, and manager conversations because all of those interactions shape frontline performance.
The same idea can extend beyond stores as well. Corporate managers also need practice delivering feedback, handling conflict, coaching employees, and navigating difficult conversations.
The technology does not care whether someone works behind a cash wrap or at headquarters. The instructional challenge is the same.
Feedback has to reflect your standard
Another lesson from years in retail is that generic feedback is not enough.
“Show more empathy” may sound helpful, but what does that mean for your brand?
“Ask better questions” is not a coaching standard.
Retailers have their own service models, selling behaviors, leadership expectations, policies, and language. Practice should be evaluated against those standards.
INCITE Coach allows scenarios to be scored against the retailer’s own behavioral rubric. The debrief can point back to evidence from the conversation, giving the learner and manager something specific to discuss.
That is important because the goal is not to create a high AI score.
The goal is to improve what happens on the floor.
Research on mastery learning and deliberate practice reinforces the importance of defined performance standards, feedback, and repeated opportunities to improve.[7]
The manager should still be part of the equation
Whenever AI enters the training conversation, there is a temptation to position it as a replacement for something.
We do not see INCITE Coach as a replacement for managers.
Good managers are still essential.
What AI can do is give employees more opportunities to practice and give managers better information to coach from.
Instead of spending the entire coaching conversation trying to determine whether an employee understands the behavior, the manager can focus on what needs to improve.
That is a better use of manager time.
The real opportunity is reinforcement
For us, this may be the most important part.
Training should not end when somebody clicks Complete.
Retail changes too quickly for that. Products change. Promotions change. Service expectations evolve. Managers change. Employees encounter situations the original training could never anticipate.
INCITE Coach creates a way to bring key behaviors back after the initial course.
A new scenario can appear a week later. A more difficult version can follow after that. A manager can reinforce the same skill with the employee.
The behavior remains active instead of disappearing into training history.
That is the difference between using AI role-play as a standalone coaching tool and using it as part of an ongoing learning strategy.
AI can finally give everyone a chance to rehearse
Retail has always been a performance business.
At some point, the associate has to speak to the customer. The manager has to have the conversation. The standard has to move from a screen into real behavior.
For years, we have been able to train people at scale. Practicing at scale was much harder.
That is what has changed.
AI now gives us the ability to create believable characters with more emotion, personality, and depth. It gives employees the opportunity to speak, listen, react, and try again. It gives organizations a way to connect that practice directly to the training and bring it back later as reinforcement.
That is what we find exciting about INCITE Coach.
Not AI for the sake of AI, but a better bridge between what people learn and what they actually have to do.
Learn it. Practice it. Reinforce it. Then use it when it counts.
About INCITE Coach
INCITE Coach brings spoken AI practice directly into INCITE®, allowing associates and managers to rehearse sales, service, peer, and management conversations as part of the learning experience.
Through MMP AI Studio, organizations can create and customize AI characters, voices, scenarios, and branded video introductions, then bring those characters into interactive practice inside INCITE.
Learn more: https://multimediaplus.com/incite-coach/
Footnotes
1. Blume, B. D., et al. Research on training transfer and the application of learned knowledge, skills, and attitudes to workplace performance. Human Resource Development Review / related systematic review literature, 2025–2026.
https://www.sciencedirect.com/science/article/pii/S136656262500035X
2. Ericsson, K. A. “Deliberate Practice and Acquisition of Expert Performance: A General Overview.” Academic Emergency Medicine, 2008.
https://pubmed.ncbi.nlm.nih.gov/18778378/
3. Gegenfurtner, A., et al. Meta-analytic research on learner-centered professional training, including practice, feedback, and role-play. Educational Research Review.
https://www.sciencedirect.com/science/article/pii/S1747938X11000157
4. Simulation-based communication training literature reviewing repeated practice, feedback, and skill development in realistic scenarios. PMC / National Library of Medicine.
https://pmc.ncbi.nlm.nih.gov/articles/PMC8990192/
5. Recent survey research on controllable text-to-speech systems, including advances in emotional expression, prosody, timbre, and speaking style. Proceedings of EMNLP 2025.
https://aclanthology.org/2025.emnlp-main.40/
6. Research on fine-grained emotional control in expressive speech synthesis, including emotion intensity and mixed emotional states. arXiv, 2024.
https://arxiv.org/abs/2410.00316
7. McGaghie, W. C., et al. Research on deliberate practice and mastery learning emphasizing measurable standards, feedback, and repeated performance improvement. Medical Education / simulation-based mastery learning literature.
https://pubmed.ncbi.nlm.nih.gov/26375272/
