Zenarate AI Coach Blog

How to Train Skills Not Scripts- Zenarate’s AI Simulation Training Methodology

Every day we work with enterprise brands helping them develop high-performing agents and incredible customer experiences. You could say we’re obsessed with it. And if there’s one thing we’ve learned through all this time spent building and working with world-class brands, it’s that the focus of any successful training methodology needs to be centered on developing agent skills, not scripts and keyword memorization.

Yet so many programs and solutions on the market today perpetuate the use of scripts and keywords as the backbone of their training. Whether it’s in traditional formats like videos and in person role play or even in newer simulation training solutions, scripts are commonly seen as the simplest default for trying to teach contact center agents how to do their jobs. Just tell them what to say, right?

At Zenarate, we believe that if you truly want to improve outcomes and customer experiences, it’s more important to train and develop your agents’ skillset around the unique parameters and requirements of YOUR business.

It’s exciting to see so many signals in the contact center space showing training and development leaders turning to AI simulation training to improve both the speed and effectiveness of their training. However, not all AI simulation training solutions are built the same way. Some are simply taking those traditional methods of script memorization and keyword analysis and automating it, with corrections and feedback delivered after the simulation is over. We believe that to truly simulate customer interactions and teach agents lasting skills, every word and intent by the agent matters and real time in-the-moment coaching throughout the simulation is critical.

So, what does that look like? Here’s a simplified look at our unique methodology that applies conversational AI, Natural Language Understanding (NLU), and Natural Language Models (NLPs) to create the most life like AI Simulated Training.

practice e1713293927712

  1. Learn It with Best Practice Examples: Agents listen to and see a best practice example.
  2. Practice It with Guided Roleplay: Agents receive specific guidance for each response and skill needing to be used throughout a conversation. With integrated Conversation AI and Screen Simulation, AI Coach verbally guides & coaches the agent through both your conversation and screen navigation best & required practices. Agents learn with their own words and natural way of speaking, not reading or memorizing scripts or keywords.
  3. Master It with Unguided Roleplay: Agents practice conversation and screen navigation without upfront guidance but receive in the moment coaching on soft skills along with best & required practices coaching throughout their roleplay. AI Coach provides real-time coaching if an agent’s use of conversation or their screen misses a best or required practice (for example, a soft skill like empathy wasn’t used, an account wasn’t properly verified, or a screen step was missed)

To make sure you really understand the difference between keyword-based simulations and intent, skill-based simulations, I’ll explain in more depth. With other AI simulation training solutions (keyword-based), as agents go through the simulated conversational roleplays, their systems are trained to match identified keywords for that scenario and once those keywords are spoken, the learner receives full credit for mastering the call. The problem is that anything else the agent says, other than the keywords, doesn’t get weighed or analyzed as part of their performance, failing to understand the agent’s true intent and tone.

For example, if I were a new agent simulating a customer call in one of those tools that required empathy, I would simply be measured on the recognition of keywords such as “I’m sorry” and “difficult.” If I responded to the customer with “I’m sorry to hear about your difficult situation, but it’s really not my problem…” the simulated roleplay would continue to play, and I would receive full credit for showing empathy. I wouldn’t receive any coaching on the rest of my statement that was clearly not up to the brand’s standards, inadvertently reinforcing bad behavior.

How would you know this? Simple, it’s all in how they build your simulation stories and report scores, keywords. Don’t get me wrong, their users will see an improvement in their training speed and likely get better outcomes than they were with traditional training methods, like going from the Model A to the Ford Pinto, a big difference. But we know that there are Ferraris on the road, and so should you.

With our unique use of conversational AI, Natural Language Understanding (NLU), and Natural Language Models (NLPs), we can simulate and analyze the ENTIRE roleplay between your agent and customer. Zenarate’s AI Simulation Training methodology uses these NLU & NLP models to learn, listen for, and coach agents on skills throughout fully branching stories, microlearning, and integrated screen and conversation simulations. This allows agents to observe best practices, practice with guided roleplay and then practice with unguided roleplay and receive coaching where needed until they are proficient and ready to take live calls.

Let’s take that same example from above. In that same simulated customer call that requires me as the agent to show empathy, I say “I’m sorry to hear about your difficult situation, but it’s really not my problem…” Our AI Coach would stop the simulation and offer me in the moment coaching and let me correct the behavior before moving on. This error would also be reflected in my score as an opportunity for improvement.

Two important things happen in that scenario. First, every word I say as an agent is understood and analyzed in real time by my AI Coach, not just specific keywords, which allows the system to gauge my intent as the agent. Second, is that the error is recognized and corrected in the moment, ensuring that I grasp the mistake as it’s happening in that moment, instead of trying to remember it after the fact. skills e1713294396380The proven result is true mastery of the skills, compliance requirements, and brand experience that you as a leader expect from your agent teams.

The Zenarate methodology behind our AI Coach platform has enabled hundreds of our enterprise customers around the globe to drive innovation, change, and improvements across the agent learning journey, resulting in faster speed to proficiency, improved call outcomes, and higher retention. To learn more, check out our Guide to AI Simulation Training or meet with one of our experts at your convenience.

dan headshot
Dan Mitzner

Dan Mitzner is a Visionary Marketing Architect and Zenarates CMO. With an illustrious career spanning over 17 years, David Mitzner is a growth-focused marketing executive renowned for his exceptional contributions to value creation. Recognized as an expert in B2B and SaaS, Mitzner brings a unique perspective to the contact center space, consistently driving organizations to unprecedented levels of success

Scroll to Top