About
A leading payments processor operates a contact center dedicated to supporting businesses using its payment platform. The support team handles a wide range of issues across multiple channels, including chat, email and voice calls, from simple queries to complex account access problems.
Challenge
The company hired 90 new employees who needed to be ramped up quickly to account for an influx of customer support needs. This led to issues with speed to proficiency and onboarding of new hires. The company needed extra support to improve agent performance and ensure high customer satisfaction, with a specific focus on increasing CSAT scores to maintain service quality during this transition.
The Fuel iX™ solution
Agent Trainer was introduced as a power skill builder, focusing on bridging the critical gap between training and learning. While the company's existing training provided foundational hard skills knowledge, they recognized that agents needed soft skills to effectively navigate the dynamic and often emotional landscape of customer interactions. These skills are difficult to teach in traditional training formats and are best developed through experience, observation and reflective practice.
The AI-driven agent training app was initially tested with a small group of underperforming agents, and later expanded to a significantly larger number of agents in a single business line.
The implementation began with chat scenarios — simulated text-based conversations mimicking real customer interactions. These chat scenarios provided an ideal starting point for agents to practice their skills in a controlled environment. Plans were then developed to expand to voice call training scenarios, in which agents can interact with AI simulations of customers with specific compliance requirements, while email scenarios were also incorporated into the app.
Agent Trainer proved to be an effective contact center training tool in simulating real user interactions, including challenging scenarios, which helped agents practice handling difficult situations in a safe environment. This was particularly useful for new hires to practice before engaging in live interactions. The app also provided prompts and reminders before interactions, such as active listening tips, which were especially beneficial for new agents. The ability to practice in a low-stakes environment where agents could restart scenarios without pressure was a significant advantage.
Even if the bot gets mad, it's good practice for them. That's part of the point — we can get them to practice these challenging calls or interactions and still get the job done in a good way.
Results
The implementation of Agent Trainer led to significant improvements across all support channels.
- 16%Improvement in CSAT across channels
- 80%Of underperformers transformed into top performers
- 29% CSAT improvement for chat channel
- Overall, agents achieved a 16% increase in CSAT scores across all channels.
- The chat channel showed the strongest improvement with a 29% CSAT increase.
- 80% of the initial test group transformed from underperformers to top performers.
- The app proved particularly valuable in preparing agents for voice calls, which traditionally caused the most anxiety for new hires.
Overall, agents found the app engaging and safe to practice with, which helped improve their knowledge, communication style and empathy.
The chatbot itself very much does mimic customer interactions ... when rolled out to the agents, the agents are like, 'But it's asking me to do the maths.' I'm like, 'Customers may ask you to do maths.'
Impact
Agent Trainer significantly improved agent performance and customer satisfaction. It helped agents feel more prepared and confident, especially for challenging interactions like calls, which often require real-time problem-solving, clear verbal communication and emotional intelligence to handle unpredictable customer reactions.
Typically, when new hires begin handling customer interactions, there's a 40% drop in customer satisfaction scores. However, Agent Trainer demonstrated the potential to substantially mitigate this decline by better preparing new agents for real-world customer interactions.
The overall feeling of that initial group was positive. There's this idea that you can't really mess up; it's not a real-life scenario. The feedback has been really positive overall, and there has been good engagement from the agents.
This optimism is shared by company leadership, who are excited about the app's potential and plan to roll it out more extensively for new-hire training to improve speed to proficiency.
From a data perspective, it has been really, really good. I don't see how or why this couldn't or wouldn't have great opportunities in the future. The tool itself is very, very promising.
Agent Trainer solves one of the biggest challenges in contact centers — the slow and costly process of getting new agents ready for customer interactions. By bridging the gap between traditional training and real-world application, it helps agents develop essential soft skills through experiential learning. This results in increased CSAT scores, minimized employee stress and boosted revenue.
