User-Centered Requirements and Design for FedEx Call Center
Goal
FedEx wanted to use the performance analysis, requirements and design activities to reduce "time to competence" and training/ramp-up time for call center agents from 18 weeks to 6 weeks while improving on-the-job performance.
Challenges
- Target user population included three types of call center agents: Domestic, International, and Hazardous Materials, and a total of 1,600 individuals at FedEx call centers across the nation
- Call center agents had minimal downtime to be available for off-call interviews so most interaction had to occur in the call center through structured observations with very brief interviews
- Process required working with top performing agents, which negatively impacted their call time metrics so the team had to find ways to compensate them in order to motivate participation
- Information and job aids required to complete calls was printed monthly, bound in a 3-inch document, and shipped to each call center agent, making access tricky and increasing costs to FedEx.
Solution
Cognitive's approach was to:
- Conduct an extensive task and resource analysis across the 1,600 call center agents, using a combination of observations, interviews, and easy-to-complete Likert scale surveys
- Analyze the information to compile a new job profile for a "universal" rep who could take any type of call
- Document user requirements for the system and translate those into preliminary designs for new system screens/user interface
- Create a new set of online coaches, wizards and other job aids (e.g., performance support tools) to assist call center agents in timely access to the information they needed and accurate completion of calls.
Cognitive Value Added
Cognitive Technologies' personnel provided value to FedEx through:
- Leading the analysis, requirements, and preliminary design activities for the team, including the compilation and analysis of huge mounds of data and delivery of the findings to the FedEx executive in charge
- Developing high-quality tools to gather the right data, including the development of a task-based Likert scale survey that enabled us to quickly gather data about difficult and complicated tasks, tasks requiring better support, etc. from the population. We achieved a 75% return rate (standard for the industry is 25-35%).
- Performance experts who created preliminary user interface designs and new performance support tools to improve agent performance while reducing face-to-face training time.