“This work had tremendous impact on the bottom line of our business, and the quality of service we gave our customers. Within months of hiring people with characteristics that Talent Analytics correlated to high business performance, employee turnover dropped by over 30%, yielding a multi-million dollar annual savings.” – EVP, Financial Services Firm
A global financial services firm was experiencing >60% voluntary attrition among Customer Service Representatives (CSRs). They hired close to 30,000 CSRs annually, making this an unrecoverable multi-million dollar expense. To be placed in a full-time role these CSRs had to pass a Series 7 exam. In preparation for this exam, new hires were paid for 12 weeks of training – culminating in a Series 7 exam. If they did not pass the first time they were fired.
They could predict the following repeatable challenges:
- New hires began dropping out of classes almost immediately
- The closer to the graduation they got the more new hires dropped out
- A significant number of new hires failed their Series 7 exam after having been trained (and paid) for 12 weeks
- To put this in perspective, of every class of 80 new employees, only 13 made it through the end of the training and passed their exam.
Analysis of traditional talent metrics like prior experience, compensation, interview questions etc., failed to uncover any systematic way to understand why people were leaving and how to find a new kind of candidate that would stay in training and pass their exams.
Talent Analytics’ Approach:
Talent Analytics was engaged to explore a different dataset – the person themselves. Our hypothesis was that there was a link/correlation between the CSR’s raw talent, their attrition and other key performance indicators. Our approach explored whether or not there was a certain “raw talent profile” that did extraordinarily well / poorly in the training, the exam and their job.
What follows is a brief description of the project and staggering outcome.
Criteria was established and standardized for identifying a top performing customer service rep. (Metrics included length of employment, skills, competencies, call quality, single call close scores, call times, etc). 775 CSRs were selected (both top and bottom performers) to be part of this study based on the established criteria.
Talent Analytics scores were gathered from study participants using our Rapid Data Collection™ process. Raw talent scores were derived (within seconds) from their answers using Talent Analytics’ Advisor, and exported to a CSV for statistical analysis. Talent Analytics scores were combined with their Key Performance Indicators. Statistical analysis was performed looking for correlations between raw talent characteristics and top / bottom CSR performance.
Strong statistical correlations were found between “raw talent characteristics” and top performance as measured by the customers Key Performance Indicators. Correlations resulted in a recommended benchmark (11 numbers) representing an ideal set of raw talent traits. Human Resources added this benchmark as an important, additional data point in their hiring process, and began advertising, interviewing and hiring guided by the Talent Analytics statistical models.
Comments: When performance relies heavily on the performer – failing to quantify and include performer in your predictive models is a major failure.
This Case Study revolves around attrition – one measure of performance. Your business performance challenge might be sales performance, customer service performance, data scientist performance, engagement or another kind of performance (errors made?) The common element is that all of these performance areas rely heavily on the performer.
Talent Analytics’ methodology and platform make it possible (easy even) to quantify and include raw talent metrics in predictive models for a much more complete view of future performance.