In most contact centers, the executive team recognizes agent attrition as a necessary evil. They know there is some expense associated with it, but the general opinion is that the cost of the cure is likely greater than the benefit. It is your challenge to change that line of thinking. Building a case stands on three credible arguments.
Argument No. 1: Cost
I know, you want quality to be the top argument, but accept the reality: Cost is more tangible and, therefore, garners more executive attention. There are some pretty straightforward elements to the turnover cost:
- Recruitment includes advertising expenses that are easy to find and to which you would add the cost of the recruiter’s time. Interviewing and selecting requires support from HR, supervisors and managers—all valuable time that could have been spent working with staff and improving results (est. $3,000 to $5,000 per hired agent).
- Training is another large, direct component of the cost of turnover. Many centers now report classroom training periods of four to six weeks or even longer. During this time, the trainee takes few, if any, calls (est. $2,400 to $3,000-plus per new-hire).
- The costs do not stop when the trainee leaves the classroom. The learning curve includes another four weeks in lost productivity (est. $2,400).
Argument No. 2: Quality
Once you lay out the cost issue, it is easier to make the case on quality by providing an objective argument that maps back to the bottom line.
Showing the lower cost of quality starts with agreement on the price of an error. Develop a clear definition of what constitutes an error (as opposed to a minor “improvement opportunity”) and then apply the same cost to each. The elements that make up the average error cost will vary by industry, but will generally include the cost of correcting the mistake and the cost of repairing good will (or the cost of not being able to repair it).
Once the cost of an error is agreed to, the cost of lower quality is easier to calculate. For example, if it takes 11 weeks on average to get to an expected error rate of 2%, all errors above that in the first 10 weeks are attributable to the learning curve. Assuming just $50 per error (a very conservative figure), the additional costs will be over $3,000 per trainee.
Argument No. 3: Consistency
Your WFM team probably has past data on annualized turnover rates that it is using to predict future rates. They are likely just taking an average of the past year or two to make this estimate, and dividing the annual rate by 12 to get a monthly number. The problem is that these numbers are an average, and averages do not often hold up very well in contact centers. Have your WFM team look at actual data month by month and compare it to the average, and you will likely see some big swings. In fact, the difference between projected and actual turnover will be far greater than the difference between projected and actual call volume.
What is the impact? If you cannot predict turnover, you cannot plan for it. To make the service level objective, the WFM team will either have to plan on higher than average turnover rates (causing overstaffing at some points throughout the year), or you will miss service level objectives when monthly turnover is higher than average. There is a cost to either outcome.
Putting It All Together
Many centers that have done this type of analysis have put the cost of turnover at $15,000 to $20,000-plus per agent. Even at the lower number, decreasing the attrition rate from 45% to 20% in a 200-seat center will result in 50 fewer trainees per year, saving the organization $750,000 annually. That represents only what we can count—the intangibles make the real effect far greater.
Now that you have a solid number, you have a case. You also have a break-even point. Spending $500,000 to get the reduction of 25 percentage points seems less like an expense, and more like an investment.
Building the argument and generating the buy-in is a starting point. You still need to identify what is causing the turnover and how it can be corrected. The difference is that when you start to work these issues, you will already have the support you need to make the changes.
