In the 90’s, I worked a number of contact center process improvement projects using Six Sigma. As with many Six Sigma projects, the idea was to come up with a predictive equation for a specific indicator, in my case, I was trying to predict customer satisfaction. It wasn’t a surprise that one of the best predictors of customer satisfaction (CSAT) was first contact resolution (FCR), a fact that is now very well known. Another significant factor to predict CSAT was agent tenure, or length of service.
Tenure is a funny metric. It drives CSAT and FCR, because of the underlying factors that come with tenure, namely experience, speed to proficiency, mastery of tools and systems – as well as how good those systems are, the knowledge of unwritten processes, and the ability to assimilate new team members to the floor through their own experience. Since that time, a myriad of studies spawned around reducing learning curve, accelerating speed to proficiency, making tenure seem less impactful to CSAT and FCR. Almost all Contact Center metrics, most obviously Handle Time (AHT), and because of those impacts workforce management needs to keep tenure in its radar.
While good capacity planning can forecast tenure, it can also drive it, if it can estimate expected tenure and direct it, it can then estimate CSAT and FCR based on the tenure forecasted. I have seen most plans incorporate AHT as part of this, but very few are incorporating customer satisfaction metrics. It is my hope that at least intuitively, you can convince decision makers in your organization that WFM planning can not only drive the traditional efficiency metrics (occupancy, utilization) and service metrics (service Level, ASA, abandon rate) but it can drive CSAT, FCR, and customer experience.
There are a few ways to do this, but one is to directly calculate the expected average tenure per week or per month in your staff plan. This isn’t difficult. You want to have the weighted average tenure of your existing staff; plus the tenure of the new hires you have in your plan. The new hires “age” week over week and therefore modify the tenure of the whole agent population. You can stay at weekly, daily, day of week or interval level, go into as much detail as it is significant for your operation. At this point, you will be able to have an expected average agent tenure for every week into the future. Now you need CSAT or FCR historical reference points for a few tenure levels, this is typically something that your analytics or quality assurance can provide. With a little bit of play, you will have a lookup table for different expected levels of tenure and then you will be able to have CSAT and FCR for all future weeks. As with any forecast or model, it will not be perfect and will need some adjustment to ensure you are able to accurately predict the actual CSAT and FCR well into the future. With patience you will have a great model.
Turn WFM knowledge into action
Now that you got a good enough model, here is where the fun begins. The number of classes per month or week, the size of the classes, the expected attrition both production and training, all the other factors on your staff plan will reflect on your future CSAT and FCR. The more reference points you build in, the more sensitive your model will be to small changes in tenure. You will be able to see how CSAT it is hurt by sudden huge waves of new agents that come in all at the same time. Demonstrate how the cost of attrition hurts the customer satisfaction and give you a reason to invest in avoiding turnover. You will be able to show how it improves with smaller, but maybe more frequent classes. At the same time, you will have numbers to show that by listening to workforce management recommendations, the customer satisfaction will improve. Your workforce management skills will be driving those metrics that are most important in today’s area where customer experience is key.
Now, why do we call this advanced workforce management? Not long ago, there was a time where WFM was all about the balance of two things; efficiency metrics represented by utilization, and service metrics represented by service level, ASA or abandoned calls. Today in the world of CX centricity, it is time to end that bi-lateral world of workforce and go into a multidimensional environment that includes customer experience metrics and more.
