THE UPS AND DOWNS OF PERFORMANCE MANAGMENT – USING A BELL CURVE

Management of teams in a call center environment is never easy. Depending on your industry and your hours of operation not only do you often have to provide services and support during holiday periods, you often have to do it with a skeleton staff.  If those employees are on the bottom end of your performance bell curve – well let’s just say, you might not have a good holiday season!

With any business, not only do you need to ensure that you have the appropriate number of agents to deal with the incoming issues, but you also need to ensure that these agents have the right skills and abilities to handle the volume that they get. You are forced to juggle resources based on the type of issues you might get – based on your historical trends – but you also need to always keep in mind the SLA you are offering to your customers and partners.

Phone calls & online chat need a significantly higher response rate than emails and the volume of issues while it can be planned for, can never account for emergencies and issues outside of your control (this is why they are called emergencies!)

One tool that you can use to help you ensure that you have the right number and type of resources is the Bell Curve.  Utilizing this tool along with other KPIs that you have in place will help you in your planning and assist you in ensuring that you have the right people available to answer the customer’s query at the right time.

Bell curve graphs are called histograms. By reviewing individual performance (let’s take output as an example) and grouping it together across your team you will have a simple scale that graphed out shows a distribution. A normal distribution will have equal numbers on either side with the bulk of your team grouped in the middle.

Importantly, the bell curve (like any statistical distribution) describes a large number (or population) of individual things. The horizontal dimension (X-Axis) of the curve describes a range of values. The vertical dimension (Y-Axis) describes the incidences of “N” occurring. If we stick to our example of output, we would have a count of the number of emails responded to (for example) on the Y-Axis with a corresponding group of employees on the X-Axis. By determining where your employees fall in the distribution you will see where they stand in the distribution.

Now this can be done for quite a few of your KPIs (calls handled, average talk time, speed to answer etc…) and the more data of this nature that you add to your employee scorecard the better as graphical representations are extremely powerful and can tell you at a glance how an employee is “doing” in comparison to his peers.  

You will need a reasonable sample size to create an accurate assessment – do not judge your teams performance on one months data, but gather several months and ensure that your measurements are accurate by taking account of time away from the job (supervisors and other special projects that would impact their output and performance) and averaging your output over the course of the period in question.

In addition, you really need a sample size of at least 20+ agents to get a proper distribution – it’s even better when you get into multiples of that! As a rule, the more data you have the better your analysis will be.

You would take that data on a weekly/monthly basis and average it out over a specific period for all of your CSRs comparing their performance against your target goal. For example, if the output is to be 40 emails/day your distribution would have staff spread around that number in a fairly equal proportion (based on the size of your sample of course). Statistically speaking, the bell curve is defined entirely by its mean (average) and its standard deviation. But for our purposes, we need to know only that its shape can tell us a lot about agent group performance.

As you can see in the graphic above, the peak is where the majority of your employees are in relation to output. You have some staff on the high side and some on the low side. You will see that the majority of employees are at the average with others above and below that average. By reviewing what the above average performers are doing in relation to the below average performers you will be able to compare and contrast behaviors and determine what actions you can best make to improve your lower performers. (NOTE: if you have a “double hump” curve that’s something I will look at in future posts so don’t think that you’ve done something wrong – double and triple humps happen … a key case in point is called volumes by time of day where you will have several different peaks).

Remember if your skew is towards the lower end of the scale you have a problem that you need to address immediately! Depending on your customer base and type of business this might be fine for you – those on the low side could be your slower performers with regards to output, but conversely, your higher performers in relation to customer retention just because they spend more time with each customer. However assuming that this is NOT the case and that all you are interested in is output, your goal is to “skew” your employee’s output towards the higher end of the scale.  This can be accomplished through training – both technical, and product – as well as feedback from other more senior staff.

How does the Bell Curve impact staffing?

At its simplest, it doesn’t … however what you need to take into account when scheduling that coverage on a holiday or even earlies and lates, is the quality of the team doing the work.  The Bell curve analysis helps you put that into a very clear picture and by combining multiple different curves on a single graph it is sometimes very obvious who your lower performers are.