Think about the metrics in your company: most likely, the metrics that you are accountable for and report on are reported as averages. Am I right? Perhaps, those metrics are averages and are lagging or leading, depending on the context.
But, from the customer’s perspective, it is always important to remember this axiom: The Customer Feels the difference in the experience, not the average. If you stop to really think about it, the concept of an “average experience” doesn’t exist. Yet, we report on statistical means.
The distinction, from a mathematical perspective, is simple: it’s about variation.
All processes are subject to some variability. More common explanations of variability describes variability as either Common-Cause or Special-Cause. The former is easiest explained as expected variation within a process that is produced by the process itself. The latter, on the other hand, is variation that is produced by the process but is assignable to some outside force. From a Queueing Theory perspective, another way to think of variation is by stratifying variation based on predictable and unpredictable.
Below is a Variability Tree, showing Predictable and Unpredictable Variability. This distinction is important because managing either type of variability will be different:
Most systems will show evidence of variability, as explained in the picture above; some will show both types, but typically one type will dominate.
For processes where arrivals are “bursty” or unpredictable, then the tools of Queueing Theory makes sense to apply; but for processes where the arrivals are predictable or seasonal, then it might make sense to apply the tools of Build-Up Diagrams.
I want to distinguish Service Variation for a moment. Now, take the ATM example above. Suppose that customer_a is serviced one way on day_1 and customer_a is serviced a different way on day_2. This type of service variation, felt by the same customer on two different days, can also be said to be “inconsistent”, from the customer’s perspective. The customer expects consistent service — the service might not be delightful — but at least it’s consistent. There is a difference and increasing consistency is the same thing as saying “reducing service variation”.
Now, from a marketing-spin, reducing variation does not mean we do not differentiate or segment our market. We can absolutely still do that and should. But, the way we service our customers ought to be consistent, but we can differentiate and segment.
My friend, Joe, recently posted on how to treat customers better. Here’s what he said, which reinforces my post today on variation, but his focus is on the more human side:
- Make it easy for customers. It seems these days that a customer never knows who to speak with regardless of the nature of their inquiry. It’s imperative that you make it simple for the customer to speak with the proper party depending on their concern. Often it seems that the customer has to speak with different departments in a company to find the answer they want. Eliminate the need for the customer to travel through the cross section of your company to receive proper attention.
- Limit who attends to a given customer. It’s unprofessional to pass a customer off to another representative if you’ve already been dealing with them. This makes the customer feel unwanted or that they’re not dealing with the best company. Whoever starts the process with a customer should be the one who finishes it.
- Chop down the phone tree. There’s nothing more frustrating for a customer then going to a company’s website and seeing a litany of phone numbers. Invariably the one the customer chooses isn’t going to bring them to the proper department. It can be difficult to figure out how you can narrow down all the different numbers your company may have, but it’s necessary that you make it easier for the customer to contact the right person.
- Own up to your mistakes. There is nothing more frustrating for a customer than when you’re just looking for someone to admit they were wrong. If you make a mistake then it’s crucial that you correct the error and admit that you were wrong. This will leave the customer with a good feeling about his or her experience and won’t turn them away from doing business with you in the future.
- Humanize your call center. There is often pressure on the people you have answering phones to handle a call in a certain amount of time and to handle a minimum number of calls in a given day. Let your representatives be real human beings and let them attend to a customer’s complaint or issue in a manner that leaves the customer feeling good. Ultimately, this is more important than statistics.
How To Manage Variability
When reducing or eliminating variation is not the answer, then managing it is the next best thing to do. Below is a simple table that shows the impact of variability on a system. I built this diagram quickly, but you can modify this table for your particular process; I’ve made it general enough so that you can apply it to your particular situation:
I’m using the term “Inventory” in a generic way here — it can mean people (patients) or actual things. For example, when the Waiting Room Size is bigger, then output can increase but there’s also an increase in the amount of patients in the room (patients = inventory). Use this trade-off table as it applies to your process.
It’s About the Customer
It’s important to always remember that it is Variation that people feel, not the average. Managing the variability in your process takes work and some knowledge of tools that are pragmatic and helpful. The “average” is an inadequate measure and is not descriptive of what the customer is feeling. If the customer is to benefit, we must take action against variation through reducing it or eliminating it and then managing it.