Many employees, from the executive table to the front lines, often shudder at the though of instituting a quality initiative because of the perceived complexity of the exercise. And, thanks to the cornucopia of buzzwords and acronyms bandied about during the past decade or so, I really don’t blame them.
However, you can experience meaningful results with minimal effort, if you follow a few very simple steps. Or, more specifically, implement a few simple tools. This article will focus on one of the most powerful, yet simplest of tools – the Pareto Chart.
We will start with a real-world example: Customer comments. Most organizations have some method of gathering comments from both internal and external customers. (Internal customer refers to someone to whom you provide services within your company. For example: An IT help desk agent serves internal customers. External customers refer to the typical consumer, client, or vendor utilizing your products or services.)
Hopefully, you have a systematic process to capture, evaluate, learn, modify, and communicate the data gleaned from the comments. (That is a discussion for another time.) As we explore the Pareto Chart, we will discuss the notion of filtering out anecdotal noise from the empirical data from your surveys. As with many customer survey’s, the quantitative data is often the most clouded. Unlike the basic questions (How likely are you to refer our company?, Pease rank your experience on a scale of 1 to 1, etc.), the Pareto Chart is most effective when summarizing the information gathered in free-from comments.
Now, back to our example. Today we will presume we are working for a legal services firm, providing low-cost assistance for small businesses. The free-form comments on your surveys will most likely fall into two main categories – complaints and praise. These should be split for obvious reasons. Let’s focus on the complaints, most of which should be managed at the process level, since 80% of your company’s failed customer experiences are process driven. Certainly these experiences are delivered by people, but driven by processes.
Noise: We have all experienced the issue of noise. Those one or two comments, often delivered verbally to a chief executive. Are these comments noise, or an indicator of a weak or broken process? Do we act, just because the CFO was browbeaten in the elevator? Include these comments (along with other data from your customer surveys) into your analysis; as they are part of the Voice of the Customer. Then, let the Pareto Chart help you to determine how important or significant the issue really is – today.
A Pareto Chart is traditionally a column or bar chart illustrating the percent of individual data elements, and then sorted (highest to lowest). Now, draw an imaginary vertical line at the 80% level. Normally, anything outside the 80% (to the right) is considered noise. Unless you see it growing over time, it (this noise) could be considered insignificant when allocating resources to fix processes. (We’ll show you simple ways to improve processes next time.)
Here’s an example of a Pareto Chart, based on complaints:
Note how Long wait, Late files, Incomplete files, and Parking fall into the rightmost 20% of the information. (This is your operational noise!) However, the Online system, Inconvenient hours, and Phone issues account for the initial 80% of the information. In this scenario, these top issues should get the attention. And, if the CFO’s information was all about someone having a difficult time parking that day, well then it’s just noise for now.
These charts are nice for demonstration purposes. However, a simple sorted table is all it takes – and this is where I promised to simplify the Pareto chart. Note the cutoff at 79% (close enough to 80%) in the cumulative column. That’s all it takes to identify the real issues and filter out the noise. So, if you are the only one looking at the information, you may not need to create the chart.
For those keeping score, this fictitious company needs to work on making themselves more available to its client base.
Finally, Pareto Charts are not just for survey data. They can be deployed to determine the significance and/or noise in many scenarios.