What your complaints process never told you

In the Six Sigma world, we give a lot of lip service to the importance of the customer; we even have an official name for it—voice of the customer, or VOC. The problem is that many businesses don’t really have a good system for giving the customer what they ask for, much less one for listening to what they say. If you have been into a retail store recently, you may wonder if anyone has a good system for that. So what should a good customer quality assurance (CQA) system look like?

Let’s begin by defining customer quality assurance, because misunderstanding what it is can contribute to messing it up.

In the old days, we thought of errors as unavoidable, and so we defined high quality as “conformance to specification” and low quality as “outside specification.” (More on this in my article on the Taguchi Loss Function.) This was the era of quality control, where we tried to control quality by inspecting out what was nonconforming and hoping that what we were sending was conforming. Because inspection is less than perfect (some sources say human inspection is about 80 percent effective) then if there were a nonconformance, it had a certain probability of getting through inspection and to the customer. Although we tried to control the quality of the output, we weren’t always successful. But that was OK, because our customers didn’t expect everything to meet specifications anyway, just as we didn’t expect everything supplied to us to meet specifications.

Nowadays, we know that conformance to target is high quality, that barely conforming to spec is low quality, and that something that’s out of spec has no quality whatsoever. This definition pertains to our suppliers and to those we supply. In this era of quality assurance, the customer is assured of quality because we make it that way to begin with by manufacturing or supplying to target. The idea is that if we give our customers what they asked for (the target), then they will be satisfied and want to buy more from us.

That still doesn’t tell us what those pesky customers want from us, and that’s part of what a customer quality assurance process is intended to do. The CQA system provides information to the business to be used in developing new products and services, as well as a means of proactively and reactively improving customer satisfaction.

Six Sigma tends to focus people on solving problems, or gaps between what we’re doing and what we want to do. However, in many situations the problem is in knowing what the customer wants us to do. CQA ensures that we have a more holographic understanding of what our customers want, and involves preventing and solving problems. A business needs a process to assure its customers of receiving high-quality products and services, and it also needs to know what the targets are to gauge the quality as perceived by the customer. (If you think your quality is great, and your customer doesn’t, they win.) While CQA can use the people of Six Sigma, it’s a process that has the customer, current and potential, as its focus.

First off, the CQA system needs to be integrated into the strategic-planning process. This is how the customer’s voice gets into the project selection-and-prioritization system for the company. A CQA system that’s disconnected from the business-planning process is one that will always have plenty of reactive work to do (a bad thing), because the business will continually be going in directions that the customers don’t care about.

The other important characteristic of a good CQA system is that the metrics used to measure customer satisfaction are related to the company’s financial performance. For example, you might have the wrong metric if you correlate your CEO’s favorite customer-satisfaction metric, number of complaints, with the sales for the same time period and you get this:

Testing the Pearson’s r of 0.2706 for these data indicates that this isn’t a significant correlation (if the fact that it looks like a shotgun blast didn’t convince you in the first place). In desperation, you might even see if the number of complaints is a leading indicator and correlate it with future sales, but (since I made up these data) that turns out to be fruitless as well. And you get the fun job of explaining to the CEO that his or her favorite metric is, in fact, bogus, at least in terms of making the company money.

Sometimes finding a satisfaction metric that relates to key business metrics isn’t a trivial exercise, but in the absence of some sort of predictive validity, you could spend a lot of time increasing customer satisfaction and go out of business with the happiest customers around.

So in our example above, why were our customers not buying more of our stuff when they had fewer complaints? One company found that of their customers who had problems with their product, 55 percent were satisfied, but of those who had no problems, only 75 percent were satisfied with the product. Just because you don’t have a complaint about a product doesn’t mean that you’re satisfied with it. Satisfaction is achieved by meeting customer expectations, and only some of their expectations are covered by a warranty or are “complainable.” These expectations are a moving target, and your CQA system is what feeds that information back into the business. A fancy feature today is tomorrow’s expectation.

Further, it turns out that while customer dissatisfaction comes about from not meeting expectations, once you meet expectations and satisfy the customer you have reached the point of indifference. Give a customer two vendors that meet their expectations for the product or service and they would choose one on some other criteria. So be cautious about working only on customer “dissatisfiers.” Six Sigma actually tends to encourage this with its sigma measure, which I explored thoroughly in a previous column, “Are You Capable?”

If satisfying a customer is a point of indifference, how do you get customers to prefer you? You give them something that will delight them—something that they didn’t expect from your product or service. This could be something small, like a mint on a pillow in a hotel room, or it could be a feature that actually meets a need they didn’t know they had, like—in the early 1980s—car doors that close easily. I’ll bet you can guess what system provides the information that allows you to select the most cost-effective delighters: CQA.

Here is a high-level model of what the CQA system does:

As you can see, in addition to providing input to the design of the process, CQA also works to improve customer satisfaction over time. One part of this is how a business reacts to customer complaints and suggestions. Businesses should maximize how easy it is for customers to complain, since these complaints are very valuable information used for future innovation in the design cycle. (This is why lowering customer complaints should not be the goal—in the short term as you improve these feedback systems, complaints go up.) How a company handles complaints is an important indicator of whether the complainer will buy from you again. If a company resolves the complaint, 54 to 70 percent of customers will do business again with the company. If the complaint is resolved quickly, that figure goes up to 95 percent. A quick turnaround for complaints implies that you invested in building the infrastructure to do so. This type of thing doesn’t happen but itself; it has to be organized. Similarly, an infrastructure to capture and prioritize customer suggestions needs to be developed and nurtured as well.

If a business is only reacting to problems as they come up, they run a real risk of not finding out about a problem until it is overwhelmingly big. While we could deploy Black Belts on such a task, I guess I would rather not have to deal with a big problem. The only way to avoid that is to have an early-warning system. And the only way that I know of to have that warning is to ask the customers.

Customer surveys are a frequently abused and misused tool, in my experience. To learn something valuable, you have to do them right, which means creating a research question to be answered. Of course, we know that any customer-satisfaction metric is tied to critical business-performance metrics, which would clearly be part of the survey. But depending on your business, there are probably a number of other topics you want to know about your customers’ experience with you and with your competitors. Developing a good customer survey is critical, and if the survey doesn’t answer important questions, it probably isn’t worth the expense of getting a high survey-return rate. Most businesses have ineffective surveys, so they’re better off not getting them to the right people, who might take a dislike to such a clumsy company.

And yes, I know there are people who say that you shouldn’t survey your customer, because they don’t tell the truth and you don’t get anything valuable from survey results. These are people who don’t understand how to design a survey to answer a research question; they ask the wrong questions for the wrong reasons and, not surprisingly, get the wrong answers.

Similarly, it’s very important to perform strong market research. Certain analytical tools such as conjoint analysis can be valuable in understanding how your customers make their buying decisions. Other tools such as focus groups can horribly mislead a company. If, for example, you find that consumers prefer a particular soft drink formula over a competitor’s in a blind taste test, then you had better aim for that part of the market that’s either blind or makes their soft drink decision wearing a blindfold. However, for everyone else, the buying decision relies on visual cues, and a blind taste test predicts nothing about actual consumer behavior.


A customer quality-assurance system is a critical component to a successful business. It feeds information into the business plan about potential product and service developments as well as ongoing customer-satisfaction improvement. By reacting to customer complaints and suggestions, and by proactively asking customers about their experience, businesses can stay ahead of their competitors and prevent small problems from growing. Such a system will make the Black Belts’ jobs easier, because it will give them a more accurate transcript of the voice of the customer, leading to projects with bigger payoffs. A company that’s serious about listening to the customer, as opposed to plugging its ears and saying, “Nyahh, nyahh! I’m not listening!” is a company that can deploy its resources to increase those satisfaction metrics that align with its success.