Software Development

Quality assurance: Does QA really assure quality?

Having a quality assurance process doesn't guarantee a quality product. This gantthead article outlines what you should be looking for in setting up the QA process and metrics.

By Kathryn M. Denton

"Customers pay only for what is of use to them and gives them value. Nothing else constitutes quality." - Peter Drucker

Quality assurance (QA) measures are valuable, but they may not tell the whole story. For example, consider call center responses. Suppose a customer’s QA survey response shows satisfaction with your call center’s answer to the customer’s problem, but the problem keeps returning. Is the customer truly satisfied? Or, maybe your monthly results are positive, yet a customer claims to be very unsatisfied due to problems from years past. In both cases, the QA survey failed to indicate the problem.

Let's explore some ways to tweak your QA measures to derive results that align with the customer's reality. Until you do that, adjustments to improve quality might miss the mark.

Who asks the questions?
First consider who should perform the QA. Certainly it shouldn’t be the service providers being measured. Is it possible that someone could fudge the numbers? Your people may be trustworthy and understand the value of seeking improvement, but they’re only human, and they may be tempted to help the numbers glow.

At the very least, those who are taking the calls might choose to record only successful service calls. This would give you very good numbers, but your customers would tell a different story. Requiring staff to provide each customer with a call number can help prevent selective call-logging. Also, try to use your own experience as a rule of thumb. Does your service experience reflect the numbers on the monthly graphs? If you turn over the QA measurement function to another individual, do the numbers stay in the same range? A periodic changing of the guard might help keep things honest and alleviate boredom and stress for those performing the role.

What are you measuring?
If you already have a program in place, are you measuring the things your customers care about? Ask your customers what they’d like you to measure. You might find that they only ask for a few things, not the exhaustive list of services you actually provide. Why? Because most of them don't know or care what the bandwidth is or how fast you can replace a router, as long as they get a fast response without downtime for their application. So it makes more sense to ask them if their application’s performance meets their expectations than to ask them jargon-laden questions about technical services.

Another factor to consider when making the measurement decision is alignment with your strategies. If you have a strategy to improve customer service, it makes sense to ask your customers if their opinion of your service has improved (their answer may not always match the graph delta).

It’s also a good idea to separate questions about products from questions about service. People tend to be biased toward specific products, and they often magnify the negatives of other products. Every product has its strong points and its faults. Certainly you want to know if a product is of poor quality and should be replaced, but don't blame your front line for bugs in purchased products.

Get prompt replies to survey questions
Now that you’re measuring what the customer cares about and your measures support your strategies, let's discuss timing. When you ask customers about their experience with a service they received, you may get different answers if you ask them immediately after the service and then ask the same question a month later. They may not even remember having received the service. An advantage to asking for customers’ opinions promptly is that you can review the results and make changes faster, improving service for others before they get hit with the same problem. If you aren't doing this already, have your help desk staff leave a short questionnaire with the customer immediately following the service—for example, a short e-mail survey, a link to an online survey, or even an old-fashioned 3x5 card.

More on gantthead
Related downloads: Related content: Note: Items in bold are available only to gantthead premium members.

Currency isn’t the only factor in determining customer satisfaction, though. People seem to have a long memory for bad service and a short memory for good service. So your prompt after-service surveys may have very high marks, but when you ask a customer what they think of service in general they may give you a negative answer. An annual survey asking more general questions can bring that disconnect to the surface. Can you fix the problem? Perhaps, but it may take time and effort to wipe out the negative history.

Help your people help your customers
Don’t be too hard on your service providers. Create a supportive environment and give them the tools they need to do a great job. Include customer service training as well as training in the various tools they support. Provide them with a good knowledge management tool. Remind them to review the logs periodically so they can learn from information others have recorded. Show them the Web sites for the tools they support, and teach them how to find and use the support pages. Give them an overview of your service contracts with vendors and the procedures for working with them. Be sure each of them has a copy of the service level agreement so they will know what turnaround the customer expects for various services.

Quality assurance should improve customer service, not become an exercise in fudging the numbers. When it breaks down, you’re not only wasting resources, you're creating a culture of dishonesty. Instead, create a supportive culture of problem resolution and great customer service. When you've properly prepared the landscape, you will be better suited to fix the problems that do occur instead of pretending they don't exist.

This article was originally published on gantthead on May 6, 2002.

Kathryn M. Denton is a technology manager with more than 20 years of experience in the field. She is also the author of Corporate Russian Roulette, a business behavior management book.

0 comments

Editor's Picks