The Next Level in Performance Management

By Cliff Hurst

Ours is still a young industry. Management practices within call centers continue to evolve as our industry matures. We can be forgiven, I suppose, for having some practices that are still a little rough around the edges. Quality Monitoring (QM) is one of those rough edges. Currently, most call centers manage their performance by concentrating on three primary measures: efficiency, responsiveness, and quality. We do a better job with some than with others.

On the whole, we are excellent at measuring and managing efficiency. In fact, efficiency is the cornerstone on which call centers were founded. The telephone is a highly efficient way of relating to customers. It is therefore, also quite cost effective. The industry is also becoming better at measuring and managing responsiveness, as measured by service level.

Quality, however, remains the stepchild of performance management. We don’t measure or manage quality very well. However, let’s not kid ourselves as to the real reasons why we monitor calls. We say that we do it to help improve overall call quality, but our actions belie that intent. In practice, call centers use Quality Monitoring (QM) primarily, if not solely, as a tool for measuring and evaluating agent performance. This is shortsighted and prevents us from achieving improvements in the performance of the organization as a whole.

The Consequences: The way we habitually approach quality isn’t working. It ends up in finger-pointing, blaming agents, lowered agent morale, a lack of commitment to the QA process from both agents and management, frustration all around, and no discernible improvement in quality over time. We can do better; we owe it to our clients and to our agents.

A New Model: I have been working to develop a new model called Getting Quality Right. This new model really excites me; I hope it will excite you. Implementation of this model will result in significant advances in QM and coaching feedback. It may well do for QM what service level has done for managing responsiveness. It will allow managers to make rational decisions based on statistically sound principles applied to a uniquely complex environment.

Why Monitor Calls? My own “Aha!” moment came when I realized that there is no one single answer to this question. Any attempt to answer it begs four additional questions, and an effective Quality Monitoring program must answer all four.

1) How are we, as an organization, doing at representing our call center?

2) What can we, as an organization, do to get better at representing our call center?

3) How is this agent doing at representing our call center?

4) What can we, as managers, do to help this agent get better at representing our call center?

Answering these four questions ought to be the fundamental goals of QM in any call center. Answer them well and your quality team will perform like an orchestra, instead of like a one-man band.

What’s Not Working Today: The problem encountered by most call centers is that we try to answer all four questions by using one methodology. That doesn’t work. It results in a nearly exclusive focus on question 3: How is this agent doing? But evaluating the level of performance of individual agents doesn’t take us very far. This is not where our biggest improvements can come from. Bigger improvements come from improving our processes, systems, training, policies, and leadership practices. The heavy lifting gets done in response to question 2, but to answer 2, we first need to answer question 1!

We crunch a lot of numbers in call centers. Mostly, we count things – lots of things. We tend to shy away, though, from analyzing statistically the things that we count. If we’ll learn to analyze numbers statistically, however, we will able to gain much more meaning from them than if we merely count.

It is especially useful to be able to use inferential statistics. With inferential statistics we are able to infer meaning from our analysis of a smaller subset of data (such as the calls that were observed) with confidence that our findings represent the whole (such as all calls handled by your center in a period of time).

How You’ll Know: You will know that you are doing this right when you can say to senior management or your outsourcing clients something like this:

“I am xx percent confident that our overall quality score this past month is zz on a scale of 100, and this score is accurate within plus or minus a yy percent margin of error. Here’s how our trend looks over time.”

The xx percent and yy percent are standards of accuracy and precision that you set. Once you set them, you can calculate how many calls in a given period of time you will need to monitor in order to obtain a representative sample. Then you measure your actual performance against your own standard.

There is no “right” percentage to use here. You may elect to establish a 90 percent confidence interval with a 5 percent margin of error. Or you may opt for a 99 percent confidence interval and a 2 percent margin of error. The choice is yours. Just be forewarned, a higher standard is going to require monitoring a larger percentage of calls. It costs money in terms of staffing and resources devoted to Quality Monitoring. This is like managing service level. You first set a standard, and then you establish the practices needed to achieve that standard.

The Key to Answering Question 1: The key to being able to express your quality results in the above manner is to first be sure you are monitoring a representative sample of calls. If you have a high level of confidence that the sample you monitor does, indeed, represent the calls taken by your organization, then you have found a way to answer question 1.

Your answer to question 1 is always an inferential statistic. This means that you can never know for certain what level your quality is. But you will have a way to express the degree of confidence you can place in your measurements. You will know how accurate and how precise your answer is. The guesswork can finally be taken out of QM. This, then, is the first practice to master in the model for Getting Quality Right.

QM is foremost about measuring, managing, and improving overall performance. This opens the entire customer interaction process to improvement opportunities. Using QM to train and coach agents becomes a secondary objective; using QM to evaluate agent performance is a distant third.

Read part 2 in this series.

Cliff Hurst is president of Career Impact, Inc, which he started in 1988. Contact Cliff at 207-499-0141, 800-813-8105, or

[From Connection Magazine April 2008]