By Peter Lyle DeHaan, Ph.D.
Benchmarking is the comparison of your call center with statistical results from the norm of industry peers. These numeric measurements are called metrics. Metrics can be in the form of financial figures, operational quality and efficiency, human resource efficacy, or whatever is deemed the most valuable to the participants, though typically and primarily they are operational in nature. “If it can be measured, it can be improved,” asserted Kelly Doran of Simcoe Message Centre in Barrie, Ontario. The “objective measurement of quality standards can help highlight areas of strength and weakness in both individuals and teams.”
Successful benchmarking follows a progressive path towards a desired outcome. First, there must be a desire to obtain, have, and use the information. Next, you need to determine who will be invited to participate. The basic requirement is for participants to have an interest in the results and a commitment to contribute. Beyond that, it is imperative that all participants are in sufficiently similar business niches within a common industry. In many cases, it is wise to select those using a common hardware or software platform, since operational metrics are hard to reliably compare when their source is different, employing dissimilar statistical standards. Some will assert, that from the caller’s perspective, a call center is a call center and therefore it doesn’t matter who your center is benchmarked with as long as they are of similar size. Yet everyone knows that a telephone triage call center is much different than an order-taking operation and a physician’s referral line has different objectives than a literature request function. Quite simply, it makes no sense to compare your call center to another one that is in a different industry, pursuing different goals, and with different cost-benefit standards.
The third step is to determine which numbers to measure or gather. It is recommended to start small, obtaining only a few key numbers. As participants become engaged in the process and realize the value of it, then other metrics can be added. This is followed by developing a standard determination of how the information will be gathered or the calculations will be made. For without a standard methodology each participant will make the calculation as they see fit, rendering any results unreliable. These two steps can be both time consuming and contentious. Assistance from someone with experience in benchmarking or a background in statistical analysis is most beneficial at this point, serving to greatly simplify the process and save valuable time. Also, if this person does not have a direct stake in the results, they are able to more objectively guide the process.
The fifth step is a critical one. It is to develop the survey form, which includes documenting the source or calculation of the data. Although this seems like a simple and straightforward process, it is one fraught with peril, as a less than ideal survey form will doom the process to misanalysis or failure. Again, someone with experience in benchmarking or developing survey forms will be most helpful. Then, regardless of the quality of the survey form, or its developer, it is of paramount importance to test it. What may seem perfectly clear to those who developed and reviewed the form, it could cause confusion or misinterpretation among those completing it. Therefore, a small field test should be conducted. Any problems uncovered in the test will need to be corrected before the benchmark survey is distributed to all participants.
The next two steps are the most important, as concerns in these areas can cause otherwise willing participants to decide not to complete the survey or to color their responses. Quite simply these steps are to gather the completed surveys and then to compile the results. Concerns reside in who performs these two items. It is imperative that this person or group be trusted, respected by all participants, and that there not be any perception of a conflict of interest. As such, it is recommended that someone not participating in, nor who will benefit from, the benchmarking results be assigned the task of both collecting and tabulating the responses.
The results of the benchmarking survey should only be presented in aggregate form and then only to those who responded. All individual answers must be fully protected. In some cases, such as providing cross-sectional or demographic analysis, certain sections may need to be eliminated due to a small number of responses that would effectively expose one or two members. The results, often along with analysis and a commentary are distributed to all who participated.
Although conducting a benchmarking study once is valuable, the real benefit comes from repeated studies over the course of time. Therefore, it is important to follow-up with those who participated to determine any problem areas needing correction or additional data to be collected. These changes must be made and the survey repeated. Depending on the nature of the information, the survey should be repeated at least annually, possibly semiannually, quarterly, or even monthly. The benchmarking results then become a periodic report card showing your successes, your shortcomings, your improvements, and your relapses – all with respect to your peers. This provides the basis for celebration and self-improvement.
Some Examples of Benchmarking Metrics
- Percent of calls answered
- Average time to answer
- Percent of calls placed on hold
- Average hold time
- Occupancy (percent of time spent working)
- Average call duration
- Average wrap up time
- Number of calls answered per month
- Amount of time spent on calls per month
- Schedule adherence
- Annual turnover rate
- Average employee (CSR) tenure
- Cost to hire one new employee
- Cost to train one new employee
- Starting pay per hour
- Average hourly rate
- Percent of revenue spent on labor
- Cost per call
- Value of the call
- Profit margin
Summary of Steps for Benchmarking
- Possess a desire to obtain, have, and use the information.
- Determine who will be invited to participate.
- Determine which numbers to measure or gather.
- Develop a standard for how calculations will be made.
- Design the survey form.
- Test the form and correct problem areas.
- Distribute the form.
- Gather the completed surveys.
- Compile the results of the collected surveys.
- Present the findings.
- Analyze and correct any problems for next time.
- Determine additional data to collect the next time.
- Repeat the process periodically (at least annually).
Peter Lyle DeHaan, PhD, is the publisher and editor-in-chief of AnswerStat. He’s a passionate wordsmith whose goal is to change the world one word at a time.
[From the Fall 2004 issue of AnswerStat magazine]