by Robert Last* for HDI®
1. Calls bunch up from time-to-time. Telephone calls arrive randomly and have peaks and valleys. Planning for a workload that arrives randomly makes support centers different from other departments in an organization. Customers (callers) decide when they want to call, and despite some general patterns, support analysts have to respond when the customers need them. Staffing and productivity tasks have to be considered in this context.
2. There is a direct link between resources and results. Certain levels of resources are required to reach a specified level of work. If it takes 36 first level analysts to achieve a service level of 90 percent with an ASA of 20 seconds for a given call load, then that is what is required to staff the support center. Incremental improvements can be achieved with training, good organization, and excellent leadership, but they cannot achieve long-term, long-lasting productivity improvements. In this case, doing more with less is counter-productive.
3. "Staffing on the cheap" is expensive. If analysts are so busy that they cannot routinely leave their phones, they will burnout and leave for other jobs. In addition, average handling time will increase as analysts find ways to take fewer calls out of exhaustion. Customers will complain about long wait times and customer satisfaction will suffer.
4. Service level has no industry standard. There is no one service level that is applicable to every support organization. Each organization places different values on customer service, and each will have different staffing costs, network costs, and numbers and types of callers. Determine a service level that makes sense for your organization according to your caller's needs, your objectives, and your cost structure.
5. When service level improves, productivity declines. Although it may appear counter-intuitive, there is logic to this phenomenon. The higher the service level, the longer the analysts are in their seats taking calls. At a certain point, fatigue sets in and productivity declines; tired people make mistakes and look for ways to lighten their workloads.
6. You will need to schedule more staff than the base staff required. Unless every single analyst has absolutely no other job responsibilities, never goes to lunch, training, the restroom, or goes on vacation or becomes ill, then the support center will need more people than is indicated in the results of staffing calculations. It is important to recognize that schedules should realistically reflect the tasks that can keep analysts from taking calls.
7. Purchase the best hardware and software possible. Hardware and software make up less than 15 percent of a support center's budget over the long term. Telecommunications hardware and software, problem tracking software, knowledge base software, headsets and amplifiers, and training equipment are critical tools for a support center and it makes sense to buy tools and systems that provide the "biggest bang for the buck."
8. Telecomm and IT people should provide service and support to the support center. Support centers do use a great deal of technology, but they are customer-facing operations, not technology operations. These systems should be managed from within to maximize their use in supporting customers with the support of telecomm and IT departments.
9. Summary ACD reports don’t tell the entire story. Interpret ACD reports for what they are worth; they provide one snapshot of the activity in a support center and need to be considered in the context of the overall operation of a support operation. Only a Balanced Scorecard can provide the comprehensive information needed to convey the details of a support operation.
* Adapted from Call Center Management On Fast Forward, Brad Cleveland and Julia Mayben,
Annapolis, Maryland: ICMI Press, 1997, Pages 146-148.
To read more about incoming calls and the support center, see HDI's focus book, The Executive's Guide to Understanding Technical Support by Robert Last. This book is available on the HDI eStore at www.thinkhdiestore.com.