Call centers should base their staff planning on the number of unique callers instead of the total number of calls. Furthermore, introducing a system that gives redialers priority over fresh callers both reduces the average waiting time and improves the fairness of the system. This can be concluded from a study on call center management conducted by CWI. PhD student Sihan Ding defended his thesis ‘Workforce Management in Call Centers: Forecasting, Staffing and Empirical Studies’ on this topic on Wednesday 23 May at VU University Amsterdam.
Around two-third of the total costs of a callcenter are personnel costs. Call center managers and planners usually aim to satisfy their service level targets with the least number of agents. Prediciting the future call volume is an essential step in this procedure. It also one of the hardest problems with no simple solutions. It is impossible to know for sure how many inbound calls will come.
In his thesis, Ding shows that redials and reconnects have a significant influence on call volumes. This makes the total number of calls an ureliable predictor for staff planning decisions: since redials and reconnects are often caused by long waiting times, these are statistically related to the staff planning. Based on this conclusion, he develops a staffing model that includes the flow of redialing and reconnected callers.
Differentiating between fresh and returning callers allows for more efficient and fair call center management. For this, Ding proposes a call-back system, in which long-waiting callers are suggested to call back some time later if the system is congested, and when they call back receive priority over fresh callers. He proves that this reduces the average waiting time, and is fair to those long-waiting callers.