Business case: Staff Planning

Staff Planning: Calculating the demand forecast for call centers via regression models

Task

The management team of an inbound call center faces a challenge that many human resource departments have to deal with: on the one hand, staffing needs to be as efficient as possible due to economic considerations, but at the same time it is necessary to ensure – often via service level agreements (SLAs) – that employees are reachable at certain times. The key to solving this optimization problem is to develop an as accurate as possible long-term forecast of the time interval for expected incoming calls. Every improvement of forecast accuracy leads directly to an improved profit situation. INWT’s demand forecasting for call centers makes it possible to realize the full long-term optimization potential by providing today the information required for the reliable planning of personnel needs in the upcoming weeks and months.

Data

INWT’s demand forecasting for call centers requires data that is typically collected by a call center’s telephone system: the number of calls handled and the call lengths as well as the times customer calls were received.

Analytics

A regression/time series analysis is performed – based on the observation of past calls – to identify systematic changes in call behavior and to also isolate such changes from random fluctuations in the data. The causal relationships identified here (workload peaks, time-of-day and day-of-week effects, call center growth, higher incoming call volume due to marketing activities, etc.) provide the basis for predicting the systematically varying call volume in the forecast period. The second step involves merging this information with the relevant performance indicators, such as the level of staff utilization necessary to achieve profitability. This produces the actual target figure – the level of staff occupancy required to handle the call center’s workload – well in advance, thus allowing call centers to optimally plan for this scenario.

Solution

The call center’s demand forecast is integrated into the existing IT system and provides information about call volume, call length, and staffing requirements for each relevant planning interval (typically half-hour intervals). One benchmark shows that the solution has the potential – based on long-term requirements planning – to generate six-figure cost savings in the first year alone. The forecast should ideally be stored in a data warehouse, because this ensures that the entire company has easy access to the information. Experience shows that different departments use the forecast data to make better decisions in many business areas. The data is used by the call center’s management in staff scheduling and vacation planning, by controllers and managers in dashboard interfaces and internal reporting, and by human resources staff in recruiting activities. The solution thus facilitates not only highly efficient staff planning for call centers, but also creates transparency across the company and helps improve numerous other internal processes relating to human resources and personnel marketing.