Use Call History Data for Better Forecasting and Scheduling

Forecasts determine schedules, but what determines forecasts? There is both art and science involved in predicting future call volume and agent staffing needs, and technology can make the forecasting process more accurate. But the starting point should always be a review of call history data.

call forecasting and scheduling

Past activity is always the best predictor of future activity, especially when broken down into ever-smaller increments of time. This makes it easier to identify anomalies and prepare accordingly. You’ll want to have monthly and weekly stats to review, and then dig deeper into daily and hourly numbers. Finally, examine work periods as short as 15 minutes. You may be surprised at the stats for these intervals, and it may help in determining when agents can take breaks, and whether personnel are beginning or ending their shifts on time. Obviously you’ll need at least one year of historic data, but it’s better to have at least 2-3 years to spot patterns and trends that can help fine-tune future forecasting. Pay particular attention to lower or higher numbers, which should be apparent as they tend to stand out amidst otherwise consistent call volumes. Determine the cause for the variation, whether it was a holiday or a new company promotion, and adjust your forecast accordingly for that same time period. Many changes in traffic volume are not likely to repeat – on a day that a major news story breaks, call volume will go down. On a day when computers are knocked offline due to a technical glitch, call volume accuracy will be thrown off. Until you determine the cause, you will not be able to forecast a point estimate (the theory that a point in the future will be comparable to a similar point in the past). Once all of this data has been reviewed, you’ll be ready to prepare a forecast, assess staff requirements and create a schedule.

Close