Forecaster

Forecasting contact volumes and average handling times is one of the most critical steps in planning your workforce. Forecasts of future contact volumes and average handling times form the basis for agent schedules, vacation plans, and long term strategic plans. Inaccurate forecasts result in lower customer service levels, higher costs, or both. Thus, accurate forecasts are essential to the success of your contact center operations.

Forecasts of contact volumes and average handling times are generated by analyzing the seasonal patterns and trends in the historical data, and any additional information available about the future contact volume drivers. Your forecast accuracy is limited by the forecasting technology used in your WFM system to analyze seasonal patterns and trends in the data. Traditional WFM systems rely on the outdated weighted moving averages based forecasting technology. Some WFM systems even require you to provide weights and other information regarding these patterns while others use simple methods to compute them. The following table compares the forecasting technologies used by the traditional WFM software and the AWO Forecaster.

View
NEXT Generation Forecasting Technologies (2:31)

In today’s complex multi-skill, multi-site, multimedia contact centers, forecasting contact volumes and average handling times accurately is a challenge. Intra-day, daily, weekly, monthly, seasonal patterns and long-term trends, contact volume drivers, and calendar events such as the 1st or 15th of a month, Mothers Day, and other special events like promotions require powerful forecasting technologies to develop accurate forecasts. The AWO Portal provides the NEXT Generation forecasting technologies with the most advanced time-series algorithms optimized with the artificial intelligence technology to achieve unsurpassed forecast accuracy.

• Expert System – BestPick, evaluates, with the click of a button, all forecasting techniques available in AWO using historical data to determine the best forecasting method and optimal parameters for each contact channel (e.g. phone, email, fax, web chat, etc.)
• State of the art forecasting methods include:
  • Box-Jenkins ARIMA models
  • Single, Double, Triple, Exponential Smoothing
    (over 10 versions)
  • Additive and Multiplicative Seasonality Models
  • Dynamic Regression
  • Curve Fitting
  • Moving Averages
• Forward looking forecasting using customer contact drivers and indicator variables, together with contact history (e.g. number of cable service subscribers with billing contact history)
• Special event period forecasting for promotions campaigns, and calendar events (e.g. 1st and 15th of a month, Thanksgiving Day, Mother’s Day)
• Daily, weekly, monthly, quarterly, and annual forecasting in addition to 15-, 30-, 45-, and 60-minute interval forecasting
• Automated statistical extreme data and missing value detection and smoothing
• Extensive “What-If” forecasting and analysis capabilities for unlimited number of scenarios
• Collaborative forecasting capabilities for web enabled collaboration among multiple locations and organizations
• Centralized database to store data and generate forecast for different contact groups, media, sites, and environments at any level (e.g. interval, daily, weekly, monthly or annual)

Forecast Accuracy Tracking and Improvement

To achieve best-in-class accuracy, your forecasting process should be a closed-loop process with continuous improvement capabilities.
The AWO Forecaster’s Forecast Accuracy Scenarios module allows you to track forecast accuracy automatically for any period and type of forecasts including short range forecasts used for scheduling, and long rage forecasts used for strategic planning and budgeting.

By tracking forecast accuracy over time, the AWO Forecaster helps you detect changes and trends in your forecasting performance, and provides the input you need for your root-cause analysis.