This blog is the second in a three-part series focused on forecasting in the contact center. The first installment discusses why forecasting is both an art and a science.
A breed of goats found only on the slopes of Mt. Etna, Europe’s largest active volcano, seem to have a sixth sense that helps keep them safe – they behave erratically just before the volcano erupts. Their unusual behavior was first noted by a goatherder who has worked on the mountain for his entire life, then corroborated by scientists who confirmed his theory. No one knows exactly why – it could be the vibration of the ground or a smell in the air – but the goats do in fact seem to be able to sense when the volcano will erupt. Perhaps, an OPB segment featuring the goats concluded, the next time you see a humble goat running in the opposite direction, you may want to follow it.
This strange yet true story of goats able to forecast Mt. Etna’s eruptions can also serve as a handy reminder of an acronym that’s useful in guiding forecasting in your contact center – GOATS. This simple process can help you ensure that you accurately align staffing with customer demand.
- Generate: Generate forecasts with the best quantities and statistical models you have access to. AI and machine can help you understand future workforce needs efficiently and accurately by incorporating data on dozens of factors, including historical data, seasonality, cyclicality, multi-skill capabilities and omni-channel routing. For maximum accuracy, create multiple forecasts using different models and compare the results.
- Others: Take those results and discuss them with others who have knowledge of your business to gather opinions. Do this in one-on-one settings to avoid group or confirmation bias and include things that affect forecasts that the models do not know about (e.g., weather, news and business changes or decisions). Ask these people to explain their reasoning; studies indicate that those that have to verbally explain their reasoning tend to develop more consistent reasoning. Above all, actively question their assumptions – and yours.
- Adjust: Use the input and your knowledge to adjust as necessary – as detailed in our previous blog, this constitutes the “art” of forecasting.
- Track: Track the results against the people you talked to. Some people will be consistently less accurate with their suggestions, while others will be more accurate – again, our previous blog addresses political scientist Phillip Tetlock’s work identifying and bolstering the ability of “super forecasters,” or people whose predictions tend to be more accurate than those of the general population.
- Specific person accuracy: Add specific person accuracy (which accounts for the accuracy of each person that provided input) into the next decision round to make forecast adjustments.
In the next installment of our blog series on forecasting, we’ll take a deeper look at forecast intervals and reforecast frequency. Learn more about how NICE WFM puts AI and machine learning to work in your contact center to understand future workforce needs efficiently and accurately.