How do you set your forecast accuracy goals? Try leveraging this tool to explore forecast accuracy…
In Defining Forecast Objective set up, is it ok to put all three – Service Level, Occupancy, and ASA?
A couple of comments - Let’s separate Occupancy for a second. Most contact centers and WFM teams pursue an accurate forecast to ensure they have the right staff to answer the forecasted demand in a timely fashion. We leverage Service level and ASA measurements to ensure we’re hitting our promise to be “available” to our customers. Whether hitting an 80% answer in 30 seconds or achieving an ASA of 25 seconds - those are all measurements that speak to your performance on that promise.
In the forecast accuracy tool we built for demonstration at WFM Labs, we’re demonstrating how realistic your forecast accuracy target may be depending on the raw size of the queue and the number of operating hours you are open. This is a mechanism to demonstrate Ger Koole’s minimal “WAPE” - which we inherently understand as “small queues are more difficult to forecast than larger queues”. Yet, when we say let’s hit +/-5% on this particular group, how are we establishing that target? The tool looks at three inputs: weekly call volume, your operating hours over which that call volume is spread, and the distribution by day of the week.
Note: These are not the only things you may wish to examine when setting forecast accuracy objectives, but it is a good place to start.
Now, with that said, you certainly could consider SL / ASA when establishing your objectives; if you’re running two groups, one which has a 90% answered in a 20-second goal and a second group with a 60% in 60-second goal - what does that tell you? It tells me the business is less concerned with the second group with a lower service objective, and hence may not be as concerned about getting the forecast as accurate as possible compared to the 90/20.
As for Occupancy - that’s a different topic altogether, which I am sure will be talked about as we get topics flowing on this board!
I don’t look at Occupancy as a objective/goal. It is an output of the other metrics you are measuring to. I will use an occupancy cap, a percent of occupancy I don’t want to exceed, because I know high occupancy will impact service level and agent satisfaction.