Constructive Use of Occupancy

Navigating through a glossary of call center metrics can be like a nightmare; at first glance, many of the terms seem like synonyms (e.g., occupancy, availability, and utilization). In addition, some call center managers use their own custom formulas - and the results can get messy. To a workforce manager whose staffing plan may hinge on a pass/fail service result if they are understaffed by even a single headcount, this is a recipe for disaster. Unfortunately, the test for gaps doesn't usually occur until the service level is missed, not made, which means it is already too late.

Occupancy vs. Utlization vs. Availability

The simplest way I have found to classify each metric is to return to the root definition, and stick to the same meanings that other systems in the industry have applied.

Occupancy defines the time the agent is occupied. This includes the time talking, being in after-call work, performing scheduled off-phone activity, handling work in-between calls, attending meetings, on a paid break, getting trained, or having a review. It only applies to working hours, not payroll hours, so eliminate the time when they're on vacation or sick from this calculation. But exclude any idle time when an agent is waiting on their next call. This is important, because that level can then be used to gauge the balance between boredom vs. burnout.

Utilization defines the phone time the agent is used. This includes the time talking, being in after call work, and that idle time when they’re waiting on their next call. This is the same definition the Erlang add-in for Excel uses when calculating its "utilization." Idle time is an output, a result of volume and service thresholds. When idle time is too high, and there is not enough in-between call activity to fill up the empty space, you may need to reevaluate the sweet spot of the service goal with your specific volumes. If you don’t already have it, a customer patience study of a scatter diagram charting the time to abandon against the service time will demonstrate what your customers will tolerate. Measuring the work that happens in-between calls (IBC) is helpful because those hours can be subtracted right out of the idle time and show instant improvement instead of falsely indicating that you have underworked agents.

Availability defines the time the agent is available in the phone system, so it includes AUX time, and is a straight-up result of idle time. Staying consistent with this definition will also help avoid confusion.

Considering the 3 "Human Factors"

Since the occupancy metric relies on knowing the working hours, there are three different "Human" factors that must be established first: Presence, Shrinkage and Non-working Staff.

The first Human Factor is Presence. Presence accounts for agents present, but also considers those agents that are paid, but may be working remotely or are performing work for a different department. Presence also considers agent vacations, sick time, personal days, holidays, collaboration efforts and travel. While you’re establishing this time, it’s also a good idea to tag those that are controllable (things you can cancel).

The second Human Factor is Shrinkage, which counts quality, training, meetings, and schedule adherence. Any scheduled work time that is a result of quality rather than a new unit of customer work should count toward shrinkage, not workload.

The third Human Factor is Non-working Staff. Your Non-working staff includes people on FMLA, extended jury duty or military leave. Non-working Staff, Shrinkage, and Presence are all Human Factors, but it will only be the Shrinkage time that belongs to working hours, which ultimately leads to the precise occupancy result.

Schedule Efficiency

Efficiency is the final layer to consider, and for a workforce manager, this refers to the efficiency of the schedules. The more flexibility the call center’s scheduler has in generating shifts, the higher the efficiency rate will be. A fully-optimized and efficient schedule set will yield the least expensive, lowest FTE required, highest service producing result. Once restrictions start burdening the schedules, such as consecutive days off or consistent start times, the efficiency rate goes down, because more people are required to handle the same workload under those assumptions. This efficiency rate is one of the most misunderstood concepts of workforce management. It is easy to understand that AHT times volume equals workload, and workload translates to bodies-in-chairs, which is inflated by Human Factors to give total required staff. But without adding in schedule efficiency, that "total required staff" number isn’t really the total amount required, unless you can force your agents to work in 30-minute segments at a time. Explaining why you are still understaffed after the hiring plan is already done is an unpleasant problem that can be avoided with a good simulation and recurring reviews.