Inbound staffing is a queueing problem, and it behaves in ways that surprise people who budget by dividing total calls by calls per agent. Calls do not arrive evenly. They cluster by hour and by day, and the number of agents needed to answer them within a target time rises sharply as the target tightens. Answering eighty percent of calls in twenty seconds may need noticeably more people than answering eighty percent in sixty, for the same volume. That is why a forecast by half-hour interval, not a monthly total, is the foundation of any honest inbound proposal, and why occupancy — the share of logged-in time agents spend on contacts — cannot safely be pushed much beyond the mid-eighties without burning people out and lengthening queues.
The measures worth agreeing fall into three groups. Accessibility: service level, average speed of answer and abandonment, read by interval rather than as a daily average that hides the bad hours. Resolution: first-contact resolution and repeat contact within seven days, which tell you whether answering quickly is achieving anything. Quality and experience: reviewed interaction scores and customer satisfaction. Average handle time belongs on the list as a planning input, not as an agent target; pressing it down reliably produces shorter calls and more of them.
Shrinkage is the term that catches buyers out. Agents are not available for every paid hour: breaks, training, coaching, meetings, holidays and sickness typically remove a substantial fraction of scheduled time, and a staffing plan that ignores it will miss its service level from the first week. Ask any provider what shrinkage assumption sits behind their numbers and how overflow is handled when a forecast is wrong — because on some days it will be, and the plan for that day is what separates one inbound operation from another.