Industry
Call center staffing ratios: shrinkage and span
Call center staffing ratios depend on shrinkage and supervisor span. Shrinkage turns scheduled agents into productive agents and affects supervisor seats.
What to take away
- Call center staffing ratios are calculated from three inputsrequired agents from a forecast, shrinkage, and span of control. Local conditions affect those inputs.
- Shrinkage is scheduled hours minus productive hours, divided by scheduled hours. Phoenix, Atlanta and Dallas differ in what drives the gap, not in the arithmetic.
- Wage floors and tax treatment shape the labor half of the math, not the ratio. Texas adopts the federal minimum wage by reference and has no state income tax, Arizona sets its own higher state minimum wage, and Georgia taxes income at a flat rate.
- No single published span fits every queue. Derive the span from coaching demand and write down the reason for the number you choose.
- Seasonality, commute time and tenure decide how much coaching time a supervisor has left.
How the ratio is built, step by step
A staffing ratio is the last number you calculate, not a target you set first. Call center workforce planning in the US usually runs on Erlang C for volume and a shrinkage table for hours.
Four steps to a ratio
- Forecast contacts per half hour
- Convert volume with Erlang C
- Gross up for shrinkage
- Divide by span of control
- Forecast contacts per half hour for a representative week, split by queue and channel, and use a service-level model to establish the productive agents required.
- Gross up that productive requirement for shrinkagescheduled agents = productive agents required ÷ (1 − shrinkage rate).
- Divide the relevant agent count by the supervisor span and round up. State whether the span uses scheduled or productive agents; the example below uses productive agents.
Erlang C can produce a productive-agent requirement from interval volume, average handle time and the service target. After that, the staffing-ratio arithmetic is direct: gross up for shrinkage, then divide by the stated span. The worked table below keeps the scheduled roster at 120 and applies the 1:15 span to productive agents, so it shows the shrinkage-and-span calculation rather than deriving demand from a contact forecast.
Shrinkage components and who owns each one
Shrinkage components and owners
| Shrinkage component | Planning input | What changes by city |
|---|---|---|
| Unplanned absence | Absence rate by day of week | Commute length and transit access |
| Breaks and rest periods | Contract and statute | Provincial rules at Canadian sites |
| Training and nesting | Ramp plan for new hires | Early attrition driven by wage pressure |
| Coaching | Hours per agent per month | Supervisor span of control |
| Team meetings | Calendar blocks | Peak season volume swings |
| System downtime | IT service level | Platform and vendor mix |
Phoenix: seasonal peaks and an indexed wage floor
Phoenix volumes swing with the winter visitor season. Insurance, travel and health plans add queues from November through March, then cut them. That swing favors a small supervisor core and a large variable agent pool.
Arizona sets its own state minimum wage, above the federal floor, and the U.S. Department of Labor's Wage and Hour Division lists the current state rate. Use it as a labor-budget input when pricing added seats; it changes roster cost, not the shrinkage formula. BLS figures for customer service representatives provide an employment and wage baseline for the labor side of the model.
Seasonal ramps widen spans further, because a new hire class needs more coaching in the months when volume is already high.
Atlanta: tenure, transit and adherence
Atlanta pulls agents from a wide metro area. Commute time on I-285 and MARTA access show up as start time variance, which is a shrinkage line item rather than a hiring problem.
Georgia's Department of Revenue has moved the state income tax to a single flat rate, which simplifies withholding assumptions. Keep any Georgia pay and withholding assumptions in the labor budget rather than changing the staffing ratio. Longer tenure cuts the training share of shrinkage.
Dallas: tenure and supervisor span
Texas has no state income tax, and its minimum wage law contains no dollar minimum of its own: the state adopts the federal minimum wage under the Fair Labor Standards Act by reference. Keep Texas pay and tax assumptions in the labor budget; they change the cost of a roster, not the conversion from productive agents to scheduled agents. The staffing calculation still depends on demand, shrinkage and the chosen span.
A wider span reduces supervisor seats in the arithmetic, but it is workable only if tenure leaves enough coaching capacity. If attrition climbs, the span narrows or coaching hours go unfunded.
Example: the same queue under three shrinkage rates
Hold the scheduled roster at 120 and vary only shrinkage. Assign supervisors against productive agents and round up at the 1:15 span.
Same queue, three shrinkage rates
Shrinkage
- 120
- 18%
- 120
- 22%
- 120
- 26%
Productive agents
- 120
- 98
- 120
- 94
- 120
- 89
Supervisors at 1:15
- 120
- 7
- 120
- 7
- 120
- 6
At the 1:15 span, 98 and 94 productive agents each require 7 supervisor seats after rounding up, while 89 requires 6. The scheduled roster stays at 120; the example isolates how shrinkage changes productive headcount and the supervisor count under this convention.
For a city-level comparison, the wage and local labor-supply detail sits in the separate staffing-pattern analysis. This calculation focuses on translating a productive-agent requirement into scheduled staffing and supervisor coverage.
