Shrinkage Calculator (Call Center)
Updated August 2026 · Free · No sign-up
Calculate call center shrinkage — the share of paid time agents aren't available for customers — and the staffing uplift needed to cover it, with the benchmark bands and reduction levers workforce managers use.
Quick answer: Shrinkage % = (non-productive hours ÷ total paid hours) × 100 — typically 30-35% in call centers. Staffing must inflate accordingly: Required staff = base requirement ÷ (1 − shrinkage); needing 100 productive agents at 30% shrinkage means scheduling ~143.
What Shrinkage Is
Shrinkage is every paid hour an agent can't take contacts — the gap between scheduled bodies and available capacity. Workforce management splits it two ways:
| In-office (planned) shrinkage | Out-of-office shrinkage |
|---|---|
| Breaks & lunches | Sick absence / no-shows |
| Team meetings & 1:1s | Vacation & holidays |
| Training & coaching | Lateness, early departures |
| System downtime, after-call admin* | FMLA/other leave |
*Definitions vary — some centers count ACW as productive. Define once, measure consistently.
The Two Formulas
Shrinkage % = Non-Productive Paid Hours ÷ Total Paid Hours × 100
Scheduled Staff Required = On-Queue Requirement ÷ (1 − Shrinkage) — division, not addition: covering 30% shrinkage takes 43% more staff, a mistake ("just add 30%") that under-staffs every interval.
Worked Examples
Example 1 — measuring. A 10-agent team, 1,600 paid hours in a month, 520 non-productive: 32.5% shrinkage — inside the normal band, but see the split below.
Example 2 — staffing. Erlang says 100 agents on-queue at peak; at 30% shrinkage schedule 100 ÷ 0.7 = 143 agents. Adding "30" instead (130) leaves every peak ~9 agents short — hello, service-level misses.
Example 3 — the improvement worth money. Cutting shrinkage 35% → 30% on a 200-seat requirement: 308 → 286 scheduled = 22 fewer FTE, roughly $900k/year at loaded cost — why WFM teams fight for every point.
Benchmarks
| Shrinkage | Reading |
|---|---|
| Under 25% | Excellent — verify nothing productive is misclassified |
| 25-35% | Normal industry band |
| 35-45% | High — audit the absence component first |
| Over 45% | Critical — staffing model is fiction until fixed |
Managing Shrinkage Without Squeezing Humans
- Split planned vs unplanned first: breaks/training are schedulable — place them off-peak and they stop hurting service level; the enemy is unplanned absence, which lands exactly where forecasts can't flex. Track that component with the absenteeism rate calculator.
- Schedule shrinkage, don't suffer it: meetings and coaching booked into forecast valleys convert "shrinkage" into planned capacity shaping.
- Interval-level truth: monthly shrinkage averages hide that Monday 9-11 AM runs 45% — measure at the interval where service level lives.
- Don't cut the good kind: training and coaching are performance investments; slashing them buys this quarter's shrinkage with next quarter's quality. The lever is absence and adherence, not development.
- Adherence ≠ shrinkage: adherence measures following the schedule; shrinkage measures what the schedule contained. Confusing them misdiagnoses every staffing miss.
Shrinkage in the Staffing Chain
The full workforce math: forecast contacts → Erlang/queuing gives on-queue requirement → ÷ (1 − shrinkage) gives scheduled requirement → ÷ occupancy targets sanity-check burnout → convert to bodies via the FTE calculator (part-timers change the headcount, not the hours). Budgeting the roster also means loaded cost per FTE (payroll taxes via the employer payroll tax calculator) — which is exactly why each shrinkage point on a big floor is six figures.
Beyond Call Centers
The same arithmetic staffs nurses (indirect care time), field service (travel/admin), retail (register vs backroom), and any operation where paid ≠ productive. Generic workplaces usually discount capacity 15-20% for meetings/admin — call centers just industrialized the measurement.
Frequently Asked Questions
What is shrinkage in a call center?
The percentage of paid time agents are unavailable for contacts — breaks, meetings, training, absence. Formula: non-productive paid hours ÷ total paid hours × 100.
What is a normal shrinkage percentage?
25-35% is the standard industry band; over 40% signals an absence or scheduling problem worth auditing immediately.
How do you staff for shrinkage?
Divide, never add: required scheduled staff = on-queue requirement ÷ (1 − shrinkage). 100 needed at 30% shrinkage means 143 scheduled, not 130.
What counts as shrinkage?
Everything paid but off-queue: breaks, lunches, meetings, training, coaching, system downtime, sick leave, vacation, lateness. Some centers treat after-call work as productive — define consistently.
What is the difference between shrinkage and occupancy?
Shrinkage removes agents from the queue; occupancy measures how busy the remaining on-queue time is. High shrinkage with high occupancy = burned-out understaffing.
What is the difference between shrinkage and adherence?
Shrinkage is what the schedule contains (planned off-queue time); adherence is whether agents follow the schedule. Different problems, different fixes.
How do I reduce shrinkage?
Attack unplanned absence (return-to-work conversations, fair scheduling), move planned activities to forecast valleys, and measure at interval level — but protect training; cutting it trades quality for a cosmetic point.
How much does a point of shrinkage cost?
On a 200-agent requirement, each point ≈ 4 scheduled FTE ≈ $160-200k/year loaded — the business case behind workforce management.
What shrinkage should a new call center plan for?
Plan 30-35% until you have your own data — optimistic 20% assumptions are the root cause of most new-center service-level failures.
Does shrinkage apply outside call centers?
Yes — any staffing model where paid time exceeds productive time: healthcare, field service, retail. Offices implicitly run 15-20% via meetings and admin.
✓ Formula verified • Last updated: August 20, 2026