Shrinkage

From WFM Labs

Shrinkage is the percentage of paid agent time that is not available for handling contacts. It is the gross-up factor between producing-FTE (the people on the phones) and required-FTE (the people on payroll). Shrinkage matters because it is the single largest correction between the Erlang-derived staffing requirement and the headcount and budget the operation actually commits to. A 25% shrinkage rate means a 100-FTE producing requirement becomes a 133-FTE required headcount — and the cost stack scales with the larger number.

Shrinkage is the most operationally controllable of the major WFM levers. AHT is a customer-by-customer property that the workforce can shape only at the margins. Volume is exogenous. Service Level is a target, not a lever. Shrinkage is a portfolio of small operational decisions — meeting cadence, training schedule, break design, off-phone activity governance — that compound into a number worth millions in any large operation.

Definition

Shrinkage is the share of paid time that is not on-phone:

Shrinkage = (Paid Time − On-Phone Time) / Paid Time × 100%

Equivalently, the staffing gross-up:

Required-FTE = Producing-FTE / (1 − Shrinkage)

A 30% shrinkage rate means each producing-FTE needs 1 / (1 − 0.30) = 1.43 required-FTE on the payroll.

Two scopes in operational use:

  • Off-phone shrinkage — every paid minute the agent is not on a contact (planned + unplanned). This is the number used in capacity planning.
  • Net shrinkage (also called non-productive shrinkage) — only the time spent on activities the operation considers non-productive. Excludes things like coaching and training that are productive but off-phone. This narrower definition is sometimes used in budgeting but is not the right number for staffing.

The right number for staffing is the off-phone number. Use a different number and the operation will be chronically understaffed.

Shrinkage categories

Shrinkage is best decomposed into four buckets:

  1. Planned, productive — training, coaching, team meetings, calibration, project work. Investments the operation has chosen to make. Largest shrinkage line in well-managed operations.
  2. Planned, non-productive — breaks, lunch, scheduled exceptions. Compliance-required.
  3. Unplanned, paid — sick time, bereavement, unscheduled-but-approved time off. Volatility line.
  4. Unplanned, behavioral — late starts, early outs, unscheduled breaks, unaccounted off-phone. Adherence-and-conformance line (see Adherence and Conformance).

A typical breakdown for a mature contact-center operation:

Category Range
Planned, productive (training, coaching, meetings) 5–12%
Planned, non-productive (breaks, lunch) 8–14%
Unplanned, paid (sick, PTO unscheduled portion) 4–8%
Unplanned, behavioral (adherence) 2–6%
Total shrinkage 19–40%

Operations with structurally lower shrinkage (BPOs, high-volume transactional queues) sit at the lower end. Operations with heavy training cycles, complex skilling, and back-office hybrids sit at the upper end.

Formula / mathematics

The capacity-planning identity:

Required-FTE = (λ × AHT × (1 + buffer)) / (utilization × (1 − shrinkage)) ÷ working-hours-per-FTE

Where utilization is the planned occupancy band (see Occupancy), shrinkage is the off-phone percentage, and buffer covers volume volatility. The two corrections are multiplicative — a 25% shrinkage and 85% utilization band require:

Multiplier = 1 / (0.85 × 0.75) = 1.57

Each producing hour requires 1.57 paid hours on the headcount line.

A 1-percentage-point reduction in shrinkage reduces required-FTE by approximately 1 / (1 − shrinkage)2 times the producing requirement. At 25% shrinkage, that's a 1.78% reduction in required-FTE per percentage-point of shrinkage. For a 1,000-FTE operation, one shrinkage point is roughly 18 FTE — large enough to fund significant operational investment with high ROI.

Practitioner use

Shrinkage shows up at three layers:

  1. Annual budgeting and headcount planning. Shrinkage assumption sets the headcount line. A budget built on 22% shrinkage and an operation realizing 28% will be chronically short.
  2. Interval-level scheduling. Schedule generation places agents into intervals respecting break, lunch, training, and meeting overlays. The shrinkage budget is realized at the interval level (see Schedule Generation).
  3. Performance management. Variance between planned and actual shrinkage is one of the cleanest indicators of operational health. Excess unplanned shrinkage indicates adherence breakdown; excess planned shrinkage indicates schedule design or training-program issues.

The big operational moves on shrinkage:

  • Stagger breaks and lunches. Naive break alignment puts everyone off-phone at peak. Properly-staggered breaks produce the same total shrinkage with much better interval coverage.
  • Time-shift training. Training during lower-volume intervals shrinks at lower opportunity cost.
  • Decouple coaching from peak. Coaching during lull intervals.
  • Reduce unplanned shrinkage. Adherence programs, schedule respect culture, root-cause analysis on absenteeism.

Each lever is small individually — collectively, they're the single largest non-AHT cost lever in WFM.

Common failure modes

  • Single-number shrinkage applied across all queues. Different queues, different shrinkage profiles. A specialty queue with high training and coaching needs has structurally higher shrinkage than a transactional queue. Apply the wrong number and one queue is overstaffed and the other is starved.
  • Reporting trimmed shrinkage and planning with full. Stripping unplanned-paid shrinkage out of the reported number for executive consumption while the WFM team plans with the full number — common pattern. Operations and finance end up working off different baselines.
  • Ignoring the multiplicative effect. Shrinkage and occupancy compound. A 5-point shrinkage error and a 5-point occupancy error don't add — they multiply, and the headcount miss can be 10–15%.
  • Excluding productive off-phone time from shrinkage. Coaching is productive but off-phone. Training is productive but off-phone. If they're not in the shrinkage number, they're invisible to the capacity plan. The operation will be understaffed in the periods around training cycles.
  • Using last year's shrinkage as this year's plan. Shrinkage shifts with hiring cycles, training-program changes, channel mix, and seasonality. A static shrinkage assumption is a Level 2 tell.
  • Treating shrinkage as fixed cost. Most shrinkage line items are choices — meeting cadence, training duration, coaching schedule. Treating it as fixed forecloses the largest controllable lever in WFM.

Maturity Model Position

  • Level 1 — Initial (Emerging Operations). Shrinkage applied as a single annualized number; categories not separated; variance treated as noise.
  • Level 2 — Foundational (Traditional WFM Excellence). Shrinkage decomposed into the four standard categories. Annual planned shrinkage feeds capacity planning; weekly actuals tracked against plan. Adherence programs run.
  • Level 3 — Progressive (Breaking the Monolith). Shrinkage modeled by queue, channel, and tenure cohort. Productive vs non-productive shrinkage explicitly distinguished in reporting and plans. Causal drivers (training cycle effects, schedule design, hiring waves) modeled. Distributional shrinkage feeds probabilistic capacity plans.
  • Level 4 — Advanced (The Ecosystem Emerges). Shrinkage treated as portfolio variable across human / hybrid / agentic pools (see Three-Pool Architecture). The economics of training and coaching investments — productive shrinkage that builds long-run capacity — modeled explicitly via Workforce Cost Modeling and speed-to-proficiency. Shrinkage variance becomes signal in Variance Harvesting loops.
  • Level 5 — Pioneering (Enterprise-Wide Intelligence). Shrinkage adapts continuously to operational state. Training, coaching, and time-off allocations optimized real time across pools. Shrinkage-as-investment formalized; the cost-vs-capability trade-off managed against learned long-run value.

References

  • Cleveland, B. (2012). Call Center Management on Fast Forward (3rd ed.). ICMI Press. Standard treatment of contact-center shrinkage categories.
  • Reynolds, P. (2010). Call Center Staffing. The Call Center School Press. Practical shrinkage decomposition for capacity planning.
  • Koole, G. (2013). Call Center Optimization. MG Books. Shrinkage in the staffing-translation math.
  • Lango, T. (2026). Adaptive: The Workforce Transformation Architecture. Shrinkage as portfolio variable; productive shrinkage as long-run investment.

See Also

Interactive tools

  • Time-to-Shrinkage Translatortime.wfmlabs.com. Converts each off-phone activity (team meetings, breaks, PTO, training, absence, coaching) from its natural unit (minutes per week, hours per year) into a daily shrinkage percentage. Shows total daily shrinkage at a glance.
  • Erlang Suiteerlangcalculator.wfmlabs.com. The staffing math that shrinkage grosses up.
  • WFM Variance Analysisoccupancy-variance-analysis.wfmlabs.com. Useful for seeing how shrinkage realization affects reported occupancy.