WFM guideBenchmarks

Contact centre shrinkage benchmarks

Q: What is a typical contact centre shrinkage rate?

Most contact centres land between 30% and 40% total shrinkage — the proportion of paid time that agents are not available to handle contacts. Many operations cluster around 33-37%. But the headline figure is far less meaningful than its composition, because the same total can be healthy or unhealthy depending on what makes it up. A rough component breakdown for a typical operation: holiday / annual leave 10-12%; sickness / unplanned absence 4-8%; breaks 6-8%; training and coaching 3-6%; meetings, 1:1s, and team time 2-4%; other off-phone (admin, system issues, special projects) 2-5%. Add those up and you get to the 30-40% range. The important distinction: 'planned' shrinkage (holiday, scheduled training, breaks) is healthy and necessary — it does not indicate a problem, and you cannot drive it to zero. 'Unplanned' shrinkage (sickness, unaccounted time, excessive ad-hoc off-phone) is the part to scrutinise and manage. A 36% shrinkage that is mostly holiday and training is a well-run operation; a 36% shrinkage where 12% is unplanned absence has a serious attendance problem hiding inside a normal-looking total. Always build and report shrinkage by component, never as a single number to be benchmarked.

People want a single shrinkage number to benchmark against. The honest answer is a range — most contact centres land between 30% and 40% — but the number only means something when you build it bottom-up. A 35% shrinkage that is mostly holiday and training is healthy; a 35% that is mostly unplanned absence is a problem.

The headline range — and why it is not enough

Most contact centres run 30–40% total shrinkage, clustering around 33–37%. But quoting that total alone is misleading, because the same number can describe a well-run operation or a struggling one. The total is the sum of very different things: necessary, planned, immovable time (holiday, breaks, training) and avoidable, unplanned, manageable time (unplanned absence, unaccounted off-phone). The benchmark that matters is per component — and especially the size of the unplanned portion.

Shrinkage by component: typical ranges

Component	Typical range	Type
Holiday / annual leave	10–12%	Planned
Sickness / unplanned absence	4–8%	Unplanned
Breaks	6–8%	Planned
Training & coaching	3–6%	Planned
Meetings, 1:1s, team time	2–4%	Planned
Other off-phone (admin, system, projects)	2–5%	Mixed
Total	≈ 30–40%	—

Holiday / annual leaveA statutory and contractual entitlement — typically the single largest component. Cannot be reduced; must be planned and spread across the year to avoid coverage gaps.

Sickness / unplanned absenceThe component to scrutinise. Above ~6% sustained suggests an attendance or wellbeing problem. This is where Bradford Factor, return-to-work, and engagement work pays back.

BreaksPaid breaks and rest periods. Necessary and contractual. The lever here is timing (placing breaks off-peak), not reduction.

Training & coachingHigher during onboarding, new product launches, or peak-ramp periods. Investment, not waste — but must be scheduled into the plan, not treated as a surprise.

Meetings, 1:1s, team timeBriefings, performance reviews, team meetings. Legitimate but worth controlling — meeting creep quietly erodes available time.

Other off-phone (admin, system, projects)The catch-all. Includes legitimate admin and the unaccounted time that hides inefficiency or system problems. A large or growing 'other' bucket is a red flag worth investigating.

Planned vs. unplanned: the distinction that matters

Planned shrinkage (~24–30%)

Holiday, breaks, scheduled training, meetings. Necessary, contractual, and largely immovable. You cannot drive it to zero and should not try — agents are entitled to leave and breaks, and training is an investment. The job here is to schedule around it, not eliminate it. A high planned-shrinkage figure is not a problem.

Unplanned shrinkage (~6–13%)

Unplanned absence (sickness), unaccounted off-phone time, ad-hoc disruptions. This is the portion to scrutinise and manage — it is partially controllable through attendance management, engagement, wellbeing, and system reliability. A rising or large unplanned bucket inside a normal-looking total is the real red flag.

Build your shrinkage bottom-up, don't borrow a benchmark

The benchmark ranges above are for sanity-checking, not for plugging into your staffing model. Borrowing "35%" because it is typical is how staffing models go wrong — your operation's real shrinkage may be 31% or 39%, and the difference is several agents. Build it from your own data:

→Pull actual hours by category (holiday, sickness, breaks, training, meetings, other) over a representative period — ideally a full year for holiday seasonality.
→Express each as a % of total paid hours. Sum them for total shrinkage.
→Compare each component against the ranges above — investigate any that are well outside.
→Use YOUR total in the staffing model. Refresh it regularly; shrinkage drifts (training surges, absence spikes, seasonal leave).

Shrinkage benchmark questions

What is a typical contact centre shrinkage rate?

Most contact centres run 30–40% total shrinkage, often clustering around 33–37%. A rough component breakdown: holiday 10–12%, sickness/unplanned absence 4–8%, breaks 6–8%, training and coaching 3–6%, meetings/1:1s 2–4%, other off-phone 2–5%. But the total matters far less than its composition. 'Planned' shrinkage (holiday, breaks, training) is healthy and necessary — you cannot and should not drive it to zero. 'Unplanned' shrinkage (sickness, unaccounted time) is the part to scrutinise and manage. A 36% shrinkage that is mostly holiday and training is a well-run operation; a 36% where 12% is unplanned absence has a serious attendance problem hidden inside a normal-looking total. Always build and report shrinkage by component, and use your own data in the staffing model rather than borrowing a benchmark number.