How to Reduce Call Center Shrinkage: A Practical Guide

Shrinkage is the share of paid agent time that is not available for handling contacts: holidays, sickness, breaks, training, coaching, meetings, and admin. Reducing it is less about squeezing people and more about planning the time you lose so that it costs you less service.

This guide covers how to measure shrinkage properly, which parts you can actually reduce, and how to plan the rest. To see what your current shrinkage means for headcount, use the free contact center shrinkage calculator.

How much shrinkage is normal?

Published figures vary with definitions. A Contact Center Shrinkage Survey of 111 respondents by Knowlagent and NECCF found an average of 24%, while Call Centre Helper reports that practitioners usually calculate 30 to 35%. Compare yourself with centers that count the same categories, and track your own trend rather than chasing a headline number.

Step 1: measure shrinkage by category and by interval

You cannot reduce what you only know as one percentage. Split shrinkage into categories and measure each one by interval and by month.

• Planned and unavoidable: annual leave, public holidays, paid breaks, statutory training.
• Planned and movable: coaching, team meetings, projects, non-urgent training, admin time.
• Unplanned: sickness, short-notice absence, lateness, system downtime.

The same survey found that unavoidable items such as absence, vacation, and breaks made up about half of total shrinkage, and that the rest (training, meetings, coaching, after-call work) is more controllable. That second half is where reduction work pays off.

Step 2: move the movable shrinkage into quiet intervals

Coaching, meetings, and training do not need to happen at 10:00 on a Monday. Use the forecast to find the intervals where you are overstaffed and schedule offline activities there. The total hours lost may not change, but the service impact drops because the time is taken when contacts are low.

• Book recurring meetings in the lowest-volume intervals of the week.
• Give team leads a weekly window per team for coaching instead of ad hoc pulls.
• Run training in blocks during known seasonal dips.
• Re-plan offline work when the day changes, rather than cancelling it entirely.

Step 3: plan leave across the year

Leave shrinkage is unavoidable, but its timing is not. Peaks in school holidays and at year end are common. Set leave capacity per week from the forecast, publish it early, and let people request against it. That spreads leave instead of concentrating it in the weeks you can least afford.

Step 4: reduce unplanned absence at the source

Unplanned absence is the hardest category to forecast and the most damaging, because it arrives without notice. The levers are mostly about schedules people can live with:

• Publish schedules further in advance and change them less often.
• Give people a fair, rule-checked way to swap shifts instead of calling in sick.
• Watch for patterns such as repeated absence after night shifts or long runs of early starts.
• Keep a small, explicit buffer for the absence you know will happen.

Step 5: plan with the shrinkage you actually have

Some shrinkage will always remain. The mistake is planning as if it did not. Convert the productive agents you need into scheduled headcount with scheduled headcount = required agents / (1 - shrinkage), using the shrinkage for that interval and season rather than one annual average. The Erlang calculator gives the productive agents; the shrinkage calculator turns them into headcount.

Common mistakes

• Using one annual shrinkage percentage for every interval and month.
• Mixing definitions, for example counting breaks in one report and not in another.
• Cancelling all training and coaching during busy weeks, which raises attrition and handle time later.
• Treating shrinkage as an agent performance metric instead of a planning input.