Shrinkage: the hours you pay for but never get
Every operation budgets for some. Remote floors quietly double it, because the unplanned half became invisible.
By Andrés Martínez, founder of WorkPulse · Updated 2026-07-12
Call center shrinkage is the share of paid agent time that is not available for handling contacts. It divides into planned shrinkage (breaks, training, meetings, coaching) and unplanned shrinkage (lateness, early departures, extended breaks, and mid-shift idle time). Planned shrinkage is a scheduling input; unplanned shrinkage is a leak, and in remote operations it is the half that grows unnoticed because nobody can see the floor.
How do you measure shrinkage without guessing?
The traditional formula (shrinkage equals paid hours minus productive hours, divided by paid hours) is only as good as its inputs. On-site, supervisors estimate the inputs by looking at the floor. Remote, the estimate collapses: a full row of green "online" statuses says nothing about who is actually working.
Activity data replaces the estimate with a measurement. Paid hours come from shift check-ins; available hours are active time (real input, work applications, within shift); the gap between them, minus sanctioned breaks, is your unplanned shrinkage: computable per agent, per team, per day, without a single self-reported number.
Where does unplanned shrinkage hide in remote teams?
Four patterns account for most of the leak.
- Ghost shifts: checked in, machine on, nobody there. Visible as idle stretches with no break, or a session that never checked out.
- Break creep: 15-minute breaks that run 25. Small per incident, large multiplied across a floor and a month.
- Edge erosion: late starts and early stops of 5-10 minutes that never appear on a timesheet but repeat daily.
- In-shift drift: active at the keyboard, but in applications that have nothing to do with the queue.
How do you reduce it without micromanaging?
The goal is not surveillance pressure; it is making the invisible half of shrinkage visible enough that normal management works again.
Automatic idle-to-break conversion removes the ambiguity between "on a break" and "gone": past an idle threshold, the time books as break automatically, so shift totals stay honest with zero supervisor effort. Soft break-policy enforcement (a notification to the agent when a break exceeds policy, visibility for the supervisor, never a lockout) corrects break creep at the moment it happens instead of in a month-end report. Auto-closing abandoned sessions kills ghost shifts. And anomaly alerts on pattern breaks (an agent whose idle profile suddenly shifts) direct supervisor attention only where it is needed.
Notice what is absent from that list: screen-watching, keystroke quotas, and always-on webcams. Reducing shrinkage is an accounting problem before it is a behavior problem; most of the win comes from measuring hours honestly.
FAQ