Why your shrinkage calculations are failing your floor

Shrinkage is the total time for which agents are paid but are not available to handle customer interactions. It is calculated by dividing the total hours of non-productive time—such as breaks, training, and absenteeism—by the total scheduled hours, then subtracting that result from one. Accurate shrinkage management ensures that staffing requirements account for the reality of the floor rather than an idealized, 100% attendance scenario.

Key takeaways

Distinguish between internal and external shrinkage to identify which metrics operations can actually control.
Calculate shrinkage monthly to account for seasonal variations like holiday leave and seasonal illness.
Factor in "shrinkage on shrinkage" to ensure your final headcount calculation accounts for the fact that the backup staff will also need breaks and training.
Use conversation intelligence to reduce manual QA time, which is a significant driver of internal shrinkage.

What is shrinkage and why does it deviate from your plan?

In a perfect world, every agent you hire would spend every minute of their shift at their desk, ready to help a customer. In reality, agents need to eat, learn, attend team huddles, and occasionally take a sick day. Shrinkage is the metric that captures this reality. It is the percentage of time that agents are being paid but are "shrunk" away from their primary job of handling interactions.

When WFM leads fail to account for shrinkage accurately, the result is a perpetual state of understaffing. You might have 100 agents on the payroll, but if your shrinkage is 35%, you only have 65 agents available to take calls. If your Erlang C model says you need 80 agents to meet your service level, you are doomed to fail before the day even begins.

Understanding shrinkage requires moving beyond a single, static number. It is a dynamic force influenced by company policy, management style, and external environment. Gartner’s Customer Service & Support practice emphasizes that data protection and domain-specific training are becoming larger parts of the agent workload, which directly increases the time agents spend away from the phones.

Categorizing the "Missing" Hours: Internal vs. External

To manage shrinkage, you must first categorize it. Most operations divide shrinkage into two buckets: internal (controllable) and external (uncontrollable).

Internal Shrinkage

This includes activities that happen within the office or under management’s direction. Because these are scheduled, they are the easiest to track but often the hardest to reduce without impacting quality.

Training and Coaching: Essential for performance but a major drain on availability.
Team Meetings and Huddles: Necessary for alignment.
1-on-1 Sessions: Critical for retention and growth.
System Downtime: Technical issues that prevent agents from working.
QA Review: The time agents spend reviewing their scores or attending feedback sessions.

External Shrinkage

These are factors that occur outside of the immediate control of the floor manager, often related to HR policies or personal agent needs.

Absenteeism: Unplanned call-outs and sick leave.
Tardiness: Late starts or extended lunch breaks.
Paid Time Off (PTO): Vacations and personal days.
FMLA/Medical Leave: Long-term absences.

By separating these, a WFM lead can see if a staffing gap is caused by a flu outbreak (external) or an over-scheduled training calendar (internal).

How to calculate your true shrinkage rate

The most common mistake in shrinkage calculation is using a simple "gut feel" or an outdated industry average. To get a precise number, you need to look at your historical data over the last 12 to 24 months.

The Formula: Shrinkage % = (1 - (Total Productive Hours / Total Scheduled Hours)) * 100

For example, if your team is scheduled for 1,000 hours in a week, but after breaks, meetings, and absences, they only spent 650 hours logged in and available, your shrinkage is 35%.

The Staffing Multiplier: Once you have your shrinkage percentage, you use it to calculate how many people you actually need to hire. If your Erlang C calculation says you need 50 agents to hit your Service Level Agreement (SLA), you don't just hire 50 people.

Total Staff Needed = Erlang Requirement / (1 - Shrinkage %) Total Staff Needed = 50 / (1 - 0.35) = 50 / 0.65 = 77 agents

In this scenario, you need 27 extra people just to ensure that 50 are actually on the floor at any given time. This is the "buffer" that keeps your service levels stable.

The Erlang C connection: Why 30% is the dangerous default

Many contact centers default to a 30% or 35% shrinkage figure because it is a common benchmark. However, Metrigy research into CX and AI success metrics suggests that high-performing centers often have higher shrinkage because they invest more in agent coaching and development.

If you use a 30% default but your actual shrinkage is 40%, you will be understaffed by a large share of your required workforce every single hour. This leads to a "death spiral":

Understaffing leads to higher occupancy (agents working back-to-back calls).
High occupancy leads to burnout.
Burnout leads to higher absenteeism (external shrinkage).
Higher shrinkage leads to more understaffing.

To break this, WFM leads must use real-time data from platforms like NICE or Genesys to adjust their shrinkage assumptions weekly. If you notice absenteeism rising on Mondays, your shrinkage factor for Mondays should be higher than it is for Tuesdays.

Tactical ways to recover lost time

You cannot eliminate shrinkage, nor should you try—agents need breaks and training to function. However, you can optimize how that time is used.

1. Optimize QA and Coaching

Manual QA is a significant driver of internal shrinkage. When supervisors have to spend hours manually pulling calls and agents have to sit through long feedback sessions, that is time away from customers. By designing QA scorecards that agents actually respect, you can make these sessions more efficient.

Furthermore, using a conversation-intelligence layer like Hear.ai allows QA teams to gain full coverage across all calls automatically. This reduces the "search and find" time for supervisors and allows coaching to be targeted and brief, rather than long and exploratory. Shorter, more effective coaching sessions return agents to the floor faster.

2. Staggered Training Blocks

Instead of pulling an entire team off the floor for a two-hour training session, break the training into 20-minute micro-learning modules. This prevents a massive spike in shrinkage that could crash your service level for the entire afternoon.

3. Address "Unaccounted For" Time

Often, the biggest drain is not the scheduled 15-minute break, but the extra 3 minutes it takes an agent to get back from the break room. This "tardy shrinkage" can be addressed through better real-time adherence monitoring in your CCaaS platform, such as Five9 or Talkdesk.

Leveraging technology for visibility

Modern workforce management requires a tech stack that talks to itself. Your ticketing system (like Zendesk or Salesforce Service Cloud) should feed data into your WFM tool to show exactly when agents are active.

When you pair these with specialized tools, the visibility improves. For instance, Hear.ai's compliance monitoring can flag risk automatically, meaning agents don't have to be pulled into "just in case" compliance retraining unless a specific issue is detected. This moves training from a broad shrinkage category to a surgical, as-needed intervention.

Forrester’s CX Index shows that customer experience quality is often tied to agent availability. If your agents are constantly "shrunk" away from their desks, your CX scores will inevitably dip as wait times rise. Managing shrinkage isn't just a math exercise; it is a fundamental part of your brand's customer experience strategy.

FAQ

What is a typical shrinkage range for a contact center? While it varies by industry, most professional contact centers see shrinkage between 30% and 40%. Centers with high training requirements or generous PTO policies may see numbers closer to 45%.

Does shrinkage include paid lunch breaks? If the agent is being paid for the time but is not available to take a call, it should be counted as shrinkage. If the lunch is unpaid, it is simply excluded from the total scheduled hours in your calculation.

How often should I update my shrinkage factor? At a minimum, you should recalculate your shrinkage monthly. However, for high-volume environments, a weekly review of the previous four weeks' rolling average provides the most accurate forecast for the upcoming month.

Is high shrinkage always a bad thing? Not necessarily. High shrinkage might indicate that a company is investing heavily in agent development, 1-on-1 coaching, and wellness. The goal is not to have the lowest shrinkage possible, but to have the most accurate shrinkage forecast so you can staff accordingly.

To learn more about optimizing your team's performance, see our guide on how to build a modern contact center QA program.