Call Center OutsourcedPowered by StealthAgents

Call Center Outsourced evidence brief · Desk review · Published

What should a workforce plan assume about shrinkage in the first ninety days?

A first-ninety-day shrinkage model that separates planned time, new-hire attrition, and coaching load before the schedule is approved.

What should a workforce plan assume about shrinkage in the first ninety days? editorial illustration
On this page
  1. Key stats
  2. Key takeaways
  3. Question and scope
  4. What the evidence supports
  5. Define the shrinkage categories
  6. Measure the baseline
  7. Separate planned from unplanned
  8. New-hire attrition
  9. Training and nesting time
  10. Coaching and meetings
  11. Absence patterns in the first ninety days
  12. Schedule adherence during onboarding
  13. The supervisor span
  14. Overtime and its limits
  15. Build the plan with a range
  16. Test the assumption
  17. Report the gap to the client
  18. Review at thirty, sixty, and ninety days
  19. Document the assumptions
  20. Limits of the evidence
  21. Manager review questions
  22. Conclusion
  23. Put this into a support lane
  24. Related operating guides
  25. FAQs
  26. Sources

Key stats

  • 2.67 million U.S. customer service representative jobs in 2025 (BLS)
  • $21.53 median hourly wage in May 2025 (BLS)
  • 289,500 projected annual openings, 2025 to 2035 (BLS)

Key takeaways

  • Build first-ninety-day shrinkage from categories, not from the tenured-floor average.
  • Separate planned training, nesting, and coaching from unplanned absence and outages.
  • Model new-hire attrition by week and keep the class size in the staffing plan.
  • Review the gap at thirty, sixty, and ninety days and adjust the reserve.

Question and scope

This brief asks what a workforce plan should assume about shrinkage during an agent's first ninety days. The population is a new-hire class in one outsourced queue, and the unit is a scheduled hour that does not produce available phone time. The question is whether the standard shrinkage figure fits a class that is still training and ramping. The analysis covers categories, measurement, and review, not the legal classification of any absence. Record the class size, the start date, and the schedule model before comparing figures.

What the evidence supports

Federal wage-and-hour guidance requires accurate records of hours worked and payment for compensable time. Contact-center standards call for forecasting, scheduling, and performance measurement, but they do not supply a first-ninety-day shrinkage rate. The operating inference is to build the assumption from the class's own schedule and observed events. A mature-floor percentage applied to a new class will usually understate the time that is not available for calls.

Define the shrinkage categories

Write the categories before measuring anything: vacation, holiday, sick leave, training, nesting, coaching, meetings, system outage, and unexcused absence. Give each category an owner and a code in the timekeeping system. Separate the time that is planned in advance from the time that appears without notice. A single shrinkage number hides the difference and makes the plan hard to defend.

Measure the baseline

Pull the last two or three classes and calculate shrinkage by category for each week of the first ninety days. Use scheduled hours as the denominator and record the source of the hours. Do not compare a class that started in a holiday month with one that started in a quiet month. Show the range across classes, not only the average.

Separate planned from unplanned

Planned time includes training, nesting, coaching, and approved leave, and it can be built into the schedule. Unplanned time includes sick leave, outages, and no-shows, and it needs a buffer. A plan that treats all shrinkage as one number cannot decide whether to add a shift or hold a reserve. Report the two groups separately to the client. Show the reserve in hours and in people.

New-hire attrition

A class can lose people during training and in the first weeks on the queue. Model the loss by week and keep the reason categories honest. Do not assume the full class will reach the floor, because the schedule and the client forecast both depend on the number who do. Review the assumption with human resources before the plan is approved.

Training and nesting time

Training and nesting are productive investments, but they are not available phone time. Put the hours in the shrinkage model and in the coaching plan so the same time is not counted twice. A new agent who is in nesting for part of a shift should be scheduled for the remaining share, not for a full queue load. Record the transition date for each agent.

Coaching and meetings

A new class needs more coaching than a tenured team, and the coaching hours belong in the plan. Include team meetings, one-to-one reviews, and quality calibration in the planned category. If the coach is also scheduled for calls, the coaching time reduces queue capacity twice. Name the coverage for the coach's own queue time. A coach who is pulled into calls cannot deliver the plan.

Absence patterns in the first ninety days

Review absence by week and by reason for the previous classes. A spike in the first month may reflect onboarding logistics rather than a pattern that will continue. A spike after the first solo week may reflect the shock of full volume. Use the pattern to set a reserve, not to predict an individual's behavior. Ask human resources which reasons need a different code.

Schedule adherence during onboarding

New agents have more schedule changes than tenured agents because training and nesting move. Measure adherence against the schedule in force and code the approved changes. A low adherence score in the first month may be a scheduling artifact. Fix the coding before using the number in a performance review. Publish the schedule version used in the review. Record the approved change in the same system the adherence report reads.

The supervisor span

A supervisor with a new class has less time for the tenured team, and that effect belongs in the plan. Count the coaching hours and the escalation time the class will consume. If the span is already at its limit, add a coach rather than stretching the supervisor. A plan that ignores the span will look affordable and fail in the first month.

Overtime and its limits

Overtime can cover a gap, but it is not a shrinkage assumption. Repeated overtime raises fatigue and absence, which raises shrinkage in the next period. Use overtime as a short buffer with a named approval and a review date. Do not build a permanent schedule on the assumption that people will always accept extra hours. Track the decline in acceptance before it becomes a crisis.

Build the plan with a range

Present the shrinkage assumption as a low, middle, and high case for the first ninety days. Tie each case to a specific event, such as a larger class or a holiday start. Show the headcount and the service effect for each case. A single number invites a false sense of precision. Name the event that would move the plan to the high case.

Test the assumption

Compare the planned shrinkage with the actual for the first four weeks. Break the gap down by category and by week so the cause is visible. If the gap is in coaching, fix the coaching plan. If it is in absence, review the reserve and the support available to new agents. Set the next check before the four weeks end.

Report the gap to the client

Give the client a short table of planned versus actual shrinkage by category. State what changed and what the team will do about it. Do not present a single percentage without the mix. The client should see whether the plan needs a new hire or a schedule change. Keep the same categories in every report. Use the same categories in the client note and the internal review.

Review at thirty, sixty, and ninety days

Set three review points and write the questions for each. At thirty days, check training and nesting. At sixty days, check the first solo weeks and the coaching load. At ninety days, check whether the class has reached the tenured assumption. A review that happens only at the end cannot correct the plan. Write the owner for each review question.

Document the assumptions

Keep the class list, the schedule, the category definitions, and the source of the hours with the plan. Note which assumptions came from the client and which came from the operations team. A later reviewer should be able to see why the reserve was set at a particular level. Undocumented assumptions make the plan impossible to defend. Attach the class roster and the approved schedule.

Limits of the evidence

This brief cannot supply a shrinkage rate for a specific site, because the answer depends on the labor market, the schedule, the client's training, and the local law. Historical classes can inform a range, but they cannot predict a new class with certainty. The sample may be small. Use the result to set a range and a review, not to promise an exact headcount. Keep the ninety-day review separate from the annual performance cycle.

Manager review questions

Managers should ask which categories are planned and which are not, how the class size is expected to change, and who owns the coaching hours. They should also ask what the trigger is for adding a coach or a shift. The remaining question is which category has the widest gap. Record the answers with the period and the class data.

Conclusion

A first-ninety-day plan needs a shrinkage assumption built by category, not borrowed from the tenured floor. Training, nesting, coaching, attrition, and absence behave differently in a new class. Present a range, measure the gap every four weeks, and keep the review dates. The plan is sound when the client can see what the reserve is for and when it will be tested.

Put this into a support lane

Bring the current queue rules, the shift handoff, and one week of cases so the review can test a real promise.

Review the queue design

FAQs

Should training time count as shrinkage?

Yes, for capacity planning. It is planned and productive, but it is not available phone time, so it belongs in the model with its own category.

Can one shrinkage number cover every queue?

No. The mix depends on the schedule, the class, and the local labor market. Use a range and review it against actual hours.

Sources

  1. Fact Sheet #64: Call Centers under the Fair Labor Standards Act
  2. Fair Labor Standards Act
  3. Occupational Outlook Handbook: Customer Service Representatives
  4. ISO 18295-1:2017 Customer contact centres