Key takeaways
- Approval time and provisioning time have different owners.
- Measure verified first use, not ticket closure.
- Record waits, not just elapsed time.
Separate the intervals
An access request passes through approval, provisioning, and verification. Combining them into one elapsed time hides the stage where the delay occurs. The design records a timestamp at each transition and reports the intervals separately.
The final interval ends at verified first use, not at ticket closure. A ticket can be closed when the account is created, while the new hire still cannot sign in because a group or license is missing.
| Stage | Start | Stop | Owner |
|---|---|---|---|
| Approval | Request submitted | Approver decision | Manager or approver |
| Provisioning | Approval recorded | Account and groups created | IT |
| Verification | Provisioning complete | New hire signs in | IT or coordinator |
Record wait reasons
At each stage, record whether the request was waiting and why: approver unavailable, missing role profile, license shortage, scheduling, or a system outage. Wait reasons are the actionable part of the dataset.
Report elapsed time and active time together. Elapsed time describes the new hire's experience. Active time describes what the operation controls.
Sample and verify
Sample requests across roles, teams, and start dates. Include requests submitted late, because they are common and often excluded from service reporting.
Verify a portion of closed requests by checking the account, the group membership, and the new hire's confirmation. This catches the gap between a completed ticket and a working account.
Use the results
Report the median and the tail by stage and by role profile. A long approval tail may call for delegated approval. A long provisioning tail may call for a standard role bundle.
Do not use the measure to rank individual technicians. Use it to test whether the process and the role definitions are realistic.
Sources and methodology
Stage-based interval measurement across sampled access requests, with wait reasons recorded at each transition.
- NIST Digital Identity Guidelines 2025.
- NIST Security and Privacy Controls 2020.
- CISA current.
- U.S. Department of Labor current.
Source count: 4. Last verification date: October 22, 2026.
Questions about the findings
Why measure first use?
Because a closed ticket does not prove the account works for the new hire.
How are waits handled?
Each wait is recorded with a reason and an owner, separately from active work.
Can this set a service-level target?
It can inform a target, but the target is a management decision based on role risk and start-date needs.
Put the finding to work
Review the full research library, compare cluster coverage inside IT Onboarding, and pair these findings with our VA candidate screening support.
