Diagnose the failure mode
Unreliable output can come from an unclear role, missing access, incomplete training, unrealistic volume, poor judgment, or weak performance. Replacing the person will not fix a workflow that only exists in the owner's head.
Choose three recent misses and trace what the person received, what decision they had to make, what standard applied, and where the handoff broke.
Turn expectations into a working system
Define weekly outputs, turnaround times, templates, examples, and escalation rules. Show both acceptable and unacceptable work. Use a checklist for high-risk tasks such as candidate records, client messages, and document handling.
Review a small sample on a predictable schedule. Feedback should identify the rule to use next time, not just correct the current item.
Decide whether the role still fits
If the person can perform with clearer inputs and feedback, expand responsibility gradually. If the work consistently requires owner judgment, redesign the role. If expectations are clear and performance remains unreliable, document the pattern and transition the work responsibly.
Review confidentiality, system access, data location, offboarding, and worker-classification considerations whenever an outsourced relationship changes.
Agency checklist
- Trace three recent failures
- Define outputs and turnaround times
- Provide examples and decision rules
- Create an escalation path
- Sample work on a set rhythm
- Remove access cleanly when the relationship ends