Start with a verified identity before any reset

A password reset begins with a person, not with a ticket. The desk must confirm that the requester is the account holder before any link or temporary code is issued. Ask for two identifiers that are already on the account record, such as an employee number and a manager name. Do not accept an identifier that the requester supplies about themselves without a second check.

Use a verification method the client has approved in writing. The method may be a callback to the number on file, a confirmation through the manager, or a challenge answer stored in the directory. Record which method was used and when it completed. An unapproved method is a finding even when the reset itself was correct.

Treat a rushed request as a reason to slow down. Urgency is a normal feature of account problems and also a common pressure in social engineering. If the requester cannot complete verification, offer the escalation path rather than a shortcut. The desk should never trade identity proof for a faster close.

Write the verification result on the request record before the reset is performed. Include the identifiers checked, the method, and the name of the person who confirmed them. A later reviewer should be able to see why the desk believed the caller. A reset without that trail cannot be defended.

Record the request in one controlled queue

Every reset request should enter one queue with a stable identifier. The queue should capture the account, the system, the requester, the time, and the verification method. A request handled only in chat disappears from the count. One queue also prevents two specialists from resetting the same account twice.

Separate the request channel from the reset action. A specialist may take the request by phone but should perform the reset in the approved admin tool. Keep the tool name on the record so the next shift knows where the change happened. The channel is not the record.

Set the queue status to verified, waiting, completed, or refused. Each status needs a named owner and a timestamp. A request left in waiting without a next action is the same as an unread message. The status should tell the next person exactly what to do.

Protect the queue from duplicate entries created by multiple channels. Search by account before creating a new request, and link any duplicate to the original. Record the duplicate so the volume report stays honest. Duplicate resets are a common source of lockouts.

Separate a reset from an account unlock

A password reset and an account unlock are different actions with different risks. A reset changes a credential, while an unlock restores access to an existing one. Record the requested action before choosing a tool. Confusing the two can leave the user locked out after a successful reset.

Confirm whether the account is locked by policy, by failed attempts, or by a security hold. A policy lock may need a manager or security owner to release it. A failed-attempt lock may clear on its own after a defined period. The desk should not override a security hold to close a ticket.

Check whether the account has an active session or a registered device before the reset. Some systems require the user to sign out everywhere before a new password works. Record that step in the reply so the user is not sent into a loop. A reset that ignores sessions looks broken to the customer.

Close the request only after the user confirms access or the desk confirms the lock is cleared. Record the confirmation time and the method used to check. Do not mark a reset complete because the tool returned success. The user's access is the outcome that matters.

Handle privileged and shared accounts differently

A privileged account needs a second approver before any credential change. The approver should be named in the client's access policy rather than chosen by the desk. Record the approval reference and the time it arrived. A privileged reset without a second approver is a policy failure, not a convenience.

A shared account has no single holder to verify. Route the request to the account owner and ask for a written confirmation that names the work reason. Record the owner and the expiry for the temporary credential. Shared credentials should be rotated after the work ends.

Keep a list of accounts that the desk may never reset. The list may include finance system administrators, payroll masters, and security tooling. State the escalation contact beside each entry. A specialist should not have to guess which accounts are out of scope.

Check that the reset does not break a service account used by an integration. A service account password may be stored in a scheduler or connector. Changing it without the owner creates an outage that looks unrelated. Flag service accounts to the system owner before touching them.

Close with evidence and a review trigger

Close the record with the verification trail, the action taken, and the user confirmation. Add the tool, the timestamp, and the specialist name. The record should stand alone if the specialist leaves. A short note is not a substitute for the fields the client asked for.

Report resets by system, by request channel, and by verification method each week. A spike in one channel may point to a broken self-service page. A drop in verification may point to queue pressure. Use the pattern to fix the source, not to blame the desk.

Set a review trigger for any reset that skipped a required check. The trigger should name who reviews it and by when. Record the outcome and the correction. A skipped check that is never reviewed becomes the new normal.

Give the client a monthly summary of volume, refusals, and exceptions. Name the systems with the most requests and the checks that failed. Ask for one change to reduce the load, such as a self-service link or a clearer verification field. The desk should not absorb a broken process indefinitely.

For the next step, compare this workflow with Build a shift-start system check for a remote operations seat and Set a document retention rule for outsourced support records. For access and oversight, use the authoritative planning reference alongside your own policy.