This research is general operational information, not legal, tax, accounting, employment, benefits, privacy, cybersecurity, insurance, or financial advice.

Question and scope
This is a proposed research protocol, not a report of completed findings. It asks what share of vendor bank-detail changes are verified by callback before the record is updated. The result would describe the verification control. Keep the evidence with the record rather than in a separate log.
The protocol does not measure whether a fraud attempt was detected. It measures whether the required verification was recorded. Keep the boundary visible in the place where the work is assigned. Keep the evidence with the record rather than in a separate log.
Population and unit of analysis
Define the population as vendor bank-detail change requests during a stated window. The unit is one request, with the requester, the verification method, the approver, and the update date recorded. Separate what was observed from what was assumed. If the source is unclear, route the question instead of guessing.
Link the request to the vendor master so the same change is not counted twice. Write the deadline as a date and a local time. Separate what was observed from what was assumed. Record the date and the source so the next reviewer can repeat the check.
Definitions and measures
Define verification as a callback to a number already on file, with the contact's name and the time recorded. Define a second approval as a separate approver who did not receive the request. Write the deadline as a date and a local time. Name the owner in the same line as the action.
Report the verification rate and the second-approval rate separately. A request can be verified but not approved. Keep the evidence with the record rather than in a separate log. Record the date and the source so the next reviewer can repeat the check.
Evidence standard
Link each request to the callback note, the approver, and the updated record. A note that says verified without the number and the contact is weak evidence. If the source is unclear, route the question instead of guessing. Keep the original note beside the corrected version.
Where the verification was done by email, mark it as a different method. The study should distinguish a callback from a reply. A control that depends on memory is not yet a control. Review the first few notes together before the process expands.
Sampling and collection
Use a census of the requests if the volume is manageable. If sampling is necessary, stratify by vendor and amount. Select the sample before reviewing verification. Separate what was observed from what was assumed.
Pilot the classification on a few requests and compare the reviewers. Clarify how a request that was rejected is treated. If the source is unclear, route the question instead of guessing. Record the exception and the person who approved it.
Analysis and reporting
Report the verification and approval rates by vendor and request source. Show the requests that were updated without a callback. A high verification rate with a weak evidence standard is not a strong control. If the source is unclear, route the question instead of guessing.
Present the results as a description of the process, not a judgment about the team. Name the step that should be strengthened first. Ask whether a new reviewer could act on the record without a conversation. Record the date and the source so the next reviewer can repeat the check.
Limits and sources of error
The study may miss changes made outside the main system. State the systems included. Name the owner in the same line as the action. Write the deadline as a date and a local time.
The callback note may be written after the fact. Record when it was created. Use the same format for every handover so the next reader can scan it. If the source is unclear, route the question instead of guessing.
Review cadence
Repeat the review after a process change or at the same interval each quarter. Compare the same request sources and the same evidence standard. Keep the original note beside the corrected version. Name the owner in the same line as the action.
Assign the follow-up to the AP or security owner and set a date. The protocol should produce a specific decision about the callback rule. Keep the original note beside the corrected version. Name the owner in the same line as the action.