There is a question worth asking before any gap analysis: the last time a written procedure changed, was it because the work changed — or because an auditor wrote a finding?
The answer separates two organisations that look identical on paper. In the first, the document follows the operation: somebody changed how a task is done, and the procedure was updated to match. In the second, the document follows the audit: a finding was raised, a paragraph was added, and the work carried on exactly as before.
The second organisation passes audits for years. Its procedures are thorough, its records are complete, and its people have never read any of it — because the procedure does not describe what they do. The system is not wrong, it is simply about something else. Then an auditor asks an operator a question instead of asking the manual, and the gap that was always there becomes a finding.
This is why re-writing the documents rarely fixes it. A better-written procedure that still describes an imaginary operation fails the same way, one cycle later. What has to change is which direction the writing flows: the system has to be built from how the work is actually done, including the parts that are inconvenient to write down.
The practical test is unglamorous. Take one high-risk task, watch it performed, and read the procedure beside it. If the two disagree, the document is not the problem to fix first — the disagreement is.
The last time a written procedure changed — was it because the work changed, or because an auditor wrote a finding?