Most organisations have a document library full of SOPs, and almost none of them are used at the moment of work. The document is not wrong, exactly. It is simply describing a process that no longer exists, in a format that cannot be consulted while doing the job.
Four failure modes
Written for audit, not for use. The SOP was produced to satisfy a reviewer, so it contains controls, signatures and version history, and omits the two decisions that actually determine the outcome.
Described by department. A procedure written from one function's point of view cannot describe a handoff, and handoffs are where the work fails.
Frozen at design time. The process changed in week three. Nobody updated the document because nobody owns it between audits.
Longer than the task. If consulting the SOP takes longer than the step it describes, people will rely on memory and colleagues instead — which is rational behaviour, not indiscipline.
The format that survives
Replace the procedure document with a role playbook. One page per role, not per department. It answers five questions: what arrives, in what state, what I do, what leaves, and who receives it. Anything else belongs in a policy or a system configuration.
Then make the workflow the source of truth and the document a derivative. If the workflow lives in a system, the system should carry the instruction at the point of work. Where it does not, the playbook should be one page and it should be dated.
A procedure nobody reads is not a control. It is an artefact of one.
What to retire
Run a rationalisation pass and expect to retire 30–40% of the library. For each document, ask which decision it changes and which risk it mitigates. If the answer is neither, retire it. A smaller library that is accurate is worth more than a complete one that is not — and it is the only version your team will actually use.