The short answer
Review an SOP whenever something it depends on changes, and give every procedure one named owner, one current copy in one known place, and a version header that says what changed and when. When a change is needed: fix the source SOP first, regenerate the working checklist from it, retire the old version the same day, and announce what changed — not just that something changed.
A calendar reminder alone will not keep procedures current, because most staleness is caused by events, not by time. The routine below uses both: events force an immediate review, and a light scheduled pass catches the quiet drift in between.
Why SOPs go stale in the first place
The work changed, the document did not
A system was replaced, a supplier changed a form, a step moved to a different person. Each change was small enough that nobody opened the SOP — but together they move the real procedure away from the written one.
Fixes live in the wrong copies
Someone corrects a printed sheet, a saved PDF, or the checklist — but not the source. The fix helps one shift, then disappears at the next revision, and the team ends up working from several slightly different versions.
Nobody owns the procedure
When a document belongs to “the team”, reporting a problem with it belongs to nobody too. Questions from new staff — the cheapest review signal there is — go unrecorded, and the same step confuses the next person in exactly the same way.
Review means re-reading everything
If the only review method is reading a long document top to bottom, reviews get postponed. Most procedures do not need a full rewrite — they need the few changed steps found, fixed, and re-issued, with the rest confirmed as still accurate.
7 events that should trigger a review the same week
Treat each of these as a defect report on the document itself. Not every trigger means a big edit — sometimes the review is five minutes and the conclusion is “still accurate, date confirmed”. What matters is that the check happens and is recorded in the header.
- A step could not be followed as written — it was unclear, out of order, or impossible with the tools actually at hand.
- A tool, system or supplier changed in a way that touches any step, screen, form or contact named in the procedure.
- A new person asked a question the document should have answered. The question marks the exact place the text failed.
- An error or near miss was traced back to the instructions — including “the instructions were fine, but nobody could find the current version”.
- A rule, form or record the procedure depends on changed — an internal policy, a customer requirement, or a form version. Check what applies to your own situation; this guide does not set those requirements for you.
- The same workaround keeps appearing. If experienced staff routinely skip or reorder a step, either the step is wrong or the work is — one of them needs to change on purpose.
- A linked checklist was regenerated or edited. That is proof the source and the working copy were about to drift; confirm they now match.
The version header that does the heavy lifting
Most review routines fail because nobody can tell, in ten seconds, whether the copy in front of them is current. A short header on every SOP — and the same fields on every checklist generated from it — fixes that.
| Header field | What it prevents |
|---|---|
| Procedure title + ID or short code | Two procedures with similar names being confused, especially in search results and shared links. |
| Named owner (a person, not a department) | Reported problems with nowhere to go. The owner triages, approves edits, and closes the loop. |
| Version number + date this version took effect | “Is this the latest?” arguments. Anyone can compare two copies in seconds. |
| One line: what changed in this version | Re-reading the whole document to find a two-step change — and missing it. |
| Where the single current copy lives | Saved PDFs and chat attachments quietly becoming the working version. |
| Next scheduled review date | Procedures that are never re-checked because no date was ever written down. Choose the interval from how fast that work changes — there is no universal number. |
The 6-step update routine
Capture the problem where the work happens
Give the person who hit the problem a one-line way to report it: the procedure name, the step, and what actually happened. A marked-up checklist is already a good report — keep it instead of throwing it away at the end of the shift.
Triage: one step, a section, or the whole procedure?
The owner sorts reports into three sizes. A wrong name or number is a one-step fix. A changed tool or supplier usually touches one section. A procedure nobody recognises any more needs a full re-check — rarer, and worth scheduling properly instead of patching.
Edit the source SOP first
Make the change in the source document, bump the version, and write the one-line “what changed” note in the header. Never start by editing a checklist, a printout, or a copy — that is how the fix gets lost at the next revision.
Regenerate the working checklist
Rebuild the checklist from the changed source so the steps people actually follow carry the new version and date. If you converted the SOP with a tool, this is the fast step — converting a clean source again is quicker and safer than hand-editing two documents into agreement. Our SOP to checklist guide covers the conversion method itself.
Retire the old version the same day
Remove or clearly archive the previous version, update shared links, and pull back printed copies you know about. Two circulating versions are worse than one imperfect one — people will reliably find and follow the old one.
Announce what changed, in one place
Post one short note in the channel the team actually reads: procedure name, new version, and the specific steps that changed. “The SOP was updated” teaches nothing; “Step 4 now uses the new form, step 6 moved before step 5” lets people adjust without re-reading everything.
A worked example (template, not a case study)
How one trigger flows through the routine. This is an illustrative example written for this guide — not a real customer procedure.
Trigger
A new team member asks why step 3 of a monthly equipment-inspection procedure names a paper logbook when the team switched to a shared spreadsheet two months earlier. Nobody updated the SOP at the switch — and step 5 still says “file the paper form”.
Through the routine
- The question is recorded against the procedure, with the step numbers — it is the review trigger, not a side conversation.
- The owner triages: a section-level fix (steps 3 and 5), not a full rewrite; the rest of the procedure is confirmed unchanged.
- The source SOP is edited: spreadsheet named in step 3, filing step rewritten, version bumped, header line added — “v4: paper logbook replaced by shared spreadsheet (steps 3, 5)”.
- The inspection checklist is regenerated from v4 and now shows v4 and the new date in its own header.
- The v3 checklist link is archived; printed v3 copies in the inspection folder are replaced the same day.
- One announcement names the two changed steps. The next newcomer gets a document that matches the work.
The total edit was two steps. What made it work was not the size of the change — it was that the question had somewhere to go, and the source, the checklist, and the old version were all handled in the same pass.
Choosing a scheduled review rhythm
Event triggers catch most staleness. A scheduled pass is the backstop for procedures that change quietly. These are starting points to adapt — not standards, and not a substitute for any review requirement that applies to your own regulated work.
| Procedure type | A reasonable starting rhythm |
|---|---|
| Changes often (tools, systems or suppliers move frequently) | A light header-and-changed-steps check every few months, plus event triggers. |
| Stable, but an error would be costly | Event triggers, plus a full read-through on a set date you choose — and after any related incident, without waiting for that date. |
| Stable and low-risk | Event triggers, plus a periodic confirmation read: does this still describe the work, yes or no, recorded in the header. |
Whichever rhythm you pick, write the next review date in the header itself. A date that lives only in someone’s calendar or memory is not part of the document — and it leaves with that person.
The upkeep checklist (pin this next to your SOP library)
- Every procedure has one named owner — a person, not a department
- Every header shows version, effective date, what changed, and the next review date
- There is one current copy in one known place, and people link to it instead of sending files
- Anyone doing the work can report a problem step in one line, where the work happens
- The 7 triggers are known: failed step, tool/supplier change, newcomer question, error or near miss, changed rule or form, repeated workaround, regenerated checklist
- Changes are edited in the source SOP first — never only in a checklist or printout
- The working checklist is regenerated from the new source, with the new version in its header
- The old version is retired the same day: links updated, known printouts replaced
- Each change is announced once, naming the specific steps that changed
Make the regeneration step the easy step
When an SOP changes, OniLens turns the updated source back into an ordered, checkable list in your browser — your documents never leave your device — so the checklist never lags a version behind. Already have a checklist? Score it free with the Checklist Quality Checker. Prefer not to do conversions yourself? Assisted Conversion turns up to 10 procedures into finished browser and PDF checklists for $29 one-time, and Team & Operations covers larger sets with branding and shared links for $49 one-time.
Free app: no account, no credit card, up to 50 saved checklists. Pro license $39/year removes the limits; Family is $79/year. 30-day refund policy on paid plans.
Frequently asked questions
How often should SOPs be reviewed?
There is no single interval that fits every procedure. Use event triggers that force a review as soon as something changes, plus a light scheduled look at the procedures that change fastest or carry the most risk. Pick the interval from how fast that work changes, and write the next review date in the document header.
What should an SOP version header contain?
Title, one named owner, version number, the date this version took effect, a one-line summary of what changed, and where the single current copy lives. If a checklist is generated from the SOP, name that link too, so anyone can check the checklist was built from the current version.
Should I edit the checklist or the SOP when something changes?
Edit the SOP first — it is the source of truth. Then regenerate the checklist from the changed source and retire the old checklist version the same day. Editing only the checklist guarantees the two drift apart.
What triggers an immediate SOP review?
A step that could not be followed as written, a tool, system or supplier change, a new person’s question the document should have answered, a near miss or error traced to the instructions, and any change to a rule, form or system the procedure depends on.
How do I stop old versions from circulating?
Keep one current copy in one known place and link to it instead of sending files. When a new version takes effect, archive the old one, update every checklist generated from it, and announce what specifically changed in the one channel your team reads.
Who should own SOP updates?
One named person per procedure. The owner triages reported problems, approves the edited source, and makes sure the checklist and the old version are handled. A procedure with no named owner goes stale by default.