An SOP review calendar has one guaranteed property: it is always late. Not sometimes late, structurally late. The document changes on a schedule; the process changes on events. A retrofit in March, a new material lot in May, a workaround invented at 2 a.m. in June that quietly became the method. If your review lands in December, you spent most of the year certified and wrong.
That gap is not an edge case, and it is not cheap. 69% of manufacturing executives report negative project or product impacts from inaccurate, unclear, or outdated process documentation (Canvas GFX survey). The question of how to keep SOPs up to date is really the question of what triggers an update — and almost every plant has the wrong trigger.
How often should SOPs be reviewed in manufacturing?
The standard guidance says annually or biannually, with risk-based cycles stretching one to three years for stable processes, and immediate updates whenever the process changes (Rostone Operations). The guidance is fine. Read it carefully and it admits the problem: the calendar is the backstop, the event trigger is the real mechanism.
Here’s what happens in practice. The event trigger fails, silently, almost everywhere. Because the only person who reliably knows the process changed is the operator running it, and no review workflow asks the operator anything. The retrofit gets a project file. The material change gets a purchasing record. The workaround gets nothing at all — it lives in the hands of the person who invented it and spreads shift to shift by word of mouth. None of those three events touches the document management system, so the document management system keeps versioning fiction on schedule.
Why do SOPs become outdated so fast?
Because drift is not a discipline problem, it’s a physics problem. The floor changes at the speed of production: parts wear, lots vary, machines get modified, people find better sequences. The documentation changes at the speed of meetings. Any system that reconciles the two on a calendar guarantees a standing gap, and the gap compounds — each unrecorded change makes the next reviewer trust the document a little less, until the review itself becomes ceremonial.
Meanwhile the work stays human. Deloitte’s 2026 Manufacturing Industry Outlook estimates 81% of task hours in manufacturing are still performed by people. The procedures describing that human work aren’t optional infrastructure; they’re the only transferable copy of the operation. When they drift, everything trained, audited, and improved against them drifts too — which is exactly how a corrective action ends up correcting back to fiction, as I wrote in Tuesday’s piece on CAPA effectiveness.
What are the signs an SOP no longer matches the process?
If the calendar can’t catch drift, what does? The floor announces it, constantly, in signals most plants log as unrelated problems.
Escalations tick up. Workarounds recur. Technicians deviate from the posted method. The same failure repeats. The process drifts between shifts. Quality varies without an assignable cause. Cycle times wobble. Seven metrics, usually owned by seven different people, each managed as its own problem with its own meeting.
Read together, they are one message: the documented process no longer matches the floor. A multi-site operations leader named Kevin VanSlooten put it precisely in a discussion on my LinkedIn: the floor changes faster than the documentation unless the organization intentionally builds maintenance discipline around the operating system, continuously. The signals are the early-warning system your review calendar was never going to be.
Document-management software doesn’t solve this either. The category is good at versioning, routing, and reminding — it makes the calendar run beautifully. But a faster calendar is still a calendar. It versions the fiction more efficiently.
Run this in your plant this week: the change-vs-revision audit
Thirty minutes, one line, three columns.
| Physical change (last 12 months) | SOP revision date that followed | Gap in days |
|---|---|---|
| Retrofit, tooling swap, material change, adopted workaround… |
List every physical change on the line: retrofits, tooling swaps, material or supplier changes, and any workaround you know became the de facto method. Next to each, the date of the SOP revision that captured it. Then the gap.
Most plants find two things. Half the changes have no revision at all — those rows are open-ended fiction. And the rows that do have revisions show gaps measured in months, not days. Every one of those gaps is a period when training, audits, and corrective actions were aimed at a document describing a machine that no longer existed.
That one-page table is also the honest answer to “how often should we review?” — the floor just told you how often it changes. Build your update trigger there.
Stop updating on a schedule. Start capturing on signal.
Keeping SOPs current isn’t a review problem, it’s a capture problem: someone has to notice the change where it happens and feed it back into the document while it’s fresh. That is what SenseiLab’s Living SOPs are built to do — the 30-day SOP Sprint captures the method that actually runs your line, validated with the operators who run it, and leaves you a system where the floor’s changes reach the document in days, not review cycles.
Diego Echenique is the CEO and co-founder of SenseiLab. He has spent more than 20 years in manufacturing operations across five continents, and now helps regulated manufacturers capture the knowledge that actually runs their floors
