Most operators use the phrase "SOP audit" to describe two different things. Sometimes they mean checking the document itself, whether the SOP is correct and current. Sometimes they mean watching to see if people follow the SOP on shift. Those are two different disciplines, and confusing them is why most SOP programs stall.
The document check is an audit. The behavior check is an adherence review. Both matter. Most operators do one seriously and treat the other as an afterthought. When adherence is low, they rewrite the SOP. When the outcome is bad, they retrain the people. Half the time they are fixing the wrong thing.
The two-by-two
Here is the frame I use. Two questions, four quadrants. Every SOP-outcome combination lives in one of them.
Fig. 1 · Four quadrants, four different playbooks.
The mistake most operators make is to treat every SOP problem as a document problem. If the outcome is bad, they rewrite the SOP. That works in the top-left quadrant, and only in the top-left quadrant. In the other three you rewrite the SOP into oblivion and the outcome does not move.
The audit: check the document
An SOP audit answers one question: is this document still correct?
What "correct" means in practice: does the SOP describe the way the work should be done today, given current menu, current tech stack, current staffing model, current regulations, and current brand standards? An SOP that was correct last year is not automatically correct this year.
What the audit checks
- Currency. When was the last update? Have there been menu changes, POS changes, or SOP-referenced tools that have changed?
- Accuracy. Does the SOP describe the correct current procedure? Is the step count right? Are the tools named correctly?
- Regulatory alignment. Does the SOP still comply with current health code, franchisor standards, and labor law?
- Decision rules. Are the decision points still correct? Have the authority thresholds changed?
- Owner. Is the named owner still in the role? Is the escalation number still valid?
The audit rhythm
Twice a year per SOP. Plus every time a menu change, POS change, incident, or regulatory update makes an SOP potentially stale. The named owner runs the audit and either updates, archives, or leaves the SOP alone. Fifteen minutes per SOP for the routine audit.
Audits are not fun. They are also not optional. An SOP library without an audit rhythm becomes a museum of things that used to be true.
The adherence check: watch the behavior
An adherence check answers a different question: are the people doing the work following the SOP?
Notice this question does not care whether the SOP is right. It only cares whether the people are doing what the SOP says. The audit and the adherence check together give you the full picture. Neither alone does.
How to check adherence
- Direct observation. Watch the work happen on shift. Have the SOP in hand. Note where the practice matches and where it drifts.
- Secret shopper. For guest-facing procedures. Third party goes in, follows a scripted scenario, reports back against SOP criteria.
- Camera review. For back-of-house procedures with camera coverage. Watch the recording, check against SOP.
- Checklist verification. For procedures with a checklist artifact (opening checklist, closing checklist). Random pull of the last 30 days, check for completeness and consistency.
- Peer audit. A general manager from another unit visits and does an adherence walk. Fresh eyes catch what home eyes miss.
The adherence rhythm
Continuous, in small doses. Every shift lead spot-checks two SOPs per shift. Every general manager runs one deep-observation adherence check per week, on a different SOP. The area director samples across units monthly. The regional director runs a quarterly cross-unit adherence review to spot drift patterns.
Continuous light checks beat quarterly deep audits. A quarterly deep audit tells you where you were three months ago. A weekly light check tells you where you are now.
Continuous light adherence checks catch drift as it starts. Quarterly deep audits catch drift as it has already spread. Prefer the first.
When adherence is low
You did the observation. Adherence on this SOP is 60 percent, target is 95 percent. Before you rewrite the SOP, ask which of three things you are actually looking at.
Impractical SOP
The SOP is technically correct but operationally impossible under real shift conditions. Example: an opening procedure that assumes 45 minutes of prep time when the schedule only gives 30. The staff created a workaround because they had to. Fix the SOP or fix the schedule. Do not blame the staff.
Weak training
Staff want to follow the SOP but were never properly trained on it. Common with SOPs written after the original training class. Or SOPs that changed and the change was not pushed through training. Retrain, then re-observe. If adherence jumps, this was the problem.
Tolerated non-adherence
Staff know the SOP, could follow it, and do not, because the general manager tolerates the workaround. This is the hardest one to diagnose because it looks like the other two from the outside. The tell: adherence is fine when the general manager is present and drops when they leave. That is a leadership issue, not a documentation issue.
The right question in the coaching moment is not "why aren't you following the SOP." The right question is "what would need to change for this SOP to be easy to follow every shift." The answer tells you which of the three you are looking at.
When adherence is high but the outcome is bad
Rarer, more dangerous. The people are following the SOP faithfully and the outcome is still wrong. Guest satisfaction is dropping. Food cost is drifting. An audit finding keeps recurring.
The SOP is wrong. Either it always was wrong and nobody caught it, or something changed and the SOP is now stale. Rewrite. Test with the outcome. Do not retrain on the old SOP.
This quadrant is where the audit rhythm earns its keep. If you had audited the SOP in the last six months, you probably would have caught the miss. If you had not, you are catching it via outcome, which is expensive.
What I got wrong
Two lessons.
I ran adherence checks as gotchas
Early in my career I ran adherence checks with a compliance mindset. Walk into the unit, catch the miss, write it up, move on. Predictably, general managers started performing adherence when they knew I was on site. Adherence during my visits was 95 percent. Adherence the rest of the year was probably 60 percent. Now I run adherence as coaching, not compliance, and I share the results with the general manager as data instead of as evidence.
I confused audit and adherence in the same conversation
I would walk into a unit, see a procedural drift, and rewrite the SOP that afternoon. Half the time I rewrote a perfectly good SOP because the real problem was that the general manager was tolerating the workaround. Now I diagnose first, quadrant second, action third. Slower. Fewer bad rewrites.
The point
The document and the behavior are two different things. Audit checks the document. Adherence checks the behavior. Both need their own rhythm and their own owner. Do the audit twice a year per SOP. Do the adherence continuously in small doses.
When something goes wrong, put it in the two-by-two before you decide what to fix. Half the time the answer is not what it looked like from the outside. That single discipline will save you from rewriting SOPs into oblivion and blaming staff for problems they did not create.