Guide
A practical sequence for groups where the standards already exist but are spread across different places and people — and where the real breakdown happens when something changes.
Before doing any work, separate the two things usually bundled together as “we need to standardize”:
For many groups past three locations, the second is far more common than the first. That matters, because the fix is different: the first needs decisions from leadership, the second needs distribution and maintenance. Groups that treat a distribution problem as a decision problem end up re-litigating standards that were already settled.
Collect everything, in whatever state it is in — SOPs, recipes, manuals, policies, training decks, checklists, PDFs, shared-drive folders, role guides, manager resources. Include the laminated sheets taped inside cabinets and the notes in a GM’s phone if you can get them.
Collect what already exists
Don’t clean it up yet. Just find it.
The mess is diagnostic. Where there are four versions of one procedure, that tells you something about where change goes to die.
For each item, capture three things: what it covers, who it applies to, and when it was last known to be accurate.
This is the step most projects skip, and it is where the real value is. Working through the inventory, four questions surface repeatedly:
Four diagnostic questions
Which version is current?
Where multiple versions exist, somebody has to decide — and it is often not obvious from the documents themselves.
Is this difference intentional?
When a location does something differently, is that approved variation or drift?
Does the written procedure match what actually happens?
Frequently it does not, and the floor version is sometimes the better one.
Who owns this decision?
Ambiguity here is why changes stall.
Bring these to the right people in batches rather than one at a time. A chef can resolve thirty recipe-standard questions in a focused session; the same thirty spread across six weeks of interruptions will never finish.
Standardization does not mean uniformity, and treating it that way is the fastest route to a system nobody uses.
One restaurant has different equipment. Another has a patio that changes the closing sequence. A local jurisdiction requires something the others do not. These are real. A system that insists every location is identical gets quietly ignored at the locations where it does not fit — and once staff learn that the company material is wrong about their restaurant, they stop trusting all of it.
Company standard
Legitimate local variation
Make the call explicitly for each area: company standard, or approved local variation attached to specific restaurants. Anything that is neither is drift, and should be resolved rather than documented.
Folder structures are built for the person filing, not the person reading. A bartender at your Harbor location should be able to see what applies to them without navigating a hierarchy designed by whoever set up the drive.
| Harbor | Downtown | Beachside | |
|---|---|---|---|
| Cook | Current SOP | Current SOP | Local variation |
| Server | Current guide | Current guide | Current guide |
| Manager | Manager standard | Manager standard | Manager standard |
Scope every piece of material by role and location, and give it a visible version and approval state. The test: can someone tell, without asking anyone, that what they are reading is the current company standard for their job at their restaurant?
This is where standardization most often unravels six months later.
If training content is a separate artifact — a deck built once, a video recorded at launch — it captures the standard as it was on the day it was made. Every subsequent change makes it slightly more wrong, and nobody notices because training is only consumed by people who do not yet know the right answer.
New hires then learn a version that is out of date, get corrected on the floor, and conclude the official material is unreliable. That conclusion is rational and very hard to reverse.
The most reliable test of whether a group is genuinely standardized: change one standard, then check what a new hire is taught the following week.
If training still teaches the old version, you have documentation, not standardization.
Standardization is not a state you reach, it is a state you maintain. Decide, in advance, what happens when leadership changes a standard:
The route for one change
A portion change from 5 oz to 4.5 oz sounds trivial. It typically touches five separate artifacts. If no one owns that propagation, four restaurants will adopt it and one will not.
Managers notice when a procedure stops matching reality. Employees ask questions the company has never formally answered. Without a route, both become text messages, and the information never re-enters the system.
Example
“The procedure says 4 oz, but the new packaging is 5 oz. Which standard do we follow?”
Line cook · Beachside
Even a simple intake — one place where “this procedure is wrong now” and “nobody has told us what to do here” can be logged and stays visible until someone resolves it — converts ongoing friction into maintenance input.
Takeaway
If a full pass isn’t realistic right now, the highest-return sequence is:
That is a week of work and removes a disproportionate share of the risk.
Every step above is achievable as a project. The reason groups end up doing it twice is that maintenance has no owner — the leadership team is lean, and keeping the system current loses to service every single week.
Standardization is not a one-time cleanup project. It is a system for keeping the current standard current.
Be explicit about who owns it afterwards. Read more on what breaks between 3 and 12 locations, or on keeping procedures current once they exist.
See it with your operation in mind
We find where your restaurants differ, bring the decisions to your leaders and keep the approved standard current across every location.
Get a Private Walkthrough →A short walkthrough tailored to your restaurants. Sent within one business day. No call required.