Keeping standards consistent across locations

Why do standards drift as you add locations?

Standards drift because they were never written down, only demonstrated. What unit one actually runs on is a set of habits people copied from watching you.

Unit two copies unit one imperfectly. Unit four copies unit two. Nobody makes a decision to change anything, and the fourth version is still recognisable but measurably different.

Drift also compounds fastest where the standard is a judgement call rather than a step. "Clean" and "friendly" travel badly. "Wipe the rail every thirty minutes" travels intact.

How do you onboard the same way at every unit?

Fix the first shift before you fix the manual. Nobody reads the handbook, and the first ninety minutes set what a new hire thinks the standard is.

Write that ninety minutes as a script: who greets them, what they are shown, in what order, and what they do unsupervised first. It is short enough that every manager will actually follow it.

Then check the same thing everywhere. If you only ever audit unit one because it is nearest, unit one is the only unit you know anything about.

What is worth documenting, and what is not?

Document what is expensive to get wrong and easy to get wrong: food safety, cash handling, opening and closing, anything with a legal consequence.

Do not document what changes weekly or what a competent person can work out. A manual nobody trusts is worse than no manual, because it teaches people that the written standard is optional.

The usable test is retrieval. If someone cannot find the answer in under a minute during a shift, the document exists for you, not for them.

How do you know a unit has stopped following the process?

You usually find out through a customer. Reviews mentioning wait times, order accuracy or cleanliness are process signals arriving through the slowest possible channel.

The earlier signals are internal and boring: checklist completion rates, how long tasks take, which unit always needs an exception. A unit that stops logging is telling you something before its ratings do.

Visit frequency is the honest constraint. Most multi-unit problems are not discovered by systems, they are discovered by someone walking in, and that is the thing that gets squeezed as you grow.