Documenting the first site before opening the second

The test of readiness to expand is not financial. It is this: could someone who did not build the first site run the second one from the documents you have today?

In most single-site businesses the honest answer is no, and that gap — not the lease, not the fit-out budget — is the real first project.

## What is actually undocumented

Ask an operator whether they have recipes and most will say yes. Ask whether those recipes include yields, sub-recipes, portion weights, plating specifications and pack sizes, and the answer changes.

The same pattern repeats across the operation.

**Recipes and specifications.** Usually partially written, rarely complete, frequently out of date relative to what the kitchen actually does.

**Purchasing.** Held in one person's relationships and one person's memory of what a reasonable price is.

**Opening and closing routines.** Learned by watching, so they vary by who is on shift.

**Training.** Delivered by demonstration, which means each new hire receives a slightly degraded copy of the original.

**The judgement calls.** When to comp, when to remake, when to stop taking orders, how much to prep on a rainy Tuesday. This is the most valuable and least documented category of all.

## Why demonstration does not scale

Training by demonstration works at one site because the person who set the standard is present. Each transfer loses a little — a step abbreviated, a reason forgotten, a tolerance widened.

At one site this is invisible, because the founder corrects the drift daily. At two sites, the correction only happens where the founder is standing, and the second site diverges quietly until something visible breaks.

## What to write, and in what order

Do not attempt to document everything. Write the things that a competent stranger would need in order to produce your product without asking.

1. **The top twenty items** by volume: full recipe, yield, portion, plating, and the common failure to avoid.
2. **The opening and closing routine**, as a checklist rather than prose.
3. **Purchasing specifications**: what is bought, from whom, at what specification, and what a reasonable price looks like.
4. **The service standards** that are non-negotiable, distinguished from the ones that can flex locally.
5. **The judgement rules**, written as if-then statements. These feel awkward to write and are the most valuable thing on the list.
6. **The reporting routine**: what is counted, how often, by whom, and who reads it.

## The test

Documentation is finished when a capable person who was not there can follow it. The only way to know is to try: hand the documents to someone new and watch what they ask. Every question is a gap.

Do this before the second site, not during it. During an opening, every gap is filled by phoning the founder, and the habit that forms is exactly the one expansion was supposed to break.

## What should deliberately differ

Consistency is not uniformity. Some things should be identical across sites — core recipes, brand expression, service standards, reporting. Others should adapt: menu length, opening hours, local additions, staffing structure, and price where the catchment genuinely differs.

The problem is not variation. It is undecided variation — differences that appeared because nobody specified, rather than because someone chose.

## Practical recommendations

1. Start with the top twenty items and the judgement rules. They carry the most value per page written.
2. Write checklists, not manuals. A document nobody opens is not documentation.
3. Test the documents on someone new before you need them.
4. Decide explicitly what is fixed and what may vary by site, and write both lists.
5. Give documentation a deadline and an owner, or it will remain permanently almost-finished.

## In short

Expansion multiplies systems, and a business with no written system multiplies its dependence on individuals instead. Writing the first site down is unglamorous, slow, and the single highest-return preparation for a second one.