Restaurant inventory: when a spreadsheet stops being enough

A spreadsheet is a genuinely good stock system for a small, stable site run by the person who built it. Anyone telling you otherwise is selling something. But there are four specific points where it stops working, and it is worth knowing which one you are at.

1. When prices change faster than you can re-cost

One ingredient can appear in thirty recipes. When its price moves, a spreadsheet needs every one of those recipes revisited unless it was built with proper lookups — and most are not. So the re-cost does not happen, and every margin figure you look at afterwards is quietly wrong.

2. When more than one person needs it

The moment a second person edits, you get versions. `stock_FINAL_v3_KITCHEN.xlsx` is a real file on somebody's desktop right now. The cost is not the confusion; it is that nobody trusts the number, so nobody acts on it.

3. When you open a second site

Two spreadsheets do not consolidate. Transfers between sites go unrecorded, each site drifts to its own ingredient names, and comparing site performance becomes a manual reconciliation job every month.

4. When the person who built it leaves

This is the one that actually bites. Every spreadsheet encodes its author's assumptions in formulas nobody else has read. When they go, the file becomes a black box that everybody is slightly afraid of, and within three months it is abandoned.

What you should expect in exchange

Moving off a spreadsheet is only worth it if you get things a spreadsheet genuinely cannot do:

• Change a price once and see every affected dish re-cost itself.
• Get invoices in without typing them.
• Count stock on a phone, on the floor, with two people.
• See theoretical against actual usage in cash, per site, per week.
• Let three people use it at once without producing three versions of the truth.

The honest catch

Any stock system is only as good as its setup. Ingredients with correct pack sizes and yields, recipes that match what the kitchen actually does — that groundwork is a day or two of real work, and skipping it produces confident, precise, wrong numbers. That setup is the part we charge for, if you want a person who has run kitchens to do it with you. The software itself is free, with no contract and no lock-in.