Organising your restaurant's service, from order to kitchen

Before changing anything, it helps to know where the errors happen. In a dine-in restaurant an order passes through four points, and each one is a chance to get it wrong.

Note that the order is handed over three times between being spoken and being cooked. The fix isn't handing it over more carefully — it's handing it over once: written where the guest says it, arriving in the kitchen exactly as written.

• The guest says the order. The waiter may mishear it in a full room, and the guest may leave out something important — an allergy, a modification.
• The waiter writes it in a notepad. Hurried handwriting, abbreviations only they understand, and sometimes nothing at all: they rely on memory.
• The order moves to the kitchen. Spoken through a pass, on a slip, or called across a noisy room.
• The kitchen interprets and prepares it. If the interpretation is wrong, the mistake surfaces only at the table, after the time and ingredients are spent.

The common mistake is letting everyone see everything. It looks like flexibility, but it produces scatter: who is following the new order if everyone can see it? And who knows what is genuinely ready?

The principle is that each role sees what it needs to work, and not what doesn't concern it.

Separating roles also solves a practical problem: teams change, and much of the team is seasonal. A limited-permission account is handed to a new employee in a minute, and withdrawn in a minute.

• The kitchen sees incoming orders and what to prepare; it has no need to see revenue or bill prices.
• The counter sees bills, payment, and orders ready to serve.
• The owner sees everything, plus the numbers: what sells, what doesn't, and which hour is tightest.
• Each order is attributed to whoever started preparing it. Not for surveillance, but because "who has this?" needs an immediate answer mid-service.

The most repeated question in service is "where is table so-and-so's order?" — and it's a question that interrupts someone else's work to answer.

The fix is for every order to carry a known status, visible to everyone at the same time, with nobody having to ask:

Those four statuses replace most of the verbal questions in service. The bigger benefit is that they make delay visible before the guest complains: an order sitting at "preparing" for twenty minutes shows on screen, without waiting for someone to notice.

• Received: the order is in, preparation hasn't started.
• Preparing: the kitchen has started, and it's known who started.
• Ready: preparation is finished, waiting to be carried to the table.
• Served: it reached the guest.

Every restaurant knows it gets tight at some point, but few know when precisely. And the gap between "evenings are busy" and "the real peak is 20:30 to 21:15" is the gap between a hunch and a decision.

Once the hour is known precisely, three decisions become possible: how many people to roster on that specific hour, what must be prepped just before it, and which slow dishes are better not pushed at the height of it.

The measurement itself asks nothing extra of the team: if orders are already recorded as they arrive, the numbers exist — they just need reading.

The biggest risk in organising service isn't the tool, it's the timing. Changing the whole working method on a busy night produces chaos that gets blamed on the new system — and then everyone goes back to the notepad.

The concrete sign of success isn't that the system works — it's that nobody asks for the notepad any more.

• Start with two or three tables, on a quiet service and not at the weekend.
• Keep the old method running alongside for the first week. The new system proves itself; it isn't imposed.
• Train one person well rather than explaining quickly to everyone; they're the one who will explain it to the rest of the team, in their language.
• After a week, ask the team two questions: what got easier, and what got harder. The second answer is the more important one.

The order is written where the guest says it — from their phone, at the table — and arrives in the kitchen exactly as written, with no second or third handover. The order board shows the four statuses to everyone at once, and roles carry limited permissions: the kitchen doesn't see the numbers, the counter sees what it needs.

There's also a page showing peak hours to the hour, built on your own orders rather than a general estimate.

What it doesn't do: no staff rostering, no stock management, and no substitute for training. It organises the order's journey and shows the numbers; the human organisation stays your job.

Subscriptions start at 2,100 DA per month on a 12-month plan, and you can test on a single table — exactly as this guide advises.