Most kitchens have a rough sense that they waste something, and almost none can say what it is worth. Until it has a dollar value attached to it, waste stays an abstract concern that gets mentioned in meetings and never changes.
A one week audit will give you that number. It is not a technology project and it does not need consultants standing in your kitchen. It needs a set of scales, a printed sheet, and someone senior who will hold the team to it for seven days. Here is how to run one.
Step 1: Decide what you are counting
Keep the scope tight or the exercise collapses by day three. Count three things only: preparation waste, which is trim and offcuts from preparing ingredients; line waste, which is food cooked or plated and then discarded before it leaves the kitchen; and plate returns, which is food that came back from the dining room.
Leave packaging, oil, and customer side waste out of this round. You can add them later.
Step 2: Set up the bins and the sheet
Put three labelled containers where the waste actually happens, not in a corner where people have to walk to reach them. Convenience determines whether this works.
Print a simple sheet with columns for date, service, category, item, weight, and a short reason. Paper on a clipboard beats an app for a one week exercise, because it requires nothing to be learned.
Step 3: Brief the team properly
Say clearly that this is a measurement exercise and not an audit of individuals. If the team believes someone is going to be blamed, the numbers will be wrong and you will have wasted the week. Explain that the purpose is to find out where the kitchen is losing product so that the process can be fixed.
Step 4: Weigh at the point of disposal
Anything going into a waste container gets weighed first and written down, with the item named and one or two words on the reason. Overproduced, wrong order, dropped, out of date, trim, and returned are usually enough categories to cover it.
The reason field is the part people skip and the part that makes the exercise useful. A kilogram of chicken is a number. A kilogram of chicken recorded as overproduced on a Tuesday lunch is a decision you can change.
Step 5: Run it for seven consecutive days
A full week matters because your kitchen behaves differently on a quiet Monday than on a Saturday night. Five days will mislead you. Include every service, and do not skip the day someone is on leave, because that is exactly when process gaps show up.
Step 6: Convert weight into dollars
At the end of the week, price each recorded item using your actual purchase cost per kilogram. Then annualise the weekly figure. This is the moment the exercise becomes real for everyone who took part, and it is worth sharing the total with the team who did the weighing.
Step 7: Sort by cause, not by ingredient
It is tempting to look at the most expensive item wasted and start there. The more useful view groups the waste by reason. Overproduction points to prep par levels or forecasting. Trim points to specification, supplier format, or knife technique. Plate returns point to portion size or a dish that is not landing. Out of date points to ordering frequency and stock rotation.
Each cause has a different fix, and only one of them is about the ingredient itself.
Step 8: Change three things, then re-count
Pick the three largest causes and address only those. Adjust a prep par level, change a portion size, renegotiate a delivery schedule. Trying to fix everything at once means you will not know what worked.
Repeat the count for one week, a month later. If the number has moved, the team will believe in the exercise and you can run it quarterly with much less effort.
Salt and Ledger runs operations and kitchen workflow reviews built on this kind of observation: covers per hour, labour per cover, and waste and prep mapped through a full service, ending in a change list with costs, owners, and dates attached to each item.
If your kitchen feels busy but the margin is not showing up, write to info@saltandledger.com.sg with your weekly covers and current food cost percentage. We will tell you what we would observe first.