The walk-in is full of last week’s over-prep. The line runs out of the most-ordered protein on a busy Friday because the prep call was off. The prep system in most restaurants is one person’s intuition about what to expect that day. When that person is wrong, the cost shows up in waste or in a table that cannot be served.

A prep system replaces intuition with documented pars. The pars are the target quantities for each prep item, calibrated to cover volume and adjusted as covers change.

Most kitchens have an informal version of this already. Whoever runs prep has a mental estimate of what to make. The problem is that the mental estimate is not documented, not reviewed, and not improving over time.

As Your Restaurant Has No SOPs covers, prep is the SOP that shows up most directly in food cost. This article is for the kitchen that has been calling prep by feel and wants a more reliable starting point. The Three-Part Prep Management System is the approach.

Why Intuition-Based Prep Produces Waste and Gaps

The obvious cost is over-prep waste. When the expected covers do not materialize, the kitchen is left with prep that carries poorly or gets discarded. Every walk-in full of forgotten prep represents a previous call that was wrong.

The less visible cost is under-prep gaps. A line that runs out of a key item during service does not recover gracefully. The table that waited for a dish is not coming back with the same enthusiasm.

The deepest cost is inconsistency. When prep varies by whoever made the call that day, the menu item varies too. A dish plated from two different prep quantities has different yield, different texture, different cook time. The guest notices, even if they do not name it.

The Three-Part Prep Management System

Part 1 establishes documented pars for each prep item

A prep par is the target quantity for one prep item for one service type: Friday dinner, Tuesday lunch, a specific known event. The par is based on historical covers for that service type and the portion size for each dish that uses that item.

Building the par starts with the last thirty days of covers by service type and the recipe quantity for each dish. The calculation multiplies average covers by the proportion of guests who order each dish, then by the portion size of the ingredient. Round up by a safety margin of ten to fifteen percent.

The result is a documented par for each prep item by service type. It is not perfect. It is more accurate than a guess, and it compounds in accuracy as covers data accumulates.

Part 2 adjusts the par based on known variables for this service

The documented par is the starting point. The daily call adjusts it based on known variables: a large reservation that orders off the menu, a special that uses an ingredient at higher volume, a forecast of unusually high or low covers.

The adjustment process is a brief review before prep begins. Compare the base par to tonight’s known variables and adjust up or down. The documentation is simple: the par, the adjustment, and the reason.

This review takes five minutes. Over a full week, it replaces the intuition-based call with a structured decision that can be reviewed and improved.

Part 3 tracks waste and feeds it back into the par

Every over-prep item that is discarded at the end of a service represents a specific amount of money. Tracking waste by item over four weeks reveals which pars consistently run over and by how much.

A par that consistently over-produces by twenty percent should be adjusted down by ten to fifteen percent. The waste record is the feedback loop that makes the par more accurate over time. Without tracking, the same par produces the same waste week after week.

How the Conductor Guides Your Prep Call

Before prep starts for Friday dinner, the kitchen lead asks the Conductor what the prep quantities should be for tonight. The Conductor is Kiluma’s context-aware AI.

It draws from the Living Library, the active layer where the restaurant’s documented prep pars, recent covers, and waste records have been saved. It returns the specific prep quantities for tonight based on the documented pars, recent covers, and the most recent waste adjustments. The prep call is no longer a guess.

Kiluma is the knowledge layer, not the inventory system. MarketMan or BlueCart manages the ordering. Kiluma holds what the kitchen has learned about how much to prep.

Review Last Week’s Waste Before the Next Prep Call

Before the next major prep session, review what was discarded from the previous one. Note which items over-produced and by how much.

Adjust the par for any item that over-produced by more than fifteen percent. Write down the new par.

The prep system improves with every review. A kitchen that does not track waste cannot improve the prep call. A kitchen that does can.

A Prep System Starts From What Happened, Not From a Guess

Does your prep call start from what actually happened last week, or from a guess about what might happen tonight? The difference between a kitchen that runs well and one that consistently wastes or runs out is almost always the prep system.

The Conductor returns the prep quantities grounded in this restaurant’s actual documented data. Try Kiluma free for 14 days at kiluma.ai.