The problem: A restaurant group knows what it bought and sold but not which batches were prepared and thrown away.
The solution: Compare planned prep, actual sales, and leftovers by shift so each kitchen starts with a better quantity.
The math
At four kitchens, $250 of avoidable prep waste per kitchen each week would amount to roughly $52,000 a year in ingredient cost to investigate.
The kitchen lead preps for a busy Friday based on last Friday. Then a nearby event ends early, rain changes the patio traffic, and half a batch remains at closing. The point-of-sale system shows sales, and the purchasing system shows ingredients bought. Neither shows the batch that never became a plate.
That missing middle makes it hard to answer the operational question: how much should this kitchen prepare for this shift?
Measure the batch, not just the purchase
For a few high-variance items, record planned quantity, prepared quantity, sold quantity, and leftover or discarded quantity. Keep the daypart and location. Some ingredients can be used safely in another service under established procedures; others cannot. Let the kitchen lead make that call and record the outcome rather than assuming every leftover is waste.
The result should be a prep range with reasons. A forecast that gives one precise number and hides uncertainty will be ignored by the people who actually run the line.
A look at a restaurant group
Consider an $8 million restaurant group with four kitchens. Suppose each kitchen discards about $250 a week in ingredients from over-prepared batches that could have been reduced without hurting service. Across four kitchens and 52 weeks, that is about $52,000 of ingredient cost. It is a hypothetical ceiling to check against prep sheets, not a claim that the group can recover every dollar.
The first test might show the loss is concentrated in two items on certain weekdays. That is more useful than a general instruction to order less. The lead can change those prep quantities, then watch sellouts and guest complaints alongside waste.
Forecast with room for judgment
AI can combine sales history, reservations, and past prep records into a suggested range. It can also point out when a manager consistently overrides the suggestion for a good reason. The final decision belongs to the kitchen, especially when quality and safe handling are at stake.
The four-step check, in your business
- Pick three items. Start with products that have meaningful ingredient cost and variable demand.
- Record each shift. Capture prepared, sold, carried safely, and discarded amounts.
- Compare like days. Use the same location and daypart, noting events and unusual traffic.
- Change one prep rule. Test a smaller range and watch both waste and sellouts.
The best prep plan is not the one with the smallest batch. It is the one that serves demand without routinely paying for food nobody eats.