Guides

Ingredient forecasting for meal-prep kitchens

Purchasing is where a meal-plan kitchen’s margin is actually decided, and it is downstream of information the business already has. Most over-buying is not a forecasting failure but a plumbing failure.

The forecast is arithmetic, not prediction

Unlike a restaurant, you are not guessing at demand. Once the cut-off passes, you know exactly how many portions of each recipe you owe. Converting that into ingredients is multiplication against a recipe, then summing.

When kitchens describe purchasing as guesswork, what they usually mean is that the arithmetic is being done by hand from a sheet that is already out of date — which is a different problem with a different fix.

The calculation, in order

  1. Take the frozen selections

    Only for days inside the cut-off, because anything later can still change.

  2. Multiply by the recipe

    Each recipe’s bill of materials times its portion count gives raw quantities.

  3. Normalise the units

    Convert everything for a given ingredient into one canonical unit, or the sum is meaningless.

  4. Deduct what you hold

    Subtract current stock to get the true requirement rather than the gross one.

  5. Flag what you cannot cover

    Where the gap cannot be closed within lead time, raise it explicitly instead of rounding it away.

One source, or the numbers drift

The rule that makes this reliable is that production counts and the ingredient forecast must come from the same selections. Maintain a purchasing sheet alongside a cook list and the two will disagree the first time somebody changes a plan mid-week.

That disagreement is expensive in both directions: buy to the higher number and you waste food, buy to the lower one and you are short on a Tuesday morning.

Common questions

Why do ingredient quantities need canonical units?
Because recipes get written in whatever unit the author was thinking in — grams, kilograms, pieces, bunches. Summing across recipes only works if every quantity for an ingredient is expressed in one unit.
Should the forecast account for waste and yield?
Yes, ideally as a factor on the recipe rather than a mental adjustment at ordering time. A yield assumption held in someone’s head cannot be reviewed, corrected, or handed to a colleague.
What if a supplier lead time is longer than the cut-off window?
Then that ingredient has to be bought against an estimate, and it should be visibly marked as such. The alternative — treating an estimate as a firm number — is how kitchens end up with confident purchase orders for meals nobody ordered.

See how this works in practice

The home page walks through the same mechanics against a real admin loaded with a demonstration kitchen: production, packing, routes and the customer side.