Optimize the management of yourrecipe knowledge

The recipe card is the reference everything else depends on

A recipe card is not a recipe. A recipe says how to make something; a card says what it costs, what it contains, what it yields and what has to be declared. It is the only document in a food business that food cost, stock issue, regulatory labelling and the production plan all depend on at once.

That is also why an approximate reference is paid for everywhere at once. A wrong weight distorts the dish cost, therefore the margin on every sale, therefore the stock issued, therefore the variance, therefore the purchasing requirement. One error, five wrong indicators.

What a complete card carries

• Ingredients and their weights — gross and net, with the yield linking the two. This is the distinction spreadsheets forget most often, and the one that shifts costs by ten to thirty percent on heavily-trimmed products.
• A calculated cost price — derived from current purchase prices, not typed in. A copied cost is out of date at the first price revision.
• Allergens and nutritional values — derived from ingredients rather than declared dish by dish. That is what guarantees a supplier change on one ingredient updates every card using it.
• The method and its steps — what the cook actually reads. A card the kitchen does not use is a management document, not a production reference.

Gross and net weight: where cost errors come from

Yield is the most structuring and most neglected parameter. A kilo of vegetables bought does not give a kilo usable: peeling, trimming, cooking loss and chilling loss all come off. Depending on the product, yield ranges from 95 % to under 50 %.

Costing on net weight without accounting for yield systematically understates the cost of the most worked dishes — often the signature dishes, the ones you believe earn the most. The profitability ranking of the menu ends up inverted.

A method that survives

• Measure once, on a real production run — weigh in and out at each step. A measured yield beats a coefficient from a manual, because it accounts for the product, the equipment and the team's technique.
• Attach the yield to the ingredient — not to the recipe. It then applies everywhere the ingredient appears, including in cards created later.
• Revisit it when the product changes — a different grade or a different supplier moves the yield. That is the moment to re-measure, not a year later.

Nested cards: intermediate preparations

A sauce, a stock, a dough or a stuffing are cards in their own right, used by other cards. That nesting is what makes the reference maintainable: when the price of butter moves it flows into the sauce, then into the six dishes using it, without a single card being reopened.

Without nesting, the same preparation is copied into every dish. The reference is correct on day one and diverges afterwards — the normal failure mode of spreadsheet recipe cards, and a silent one.

What nesting makes possible

• Automatic cost pass-through — a price change propagates along the whole chain, down to the finished dish.
• Separate production — the intermediate preparation is planned and produced for itself, ahead of the dish, through the production plan.
• End-to-end traceability — the stock's batch shows up in the dish's batch, the chain required by food-safety traceability.
• Consistency across sites — the same preparation, produced in several locations, keeps one definition. See multi-site management.

Keeping the reference alive without a full-time job

A reference decays by default. Recipes evolve in the kitchen without the card following, suppliers change, portions drift. The question is not how to build a perfect reference, it is what keeps it current without tying up a person full time.

• Let prices update themselves — that is half the maintenance work, and it disappears as soon as cards are linked to supplier references instead of typed-in prices.
• Revise on signal, not on schedule — a repeated gap between planned and actual production, or a food cost ratio moving with no price increase behind it, points at the card to reopen. More effective than an annual review nobody finishes.
• Let the people producing correct them — a card edited by the section chef using it stays accurate. A card maintained remotely by finance drifts from kitchen reality within months.

Frequently asked questions about recipe cards

How many cards do we need to start?

The ones carrying the volume. Twenty to thirty cards usually cover most of revenue and are enough to produce a credible food cost. Aiming for completeness before you have a first figure is the surest way never to finish.

Are allergens really derived automatically?

Yes, from the ingredients and their own compositions. The declaration remains the business's responsibility, but it starts from a consistent base instead of being re-entered dish by dish — and it updates when an ingredient changes.

Can an existing reference be imported?

Yes, by file import. That is the usual path when cards already exist in a spreadsheet: the import recovers the structure, and the remaining work is on yields and on linking to supplier references.

What happens when a purchase price changes?

The cost of every affected card is recalculated, including through intermediate preparations. That is the main point of the reference: the update requires no action.

Do we need a card for every variation of a dish?

For variants that change cost or allergens, yes — a gluten-free version has neither the same cost nor the same declaration. For a plating variation, no: multiplying cards with no substantive difference makes the reference heavier without making it more accurate.

How are seasonal recipes handled?

They stay in the reference between seasons, with costs still tracking supplier prices. When the season returns the card is already current and the selling price can be set on a real cost rather than last year's.

Is the reference useful beyond costing?

It also carries regulatory labelling, allergen declarations, the production plan and batch traceability. That is what justifies the initial effort: data entered once feeds four separate obligations instead of being re-entered into four tools that drift apart.

How do we know a card is wrong?

Three signals point at it: a recurring stock variance on its ingredients, a systematic gap between planned and produced quantity, and a food cost ratio drifting from its neighbours with no menu reason. None requires an audit — they surface on their own once sales, production and stock share the same reference.