Chapter 11 — Key Takeaways

The core claims

1. Food cost is what you used, not what you bought. Cost of goods sold is beginning inventory plus purchases minus ending inventory. Without a count you do not have the numerator, and no software supplies one.

2. The invoice shortcut is wrong by exactly the change in inventory divided by sales. In a single period that error can be several points and can point in the opposite direction from the truth. Over a long window it shrinks toward zero, which is why an operator can be wrong for eleven months and have the annual figure look about right.

3. A cost card is a purchasing document, not a profitability document. The Hearth Chicken costs \$8.52 in product and consumes \$9.94 once you add butchery labor and hearth fuel — 34.3% of its price rather than 29.4%. The card is not wrong; it answers a narrower question than an owner has. The wider question is prime cost.

4. Food cost percentage lies in four structural ways: the denominator moves with price; it is blind to mix; it is blind to labor; and it is usually mis-measured.

5. AP is not EP. Dividing by a yield percentage raises cost, because you paid for the whole thing and can only sell part of it. A 75% yield means the ingredient costs 133% of the invoice price.

6. Your supplier's yield is not your yield. Different conditions, different spec, different use of by-products, and a product that varies by season and lot. Test your own, three times, and cost from the conservative end.

7. Target-cost pricing is a smoke detector, not a thermostat. Obeying it on a cheap-to-make item gives away real money; obeying it on an expensive one prices you out of your market. You need the blend at target, not every item.

8. Portion control is a physical-systems problem. Three invisible drifts — a heavier bird, a fuller spoon of roots, a free-poured sauce — took the Hearth Chicken from 29.4% to 32.8% in six weeks, \$4,892 a year on one item, with no report in the building showing it.

9. Ideal versus actual is the report that tells the truth. It does not say what happened; it says how much happened, which buys you the right to go looking — in order, with theft last.

10. The Q factor is real money. Bread, butter, condiments, and coffee accompaniments at \$0.48 a cover is 1.55 points of food cost, and almost nobody costs it.

The formulas

$$\text{Food cost \%} = \frac{\text{beginning inventory} + \text{purchases} - \text{ending inventory}}{\text{food sales}}$$

$$\text{shortcut \%} - \text{true \%} = \frac{\text{ending inventory} - \text{beginning inventory}}{\text{food sales}}$$

$$\text{Plate cost} = \left(\sum \text{costed components}\right) \times (1 + \text{waste allowance})$$

$$\text{Yield \%} = \frac{\text{EP weight}}{\text{AP weight}} \qquad\qquad \text{EP cost} = \frac{\text{AP cost}}{\text{Yield \%}}$$

$$\text{Menu price} = \frac{\text{plate cost}}{\text{target food cost \%}} \qquad\qquad \text{Allowable plate cost} = \text{menu price} \times \text{target food cost \%}$$

$$\text{Ideal food cost \$} = \sum (\text{units sold} \times \text{plate cost}) \qquad\qquad \text{Variance} = \text{actual usage} - \text{ideal usage}$$

The anchor card — memorize this one

Menu price \$29.00** → food cost **29.4%** → contribution margin **\$20.48.

The benchmarks (rules of thumb — ranges, not laws)

Key terms

food cost percentage · plate cost · recipe cost card · as-purchased (AP) cost · edible-portion (EP) cost · yield percentage · yield test · waste allowance · portion control · target-cost pricing · ideal vs. actual food cost · Q factor

What you should be able to do Monday morning

Take one dish off your menu, build its cost card from the batch recipes and the actual invoices, weigh three plated portions against the spec, and state the item's plate cost, food cost percentage, and contribution margin to the cent. Then count your walk-in, compute food cost from the count rather than the invoices, and compare it with the ideal your cards predict.

If the two numbers are more than a point apart, you have found something — and you have found it in a week rather than in eleven months.