Restaurant Cost of Goods Sold: Control Vendor Changes Without Losing the Trail
A vendor substitution can keep dinner service running and still change a dish's cost. The important question is not just whether the replacement case costs more. It is whether the usable quantity, yield, and portion cost changed.
For a multi-location group, record that change where it happened. Otherwise the next food-cost review may show a variance without enough detail to explain it.
Start with a consistent COGS calculation
A basic inventory calculation is beginning inventory plus purchases minus ending inventory. MarginEdge documents this formula in its controllable P&L guidance. Transfers, credits, and category treatment should follow your accounting setup so location comparisons remain consistent. MarginEdge calculation guide.
Compare the same period and use a sales denominator that matches the category you are measuring. Food-only cost divided by total food-and-beverage sales tells a different story from food cost divided by food sales. Label the measure clearly.
Normalize vendor prices before comparing them
For each important ingredient, track the item, pack size, purchased quantity, usable yield, delivery charge treatment, and effective date. A cheaper case may contain less product. A higher purchase price may come with a different yield.
Use a receiving check: did the delivered quantity and product match the order, and did the invoice match the delivery? Keep disputed charges and expected credits open until the credit is documented. An email requesting a credit is not the same as a posted credit.
Record substitutions where they occur
A lightweight substitution log should name the location, original ingredient, replacement, price and pack-size difference, recipe affected, and approver. Check any recipe or menu consequences before making the substitute routine.
Prioritize ingredients with the largest spend or repeated unexplained variance. A fixed number of ingredients or a universal price-change threshold is less useful than your own materiality rule: which changes are large enough, frequent enough, or sensitive enough to deserve review?
Compare actual and theoretical usage
Actual usage comes from the inventory movement recorded during the period. Theoretical usage estimates what recipes and items sold should have consumed. A gap points to questions, not a diagnosis by itself. The food-cost variance guide gives the next checks for counts, units, timing, and recipe assumptions before assigning a cause.
Check count accuracy, unit conversions, transfers, recipe versions, waste records, and substitutions before concluding that portioning or theft caused the difference. An item voided before preparation has a different food-cost implication from one discarded after cooking.
Give the finding a next step
A useful finding might read: “Location A's substitute has a higher cost per usable portion; purchasing will confirm the agreed price and the kitchen lead will update the recipe record.” That is more actionable than “food cost is high.” This is an illustrative workflow, not a customer result. Record purchasing and kitchen responsibilities in the manager action log; the person requesting a credit should know who checks that it posted.
Work the calculation before explaining the variance
Hypothetical period: opening food inventory $8,000 + food purchases $20,000 − closing food inventory $7,000 = $21,000 food COGS, before any separately required adjustments. Against $70,000 food sales, food COGS percentage is $21,000 ÷ $70,000 × 100 = 30%. These are arithmetic inputs, not recommended targets. Use reconciled transfers and credits and the same period for both sides.
Now compare usable cost. A hypothetical 20-pound case at $80 with 80% usable yield provides 16 usable pounds, costing $5 per usable pound. A 25-pound substitute at $90 with 60% yield provides 15 usable pounds, costing $6 per usable pound. The cheaper purchased pound produces the more expensive usable pound. Confirm actual yield in your kitchen before acting.
| Variance worksheet | Record |
|---|---|
| Period, location, ingredient | Same unit and date range across sources |
| Actual usage | Opening quantity + receipts + net transfers − closing quantity, adjusted consistently |
| Theoretical usage | Items sold × recipe quantity, using the relevant recipe version |
| Difference | Actual minus theoretical, in a common unit |
| Checks before diagnosis | Counts, conversions, waste, substitutions, credits, and transfers |
| Next action | Purchasing or kitchen owner, due date, evidence needed |
Keep a purchasing correction from fading away
Marty is a managed profit-control layer for multi-unit restaurants.
Find → Assign → Follow up → Verify. Find a supported price or usage change. Assign purchasing or the kitchen lead a specific action. Follow up on the credit request, recipe update, or receiving correction. Verify the later invoice, posted credit, or comparable usage record before claiming a result.
The paid Cash Recovery Plan establishes the baseline that tells Marty what to watch first.
Optional Managed Profit Control adds monitoring, alerts, follow-up, escalation, and verification for the purchasing and usage controls selected in the Review.
Sources checked September 15, 2026: linked MarginEdge documentation and Marty's Review offer.