Multi-Location Restaurant Inventory, Without the Spreadsheets

Running one restaurant on a spreadsheet inventory is manageable. The chef knows the walk-in, the counts happen on Sunday night, and the numbers roughly match what the owner sees on the invoices. Add a second location and the same spreadsheet starts to drift. Add a third and it usually stops telling you anything useful at all.

This is not a discipline problem. It is a structural one. Multi-location restaurant inventory management has to answer questions a single spreadsheet was never built for: which kitchen used the product, which one paid for it, and whether the difference between what you sold and what you used is waste, theft, a bad recipe card, or a case of chicken that moved across town and was never written down.

Why food cost control got less forgiving

Operators have less room for error than they did a few years ago. The National Restaurant Association reported in March that average food costs are 34% above pre-pandemic levels, and 82% of operators said their costs were higher than the year before. In its 2026 State of the Restaurant Industry release, the Association said 42% of operators reported their restaurant was not profitable in the prior year.

Menu prices have followed, but not without limit. The Bureau of Labor Statistics' April CPI release showed food away from home up 3.6% over twelve months. When guests are already feeling price increases, finding a point or two of food cost in the kitchen is often more realistic than finding it on the menu.

Waste is part of that picture. In data behind its 2026 food waste report, ReFED estimates that restaurants and foodservice generated 12.5 million tons of surplus food in 2024, including about 1.49 million tons from overproduction. Most of that surplus is plate waste you cannot count in a walk-in, but overproduction and over-ordering are exactly what good inventory practice is meant to catch.

The four pieces of multi-location inventory

1. Counts that mean the same thing everywhere

A count is only comparable across locations if everyone counts the same items, in the same units, at the same point in the day. That sounds obvious until you look at three kitchens' sheets side by side: one counts shredded cheese in pounds, another in bags, a third in "half a lexan." Standardize the item list, the unit of measure, and the count time (for example, after close on the last day of the period, before the morning delivery). Arrange the sheet in shelf order for each location so counting is fast and nothing gets skipped.

2. Par levels set per location, not per company

A par level is the amount you want on hand after a delivery so you can get to the next one without running out. A simple version: average daily usage multiplied by the days between deliveries, plus a safety buffer. If a location uses 30 pounds of chicken thighs a day, receives deliveries every three days, and wants a 20-pound cushion, its par is 110 pounds.

The trap in a group is copying one location's pars to the others. A downtown lunch spot and a suburban dinner spot with the same menu will use products at very different rates. Pars should come from each location's own sales history and be reviewed when the menu or the delivery schedule changes.

3. Transfers recorded on both sides

Transfers are where spreadsheet inventory most often breaks. The flagship runs short on brisket, the commissary sends over two cases, and someone texts the kitchen manager about it. The sending location's usage now looks high, the receiving location's usage looks low, and neither number is true.

A transfer needs three things recorded at the time it happens: the item and quantity, the cost at which it moves, and both the sending and receiving location. If it only lives in one kitchen's spreadsheet, it will distort both locations' food cost for that period.

4. Variance: actual versus theoretical food cost

This is the number that tells you whether the kitchen is running the way the recipes say it should.

• Theoretical food cost is what you should have used, based on what you sold and what each recipe costs.
• Actual food cost is what you did use: beginning inventory, plus purchases, plus transfers in, minus transfers out, minus ending inventory.

The gap between them is variance. Some variance is normal. A gap that is large, growing, or concentrated in a few items is a signal worth chasing.

A worked example (hypothetical)

Picture a three-location group. Here is one week at its second location. All figures are illustrative.

A $1,420 weekly gap at one location adds up to roughly $74,000 over a year if nothing changes. Now look at what happens if the $1,100 transfer out had never been recorded. Actual food cost at this location would show $14,700, or 35.0%, and the location that received the product would look $1,100 better than it really was. The owner would be coaching the wrong kitchen.

This is why the transfer log matters as much as the count itself. It is also why theoretical food cost needs current recipe costs: if the recipe cards still carry last quarter's prices, the "variance" is partly a pricing error, not a kitchen problem.

Where spreadsheets break at two or more locations

A single spreadsheet can hold the math above. What it cannot do well is keep the inputs honest across several kitchens:

• Version drift. Each location saves its own copy. Within a few months the recipe tabs, item lists, and units no longer match.
• Stale costs. Invoice prices change weekly, but recipe costs in the sheet get updated when someone has time. Theoretical food cost quietly goes wrong.
• One-sided transfers. The transfer is logged by the sender or the receiver, rarely both, and never reconciled.
• No link to sales. Someone has to export item-level sales from the POS, paste them in, and match item names by hand. At three locations that is a real part of someone's week.
• Late answers. By the time the numbers are assembled, the period is over and the product is gone.

If this sounds familiar, the pattern is the same one we described in the $47,000 spreadsheet that ran a practice: the sheet works until the business grows around it, and then the cost is hidden in labor and mistakes rather than in a line item.

What to do this week

You do not need new software to improve this. Start here:

1. Pick your top 20 items by spend and confirm they are counted in the same unit at every location.
2. Set one count time for all locations and hold to it for a month.
3. Start a shared transfer log with sender, receiver, item, quantity, and cost. Have the receiving manager confirm each entry.
4. Update recipe costs for those 20 items using the latest invoices.
5. Calculate variance for those items only, per location. Look for the two or three with the largest dollar gap and ask why.

A narrow, accurate variance report beats a complete, unreliable one.

When it is time to move off the spreadsheet

The signal is usually not a single bad week. It is when the person who maintains the sheets becomes the bottleneck, when location managers stop trusting the numbers, or when you cannot say which kitchen is driving a food cost increase. At that point, the answer is a system where counts, purchases, transfers, and POS sales land in one place with one item list, so variance is calculated rather than assembled.

That is the kind of connected system we build for multi-location restaurant groups, and it is worth pricing against the tools and hours you already spend; our software cost calculator is a reasonable place to start. For a broader look at what disconnected tools cost a growing business, see the real cost of running your business on seven tools.

If you would like to talk through how your locations handle counts and transfers today, you can book a conversation whenever it is useful.