How to Manage Restaurant Inventory Without Spreadsheets

Most restaurant inventory is still managed on a clipboard and a spreadsheet. A manager walks the walk-in on Sunday night with a printed sheet, scribbles counts, takes it to the office, and types them into a Google Sheet that has not had its formulas checked in eighteen months. The variance number at the bottom is treated as gospel even though half the recipes feeding into it have not been updated since the last menu change.

Meanwhile, food cost is the single largest controllable expense in a restaurant. A point of food cost on a $1.5M operation is $15,000/year. Most restaurants running on spreadsheets are leaking 2-4 points to invisible waste, theft, portion drift, and over-ordering they cannot see because their tracking is too coarse to spot it. This guide is the practical version of getting off the spreadsheet -- six steps, real tools, no fluff.

Before the fix, the diagnosis. Three reasons spreadsheets fail in a real kitchen:

They do not connect to sales data. Your spreadsheet tells you how much chicken you had on Sunday and how much you have today. It does not know you sold 87 chicken sandwiches between those counts, what each sandwich is supposed to use, and therefore what your actual variance is.

They do not catch in-period changes. A 50 lb case of tomatoes arrives on Tuesday. A drop-off invoice gets stuffed in a folder. The Sunday count two weeks later is wrong because nobody updated the spreadsheet between counts. Multiply by every delivery in a month.

They are read-only by the people who matter. The line cook prepping for service does not open the inventory spreadsheet. They eyeball the walk-in, decide they have enough, and move on. By Friday the bisque is 86'd and nobody saw it coming. Spreadsheets are not operational documents -- they are managerial postmortems.

Every inventory system starts here. You need one canonical list of every SKU you purchase, with the unit you order in, the unit you store in, and the unit you use in recipes.

A common entry looks like this: - Item: Boneless skinless chicken thighs - Vendor: US Foods - SKU: 11293-44 - Pack size: 40 lb case - Storage unit: lb - Recipe unit: oz - Conversion: 1 lb = 16 oz, 1 case = 640 oz - Current price: $3.85/lb - Storage location: Walk-in shelf B2

The upfront pain of building this list is real. Expect 4-8 hours for a typical full-service restaurant with 200-300 active SKUs. The good news: most modern inventory tools can import vendor catalogs (US Foods, Sysco, Performance Food Group all support this) so you are editing rather than building from scratch.

No item master, no inventory system. This is the load-bearing step.

Recipe costing is the conversion that turns inventory data into food cost insight. Without it, your inventory system is just a counting app.

For each menu item, build a recipe card that lists every ingredient in its recipe unit, with the cost calculated from the item master.

Example -- 8 oz chicken thigh sandwich: - 8 oz chicken thigh (at $3.85/lb = $0.24/oz) = $1.93 - 1 brioche bun (at $4.20/dozen = $0.35/each) = $0.35 - 1 oz house pickles (at $11.00/gallon = $0.09/oz) = $0.09 - 0.5 oz aioli (at $14.50/quart = $0.45/oz) = $0.23 - 0.5 oz lettuce (at $1.85/lb = $0.06/oz) = $0.03 - 0.05 lb fries (at $0.95/lb) = $0.05 - Total food cost: $2.68 - Menu price: $14.00 - Food cost %: 19.1%

Three things happen when you do this for every item:

1. You discover the menu items that are quietly bleeding margin (the $26 ribeye that actually costs $11.50 plated, not the $7.20 the old spreadsheet said). 2. You can run depletion automatically — every time a sandwich rings on the POS, the inventory system subtracts 8 oz of chicken, 1 bun, 1 oz pickles, etc. 3. You can run variance — the difference between what was sold (depleted) and what is actually missing from the walk-in.

That third number is where the money is. Variance is what every spreadsheet operation is missing.

The Sunday clipboard walk has to go. Counts on paper introduce transcription errors, slow down the process, and create a delay between the count and the variance calculation.

A modern inventory tool runs counts on a phone or tablet. The person counting walks the walk-in, scans the storage location, sees the expected count from depletion math, and enters the actual count. Variance is calculated in real time. Notes (e.g., 'found 2 lbs spoiled, threw out') are attached to the line.

The practical wins: - Counts take 30-50% less time because no transcription step. - Three people can count different sections simultaneously. - Variance is visible immediately, not on Tuesday when the manager has time to update the spreadsheet. - Photos can be attached to spoilage entries for audit trail.

Most platforms support barcode scanning for packaged goods, which speeds counts even further. Whether you scan or hand-count, the key is the device replacing the clipboard.

This is where restaurants lose money silently. A vendor delivers a case of striploin marked 14 lbs. You pay for 14 lbs. The case actually contains 12.8 lbs. Over a month of deliveries, that 8% shrink at the vendor level compounds.

Good inventory tools require receiving against the original purchase order. The receiver scans the delivery, the system shows what was ordered, and the receiver enters the actual quantity and weight received. If there is a discrepancy, the system flags it -- so you can call the vendor for a credit before the food is consumed and the paper trail is gone.

Receiving discipline is often the single highest-ROI process change a restaurant makes when getting off spreadsheets. The variance you catch in delivered weight is dollar-for-dollar recoverable from suppliers, and most operators do not realize how much they are giving away until they start counting at the back door.

A par level is the target quantity you want on hand at the start of each ordering period. Set pars correctly and ordering becomes mostly automatic.

The formula for a healthy par:

Par = (Average daily usage x days until next delivery) + safety stock

For chicken thighs at a restaurant using 60 lbs/day with deliveries every 3 days and a safety stock of 30 lbs: - Par = (60 x 3) + 30 = 210 lbs

When your inventory hits below 210 lbs heading into ordering day, the system suggests adding chicken to the order with a recommended quantity to bring you back to par.

The sophisticated version (which most modern platforms now support) uses forecasted demand instead of a flat daily average. If next Friday's POS forecast says you will do 30% more sales than the trailing average, the par for that delivery window adjusts up. If it is Restaurant Week and you are pulling 2x volume, the par doubles automatically.

This is the difference between always-running-out and always-over-stocked. Both are expensive. Forecasted-demand pars solve both.

Variance = (Theoretical usage from POS depletion) - (Actual usage from beginning count + receipts - ending count).

This is the single most important number in restaurant inventory. It tells you the gap between what should have been used (based on recipes and sales) and what was actually used (based on physical counts and receiving).

A healthy variance is under 2% for most categories. Above 4% means you have a meaningful issue -- could be theft, could be portion drift, could be missed comps, could be unrecorded spoilage. The variance report does not tell you which, but it tells you where to look.

A practical weekly ritual:

1. Pull the variance report Monday morning. 2. Identify the top 5 SKUs by dollar variance (not by percentage -- one bottle of premium tequila with 12% variance matters more than 30% variance on flour). 3. Pick one to investigate this week. Talk to the kitchen, watch portion sizes, review POS comps and voids, check the spoilage log. 4. The next week, address a different top-variance SKU.

This is what 'managing food cost' actually looks like day to day. Not a spreadsheet at the end of the month. A weekly variance review that turns into one focused investigation.

Getting off the spreadsheet is not optional for restaurants serious about controlling food cost. The variance you cannot see is the variance you cannot fix.

The six steps -- item master, recipe costing, mobile counts, receiving discipline, par-based ordering, weekly variance review -- are the same regardless of tool. The tool just makes each step faster and more accurate.

If you want a dedicated inventory platform with deep specialization, MarketMan and MarginEdge are the strong choices. If you already pay for Toast and want to stay there, Toast Inventory is the path of least resistance. If you would rather run inventory, POS, scheduling, CRM, marketing, and online ordering on one platform at a per-seat price, an all-in-one like Deelo gets you there. The math gets compelling once you are paying for three or four subscriptions to do what one platform can.