Every operator I've talked to watches their food cost percentage. They check it weekly, sometimes daily, and they know when it's crept up by half a point. What most of them can't tell me is why. The P&L shows the variance. It does not show where the variance came from. In my experience, the answer is almost always the same place: back-of-house prep.

The Gap Between Theory and the Walk-In

Theoretical food cost is the number you calculate when you multiply each menu item's recipe cost by how many you sold. It tells you what your COGS should have been if everything went perfectly — the right portions, zero waste, no over-prep left at end of night. Actual food cost is what you paid your suppliers divided by your revenue. The gap between those two numbers is called variance, and for most restaurant groups running 5 to 15 locations, that gap sits somewhere between 2 and 5 percentage points.

Two to five percent sounds abstract. On a location doing $90,000 in monthly revenue, that variance is $1,800 to $4,500 per month. Multiply that across a 10-location group and you're looking at $18,000 to $45,000 disappearing every single month without a clear explanation on any report you're already running.

We've seen operators spend months chasing that variance through supplier invoices, looking for price increases. Sometimes that's the culprit. But more often, the invoices look fine. The prices are stable. The problem is in the prep.

What Prep Invisibility Actually Looks Like

Here's the situation that creates food cost variance without anyone noticing it in real time. Your kitchen lead comes in at 7 a.m. and looks at yesterday's handwritten prep notes. The notes say to make 40 portions of the braised protein and 30 portions of the grain side. No one attached those quantities to a forecast. No one adjusted them based on the reservation count or Tuesday's historical pattern. The lead preps 40 and 30 because that's what the note said.

By 9 p.m., the walk-in has 18 portions of braised protein that will degrade overnight. That's roughly $2.70 per portion at cost, or about $48.60 in a single shift. Not catastrophic on its own. But this happens at different stations, in different forms, at every one of your locations, every single day. That accumulation is your food cost variance.

What makes it hard to fix is that no one sees it as a connected problem. The kitchen lead didn't do anything wrong — they followed the notes. The manager reviewing the P&L sees the variance number but can't tie it to a specific prep decision. The gap between the prep line and the cost report is where money goes quiet.

Why Standard Inventory Tools Miss This

Most restaurant inventory systems are designed to answer one question: did we get what we ordered? They track purchase orders, log deliveries, and flag when a supplier sends 11 cases instead of 12. That is genuinely useful. But it solves a different problem than prep visibility.

What those tools don't track is what happened between the delivery and the customer's plate. Specifically:

  • How much was prepped versus how much was needed
  • How much prep went unsold and had to be discarded
  • Whether the prep quantity was driven by a demand forecast or by habit
  • Which locations over-prep systematically versus which under-prep and create ticket time problems during service

Without answers to those questions, your food cost analysis is working from incomplete information. You know your costs went up. You don't know what generated them.

Prep Visibility Closes the Loop

The shift happens when prep quantities are forecast-driven rather than habit-driven. When your morning prep list is calculated from the previous night's cover count and a 90-day sales pattern, you're not guessing. You're producing an estimate grounded in actual demand. That forecast won't be perfect — seasonal swings, unexpected weather, large parties — but it will be systematically better than a kitchen lead eyeballing yesterday's notes.

More importantly, when actual prep output is tracked against those quantities, you start to see exactly where your COGS variance originates. If your theoretical food cost calculation shows chicken thigh portions costing $3.20 each, and your waste log shows an average of 9 unsold portions per night across a 6-location group, you now have a $172.80 per-night waste item that your P&L was obscuring as aggregate food cost variance.

That is a specific, fixable problem. Adjust the forecast model for that item, calibrate the par level for peak nights versus off-peak, track whether the change holds. This is what food cost management looks like when the prep layer is visible.

"Knowing your food cost percentage is table stakes. Knowing which prep decisions created it is how you actually control it."

The Multi-Location Dimension

Prep visibility matters even more when you're running multiple locations. At a single restaurant, an experienced manager can walk the line before service and develop an intuition for over-prep patterns. At 8 or 12 locations, that walk isn't possible. You're relying on each kitchen lead to make good prep decisions, with no shared system to surface when one location is consistently over-prepping a specific category.

We work with operators who discovered that one of their locations was over-prepping their highest-cost protein by 15 to 20% every Friday because the kitchen lead assumed Friday was always a big cover night. Some Fridays it was. Others it wasn't. Without cross-location prep data, that assumption never got challenged. It just showed up as slightly elevated food cost at that location relative to the others, which the ops manager attributed to a "rough location" rather than a specific, identifiable prep habit.

Building Toward Prep-Informed COGS

The practical path forward for most multi-unit operators doesn't require a full ERP or a warehouse management system. It requires connecting two things that currently live in separate places: the demand forecast your POS generates and the prep quantities your kitchen actually produces.

When those two data streams are linked — when your kitchen lead's prep list is generated from the same forecast your POS is running — and when actual prep output is logged and compared to that list, you have the foundation for genuinely accurate food cost attribution. Not just "our food cost is 31% this month," but "our food cost is elevated at three locations because of consistent over-prep on two high-cost prep items during Tuesday and Wednesday dayparts."

That level of specificity changes how you manage COGS. It moves food cost control from a reactive monthly exercise into an ongoing operational discipline. The P&L stops being a mystery and starts being a report that matches what you already know is happening in the kitchen.

If your food cost variance is persistent and unexplained, don't start with your supplier invoices. Start with your prep process. That's where the answer almost certainly lives — and it's the one place most operations haven't made visible yet.