A QSR snack brand location saw COGS decline by 9.7% year over year and COGS margin improve from 15.67% to 15.34% while using consistent weekly counting and incremental PAR level corrections. No dramatic changes, no menu overhauls, no supplier switches. Just slightly better data every week than the week before.
Not every story is a comeback. Not every location starts with a 20% COGS ratio and slashes it in half. Some locations are already running reasonably well, and the improvement is quiet. A little better each week. A slightly smaller order. One fewer case of something that was sitting too long. Over a full year, those small corrections add up to real money.
This is the story of a QSR snack brand location that was operating fine. Not great, not terrible. COGS at 15.67% of sales. Reasonable for the category. But the operator knew there was room. The problem was finding it without data.
When COGS is at 15.67% and the location is profitable, nobody panics. The monthly review is short. The numbers look OK. But "OK" is where waste hides. A few items over-ordered here. A PAR level set too high there. None of it is dramatic enough to trigger a conversation, but all of it is money left on the table.
Without weekly counts, there is no way to see the small inefficiencies. They are invisible in a monthly report. A 2% over-order on five items does not show up as a problem. It shows up as "normal."
The location started counting every item every week. Within the first few weeks, the data started telling a story. Not a dramatic story. A quiet one.
A few items were consistently showing usage below their PAR levels. Not by a lot. Maybe 10% below. But 10% below PAR on 20 items, every week, for a year, is real money. The rolling 3-week average caught it. The PAR levels came down. The order got slightly smaller.
Here is what the full-year numbers looked like:
| Period | COGS as % of Sales | Total COGS | Total Sales |
|---|---|---|---|
| Prior Year (full year-to-date) | 15.67% | $83,411 | $532,164 |
| Current Year (full year-to-date) | 15.34% | $75,343 | $491,190 |
| Change | -0.33 percentage points | -$8,068 | -$40,974 |
$8,068 in COGS reduction was observed at a single location in one year. Not from a dramatic turnaround. From small, consistent corrections that compounded over 52 weeks of counting.
Most operators will never see a 45% COGS reduction. That happens when a location is significantly over-ordering and the correction is dramatic. Most operators are already doing a reasonable job. They are not ordering 40% too much. They are ordering 5% too much on some items and 10% too much on others.
The weekly count does not address this with one dramatic move. It addresses it with 52 small moves. Each week, the rolling average gets a little more data. Each week, a PAR level might come down by a small amount. Each week, the order is slightly more accurate than the last.
This is the story most operators can actually see themselves in. Not the turnaround. Not the big save. The steady grind of slightly better data, week after week, until the annual number looks different.
Here is what happens when the manager counts every week. The app shows a Stock Dot for each item: green, yellow, red, or gray. Green means you are at or above PAR. Yellow means you are approaching PAR. Red means you are below PAR and need to order. Gray means no data yet.
Over time, items that are consistently green start to get their PAR levels questioned. If an item is green every week, the PAR level might be too high. The rolling average catches it and brings it down. The next order is smaller for that item.
Items that are hitting red too often get their PAR levels raised. The rolling average catches that too.
None of this requires the manager to make a decision. The app does the math. The manager counts. The app adjusts. The order improves.
If your COGS looks reasonable and you are not sure where the savings are, the answer is in the weekly data you are not collecting. You do not need a turnaround to justify weekly counting. You need the small corrections that happen when you have real usage data instead of habit.
$8,068 at one location. If you have five locations, that is $40,000. If you have 15, that is $121,000. None of it from dramatic changes. All of it alongside counting every item every week and letting the data adjust the order.
No. This location was already operating at a reasonable 15.67% COGS ratio. The improvement came from finding small inefficiencies that were invisible in monthly reports. Weekly counting does not require a crisis to generate value. It generates value by making every order slightly more accurate than the last one.
$8,068 in COGS reduction was observed at this location over one year alongside incremental PAR corrections. The reduction comes from items that were consistently over-ordered at their existing PAR levels. Each correction is small, but over 52 weeks across all items, the total adds up.
Yes. The more weeks of data the rolling average has, the more accurately it reflects real usage. The first few weeks are the least accurate because the average is working with limited data. By week four or five, the average is stable and PAR adjustments become reliable.