A prime cost that looks great is usually a prime cost that is missing something
Full-service prime cost lives between 55 and 60% of sales. If yours reads well under that and nothing about how you buy or staff has changed, the likeliest explanation is not a brilliant month. It is a cost that is not being counted.
· 7 min
Prime cost is food plus labour over sales. Three inputs, one ratio, and the whole thing rests on an assumption almost nobody checks: that all three inputs are actually there.
When one of them is missing, the arithmetic does not break. That is the problem.
How a missing cost becomes a good month
A day has no invoice entered. The food cost for that day is not unknown — in most systems it is zero. Zero is a number. It divides. It averages. It flows into the month exactly like a real figure, and it pulls the ratio down.
Do that for a handful of days and the effect is small. Do it for most of a month and prime cost can read fifteen or twenty points below the truth, on a screen that gives no sign anything is missing.
A month where food cost is recorded on 1 day in 9:
What the screen shows ≈ 29% prime
What the complete days say ≈ 50% prime
Same restaurant. Same month. One number is a report on the business and the other is a report on the paperwork.
:::key The asymmetry that makes this dangerous An OVERSTATED cost gets caught within a day, because an operator looking at 71% prime picks up the phone. An UNDERSTATED one is never investigated. Nobody files a bug about good news, so the wrong number in this direction can sit there for half a year while everybody feels fine. :::
Why food is nearly always the input that goes missing
Labour tends to be complete because it arrives on its own. Hours come off a clock or a schedule whether anybody thinks about it or not.
Food cost has to be entered. Someone photographs or keys the invoices, and that is a task with no deadline attached to it and no immediate consequence for skipping. So it slips, and it slips silently.
There is meant to be a second rail for exactly this reason: if every menu item is linked to a costed recipe, the POS can produce a theoretical food cost every night without anyone touching an invoice. Most kitchens have not connected that rail — the recipes exist, the menu exists, and nothing joins them — which is why a missed invoice week becomes a zero instead of an estimate.
The three-question check
Run this on your own last six months, in whatever tool you already have.
1. On how many of the days in this window is each input actually PRESENT?
2. What does the ratio say on ONLY the days where all of them are?
3. If (2) is much worse than the headline, the headline is a coverage report.
Most dashboards will not answer question one, which is itself the answer to a different question you should be asking about that dashboard.
What a tool should do instead
Show the days. A prime cost figure should carry the denominator it was built from — days traded, days with a labour record, days with a food cost — so that a period with thin coverage announces itself on the face of the number rather than averaging its own gaps into a flattering result.
And a missing figure should render as missing. Not as zero, not as a silent omission from an average. We wrote about that design rule, and what it costs to hold, in withheld, not zero.
The free prime-cost calculator runs the arithmetic on figures you type in and stores nothing. Feed it a food cost you are not confident about and it will give you an answer exactly as confident as that input — which is the honest behaviour for a tool handed an honest guess.
heard, operator.
SavorVitals runs this number every day — food from your confirmed invoices, labor from SavorCrew or your own entry, your target band drawn on the chart.
See your own numbers →