The same vendor price increase costs about 7x more when you catch it at month-end instead of the same week.
It is the 22nd. Last month's P&L finally lands in your inbox. Food cost came in at 34.1% against a 30% target — four points of margin, gone. You read it twice, and then you do the only thing anyone can do with a number that old: you try to remember what happened.
Was it the beef price? The new brunch menu? The week the walk-in ran warm? Three weeks have passed since the month ended. Whatever caused it has been running quietly the whole time, and it is still running today.
This is the part of restaurant accounting nobody puts on a feature list: the report is not the problem. The delay is.
Say a distributor raises the price of your main protein by 9% on the 3rd of March. Nothing changes on the floor. The cases arrive, the line runs, the guests are happy. The increase shows up as a slightly larger invoice that nobody reads line by line.
March closes on the 31st. The books close on April 22nd. You spot the variance, pull the invoices, confirm it, and call the rep on the 25th. By then you have paid the higher price for 53 days.
Monthly closeWeekly closeIncrease startsMar 3Mar 3You see itApr 22Mar 9Days at the higher price~53~7Cost at $12,000/mo protein spend≈ $1,890≈ $250
Illustrative figures. The exact dollars will be different in your restaurant — the ratio will not. The same increase costs roughly seven times more when you find it a month later.
Multiply that by the number of line items on a single week of invoices. Nobody is losing four points of margin to one dramatic event. It is thirty small ones, each invisible on its own, each running for seven weeks before anyone gets a chance to notice.
Restaurants do not close slowly because the people doing the closing are slow. They close slowly because of three structural facts that have nothing to do with effort.
They arrive as paper on a clipboard by the back door, a PDF attached to an email, a photo texted at 11pm by a manager who is closing out. Someone has to turn all of that into rows in a system — and because it is a batch job, it gets done as a batch job, once, after the period ends.
Every distributor names things differently. Is CHKN WNG JMB 40# poultry or is it under a generic protein bucket? Was that shredded mozzarella coded to dairy last month or to food — other? Consistency across hundreds of line items depends on one person remembering what they decided in a previous period. See how automated invoice coding works.
February has 28 days. March has 31. One of them has five Fridays and the other has four. Compare the two P&Ls and the busier month wins, every time, regardless of how well either one was actually run. A number you cannot compare is not a number you can manage.
Shortening the loop does not mean doing month-end four times a month. It means moving the work that creates the delay to the moment the delivery lands, so the numbers are simply there when you want them.
The one number worth watching weekly
Prime cost — food, beverage, and labor as a share of sales — moves fast enough to be worth checking every week and matters enough to act on when it moves. Almost everything else on a P&L can wait for the period close.
“Protein cost is running 4.1% above last week at Riverside — driven by a single vendor price change on Tuesday.”
The kind of thing Dex surfaces on a Monday morning, instead of on the 22nd
Once you are reading numbers weekly, the calendar problem gets worse before it gets better — four-week months next to five-week months make every trend look like noise. This is what fiscal period accounting solves. A 4-4-5 calendar splits the year into periods of equal shape, each with the same number of Fridays and Saturdays, so period three and period four are genuinely the same size.
It sounds like a bookkeeping detail. In practice it is the difference between a trend you can trust and a chart that mostly tracks how many weekends fell inside the month. More on fiscal calendar accounting.
None of the following requires new software. They are worth doing whatever you run your books on.
You cannot manage a number you meet three weeks after it happened. You can only explain it. The goal was never a faster month-end — it is a shorter distance between something changing in your restaurant and you finding out. Close that gap and the monthly P&L becomes what it should have been all along: a confirmation of what you already knew, rather than the first you heard of it.
