Bookkeeping Hospitality Small Business
How to Read Your Weekly Restaurant P&L in About Fifteen Minutes
Published September 14, 2026 by Invisible LLC Team · 9 min read
Here is the sentence we hear most often from operators, in some version or another:
"We had a great Saturday and a brutal Tuesday and the week was a wash. I need to see that on Sunday morning, not the fifteenth of next month."
That is the whole argument for a weekly P&L. A monthly statement that arrives two weeks after the month closes is a historical document. By the time it tells you labor ran at 38%, you have already worked four more weeks at 38%.
The takeaways. A weekly read is five lines, in order, and it takes about fifteen minutes once someone has set it up properly. The line that does the most work is labor cost percentage — and most operators calculate it in a way that isn't comparable week to week. The hard part isn't the reading. It's getting the POS, the processor, and the bank to agree before you read anything.
Why the monthly close is the wrong instrument
Nothing is wrong with a monthly P&L. It is the right document for your accountant, your lender, and your tax return. It is simply the wrong tool for running a restaurant, for three reasons.
It is too slow. Restaurant problems are weekly problems. A schedule that was built wrong, a vendor price increase that nobody caught, a promotion that ran at negative margin — all of those compound for a month before a monthly P&L surfaces them.
It averages away the signal. A month with two great weeks and two terrible ones looks like an average month. You cannot fix an average. You can fix a Tuesday.
It can't answer the question you actually have. "Was my labor high because I overhired in January or because we got slow in February?" is unanswerable from a monthly number. It is obvious from four weekly ones.
The weekly P&L doesn't replace the monthly close. It sits in front of it.
The five lines, in the order you read them
Read them in this sequence. The order matters, because each line frames the next.
1. Net sales
Not gross rings. Net sales after comps, voids, discounts, and promotions.
This is the first place a weekly read goes wrong. Your POS "sales" number and your net sales are not the same figure, and the gap between them is often the most interesting number of the week. A comp line that quietly doubled is a management problem wearing a sales costume.
Break it at least into food and beverage. If you run a bar program or a brewpub, break beverage further — beer, wine, spirits behave differently and blend into a meaningless average.
2. Cost of goods sold, as a percentage
Food cost and beverage cost, separately. A combined COGS percentage hides the thing you need to see, because a good beverage week can mask a bad food week almost perfectly.
Weekly COGS is where operators most often decide the whole exercise is too hard, because doing it properly means counting inventory. Our position: a full count weekly is unrealistic for most independents. A key-item count — your ten or fifteen highest-cost, highest-volume items — captures most of the variance in a fraction of the time and is far better than purchases-as-a-proxy.
3. Labor cost, as a percentage
The line that repays the most attention, and the one that is most often computed in a way that can't be compared to last week. It gets its own section below.
4. Prime cost
COGS plus labor. This is the number most operators should manage to, because it is the sum of the two costs you can actually move inside a week.
We wrote a full walkthrough of prime cost for restaurants and brewpubs — what to include, what to leave out, and how to track it without a controller. If you only take one number from a weekly read, take this one.
5. Controllables and occupancy — but below the line
Rent, insurance, and debt service do not change based on how you ran service this week. Put them below the line so they don't distort the weekly read. What does belong up top is the controllable middle: repairs, smallwares, cleaning, delivery commissions, credit card fees.
Third-party delivery commission in particular deserves its own line rather than being buried. It is frequently the difference between a channel you think is profitable and one that isn't.
Labor cost percentage: how to compute it so it's comparable
Labor cost percentage is simple arithmetic — total labor cost divided by net sales — and almost every operator we meet computes the numerator differently than they did last month.
Use fully-loaded labor. Wages alone understates your real cost by a meaningful margin. Fully loaded means:
- Gross wages (hourly, salaried, and overtime premium)
- Employer-side payroll taxes
- Benefits you actually pay
- Any employer-funded portion of insurance
Management salary is the judgment call. Include it, exclude it, or track both — but pick one and never change it, because the whole value of this number is period-over-period comparability. A labor percentage that means something different in week 12 than it did in week 8 is worse than no number at all.
Split front of house and back of house. A blended labor percentage tells you labor is high. A split one tells you the kitchen is overstaffed on Mondays. Only one of those is actionable.
Watch for the tipped-wage distortion. If you take a tip credit, your labor cost line reflects the cash wage you pay, not what the employee earns. That is correct — it is your cost. But it means your labor percentage is not comparable to an operation that doesn't take a credit, and it is not comparable to your own prior periods if your tipped wage rate changed. Tipped wage rates move on a schedule in many jurisdictions, and 2026 has been an unusually active year for them. If your labor line stepped up without your schedule changing, that is probably why — we walk through the current state of the rules in our tip credit guide.
Sales per labor hour is the better companion metric. Labor percentage moves when sales move, which means a slow week looks like a staffing failure even when you scheduled perfectly. Sales per labor hour separates the two.
The unglamorous part: making the POS agree with the bank
This is where the weekly P&L actually lives or dies, and it is the reason most operators who try to do this themselves give up by week three.
"The DoorDash deposit is net of commission and I can't reconcile it to the day's POS."
Right. And the card processor deposits net of fees, on a lag, sometimes batching a weekend into one Monday transaction. And gift cards are a liability when sold and revenue when redeemed, which most POS reports do not present that way.
The four reconciliation points that cause almost all of the pain:
- Card settlement timing. What the POS recorded Saturday hits the bank Monday or Tuesday, often batched. Match on settlement batch, not on day.
- Processor fees. Deposits arrive net. If you book the net figure as revenue, you understate both sales and expenses, and your COGS and labor percentages are computed against a denominator that is too small.
- Third-party delivery. Same problem, larger percentage. Book gross sales and the commission as an expense, every time. If you net it, you will never know what delivery actually costs you.
- Gift cards and house accounts. Sold is a liability. Redeemed is revenue. Getting this backwards inflates the week you sold them and deflates the week they're used.
None of this is conceptually hard. It is just genuinely tedious, it has to happen every week, and it has to happen before the numbers mean anything. This is the part that belongs on someone else's desk.
The Sunday morning routine
The cadence that works, in our experience with operators:
- Books close Sunday for the week ending Sunday. One consistent week definition. Don't switch between calendar weeks and fiscal weeks.
- The packet is one page. Five lines, this week, last week, and same week last year if you have it. Not a dashboard. Not a twelve-tab workbook.
- You read it Sunday or Monday morning, before you build next week's schedule — because the schedule is the lever the numbers are supposed to move.
- Two questions per week, maximum. "Why did beverage cost jump?" and "Why was BOH labor high Monday?" You are not conducting an audit. You are steering.
The discipline is the cadence, not the sophistication. A rough weekly number you actually look at beats a precise monthly one you don't.
What "good" looks like — and why published benchmarks mislead
You will find widely circulated targets for food cost, beverage cost, labor, and prime cost. Treat all of them as orientation, not as standards.
They vary enormously by concept. A quick-service operation and a full-service steakhouse have structurally different cost profiles. A brewpub producing its own beer has a beverage cost that is not comparable to a bar buying kegs wholesale. A market with a $17 minimum wage and no tip credit has a labor line that would look alarming in a market with a $7.25 floor — and both can be healthy businesses.
The benchmark that matters is your own trailing thirteen weeks. Direction and consistency tell you more than any published range. If prime cost has drifted up three points over a quarter, that is a real finding regardless of where it sits against someone else's target.
If you want the general mechanics of reading a P&L outside the restaurant context, our plain-English income statement guide covers the underlying structure.
Where this fits
The weekly P&L is not a reporting exercise. It is the instrument that tells you whether last week's decisions worked, in time to change this week's.
Getting there takes two things. Someone has to reconcile POS, processor, delivery, and bank every week so the numbers are true — that's bookkeeping. And someone has to build the one-page packet so it reads in fifteen minutes instead of requiring interpretation — that's custom reporting.
We run this weekly for hospitality operators, alongside payroll for mixed tipped and salaried staff. Not a dashboard you have to learn. A person who knows what 86'd means, sending you one page on Sunday.
Tell us what last week's numbers didn't tell you. Thirty minutes, free, and we'll be honest if we're not the right fit.