How Many Employees Should I Schedule Each Shift at My Food Hall in 2026?

Two ways to set the number: gross-profit division versus labor-percentage targeting

Almost every food hall staffing conversation collapses into one of two methods, and they produce genuinely different schedules on the same week of sales data.

Method one is gross-profit division. You agree on a single number — the gross profit an average employee should generate on an average shift doing an average job — and you divide each shift's gross profit by it. Set the floor at $200 a shift and a Saturday dinner doing $2,400 in operator-controlled gross profit gets twelve bodies. The number is absolute. It does not care what your labor percentage was last month, it does not care that you "always run ten on a Saturday," and it does not care that a manager wants to give a friend hours. It asks one question of every person on the schedule: is this shift producing at least $200 of gross profit for you to stand on? A fourth busser on a dead Tuesday cannot answer yes, so the fourth busser does not get scheduled.

Method two is labor-percentage targeting. You set labor cost as a share of revenue — most full-service operations run shared front-of-house labor somewhere between 18% and 30% of the revenue that labor supports — and you schedule up to that ceiling. If Saturday dinner is forecast at $9,000 in shared revenue and you target 22% labor, you have roughly $1,980 to spend. At a blended $22 per hour fully loaded, that is about 90 labor hours, which across a six-hour dinner block is fifteen people. Notice that the two methods disagree. Percentage targeting bought you three more bodies than gross-profit division did on the same night.

Neither answer is automatically right, and the disagreement is the useful part. Gross-profit division anchors to margin — it uses gross profit, which is revenue minus cost of goods, so a night that sells a lot of low-margin product does not earn extra staff. Labor-percentage targeting anchors to top-line revenue, so it will happily staff up a high-volume, low-margin night and quietly shred your contribution margin. In a food hall this matters more than in a single restaurant, because your shared-space revenue mix swings hard. A beer-heavy Friday at 75-80% pour margin and a Sunday brunch heavy on prepared food at 55-60% margin can post identical top-line numbers and support wildly different crews.

The counter-argument is real, though. Gross-profit division needs clean cost-of-goods data by daypart, and plenty of halls do not have it. Percentage targeting only needs a POS sales report, which every hall has on day one. If you cannot yet split COGS by shift, run percentage targeting as a bridge, get the COGS reporting built over a quarter, then switch. Do not pretend you are running gross-profit division on numbers you are actually guessing at.

A third option deserves a mention only to be dismissed: fixed-template scheduling, where every Saturday gets the same ten people because that is what the last general manager did. It is not a method, it is an inherited habit, and it is the single most common reason a hall is simultaneously overstaffed on Tuesday and drowning on Friday.

How to decide between them

The decision hinges on three things: the quality of your cost data, the volatility of your margin mix, and whether you are staffing to protect margin or to protect guest experience during peaks.

Start with data quality. If your POS can attribute cost of goods to a shift — because your bars ring through a system that knows pour cost and your shared-service revenue is categorized — gross-profit division is available to you and you should use it. If your COGS only closes monthly at the entity level, you cannot divide by shift honestly, and forcing it produces a schedule built on invented numbers, which is worse than an admittedly crude percentage target.

Then look at margin volatility. Pull twelve weeks and compute gross margin percentage by shift. If your best and worst shifts sit within about four points of each other, the two methods converge and you should pick the one that is cheaper to operate — usually percentage targeting, because it needs less reporting. If your spread is ten points or more, which is typical for a hall with a strong bar program and a soft weekday lunch, gross-profit division will save you real money because it refuses to reward volume that does not carry margin.

Finally, decide what the schedule is for. There is a legitimate service-floor argument that overrides both methods at the edges: a hall with six vendor stalls open needs enough runners and bussers to keep tables turning even on a shift where the math says two people. If dirty tables cost you vendor tenants, you staff above the math and you write down why. The RevOps discipline here is the same one you would apply to a sales territory — the model sets the baseline, documented exceptions adjust it, and you never let an undocumented exception become the new normal.

Run this loop weekly for the first quarter, then monthly. The feedback edge at the bottom is not decorative — the whole system depends on comparing what each shift actually produced per scheduled employee against the target you set, and adjusting the target when reality disagrees with it for four weeks running.

The concrete numbers behind each method

Here is a worked week for a mid-size food hall with eight vendor stalls, two operator-run bars, and shared seating for about 180. The operator controls bartenders, barbacks, runners, bussers, dish, and floor leads. Vendors staff their own counters and cooks, and none of that headcount belongs in this math.

Trailing thirteen-week gross profit by shift, operator-controlled only:

• Monday lunch: $410. Monday dinner: $780.
• Tuesday lunch: $600. Tuesday dinner: $940.
• Wednesday lunch: $650. Wednesday dinner: $1,150.
• Thursday lunch: $720. Thursday dinner: $1,600.
• Friday lunch: $980. Friday dinner: $2,100. Friday late bar: $900.
• Saturday lunch: $1,300. Saturday dinner: $2,400. Saturday late bar: $1,100.
• Sunday brunch: $1,450. Sunday dinner: $820.

Divide by a $200 per-employee target and the schedule writes itself: Monday lunch 2, Monday dinner 4, Tuesday lunch 3, Tuesday dinner 5, Wednesday lunch 3, Wednesday dinner 6, Thursday lunch 4, Thursday dinner 8, Friday lunch 5, Friday dinner 11 (10.5 rounded up), Friday late bar 5 (4.5 rounded up), Saturday lunch 7 (6.5 rounded up), Saturday dinner 12, Saturday late bar 6 (5.5 rounded up), Sunday brunch 7 (7.25 rounded down), Sunday dinner 4.

That totals 92 shift-slots across the week. At an average six-hour shift, roughly 552 scheduled hours. At a fully loaded blended rate of $22 per hour — wage plus payroll tax, workers' comp, and any benefit load — that is about $12,144 in weekly operator-controlled labor against roughly $17,900 in weekly gross profit. Labor consumes about 68% of the gross profit those employees generate, leaving about $5,750 a week toward rent, utilities, insurance, and management salary. If that residual does not cover your fixed costs, your $200 target is too low, not your headcount too high.

Now run the same week on labor-percentage targeting. Assume shared-space revenue of about $31,000 for the week and a 22% labor target: $6,820 of budget, which at $22 loaded buys 310 hours — about 52 shift-slots. That is 40 fewer bodies than gross-profit division produced. The methods diverged wildly, and the reason is instructive: at $17,900 gross profit on $31,000 revenue, this hall runs a 58% blended gross margin, so a percentage-of-revenue target is anchoring to a number nearly twice the size of the money actually available. If you use percentage targeting, target a percentage of *gross profit*, not revenue — 60-70% of operator-controlled gross profit is a defensible band — or you will systematically over- or under-staff depending on your margin.

How to set your own per-employee target rather than copying $200. Take trailing annual operator-controlled gross profit, subtract your fixed costs and the profit you require, and divide the remainder by the number of shift-slots you expect to schedule. If the hall produces $930,000 in annual shared gross profit, fixed costs run $340,000, and you require $90,000 of owner profit, you have $500,000 for labor. At 4,800 annual shift-slots, that is roughly $104 of labor per slot — and at a $22 loaded rate over six hours ($132 per slot), the math does not close. Either the target rises, the shifts shorten, or the fixed-cost base has to come down. Most halls land their per-employee target between $150 and $300 depending on wage market and margin structure; $200 is a common center, not a law.

Rounding rules matter more than they look. Always round up on peak shifts and down on troughs. A Friday dinner at 10.5 rounds to 11 because being one runner short during the peak costs you table turns worth far more than one wage; a Sunday brunch at 7.25 rounds to 7 because the marginal person on a moderate shift is genuinely marginal. Apply the same asymmetry every week so the schedule does not drift.

Position mix inside the count. The formula gives you a number, not a roster. For a twelve-person Saturday dinner in this hall: 4 bartenders, 1 barback, 3 runners, 2 bussers, 1 dish, 1 floor lead. For a five-person Tuesday dinner: 2 bartenders, 1 runner, 1 busser combined with light dish, 1 floor lead who also works the floor. Below about four people, roles merge and you should expect service quality to depend heavily on who specifically is working — which is why cross-trained staff are worth a premium in a hall.

Implementation details and sequencing

Rolling this out badly gets you a mutiny and a worse schedule than the one you replaced. Sequence it deliberately over about eight weeks.

Weeks one and two — build the data. Get operator-controlled revenue and COGS split by shift in your POS. Define shift boundaries explicitly (lunch 11:00-16:00, dinner 16:00-22:00, late bar 22:00-close) and make sure every ring lands in exactly one bucket. Pull thirteen weeks of history if you have it; eight is workable; four is too noisy to trust. Strip out anomalies — a street festival Saturday, a two-day closure — and note them separately rather than letting them distort the average.

Week three — set and socialize the target. Do not announce the number, derive it in a room with your managers using the fixed-cost math above, and say the standard out loud: in this food hall, if you show up, keep the floor turning and the bars pouring, and give average service, you should be covered by no less than $200 a shift in gross profit. Framing it as a floor everyone can clear is the difference between a standard and a threat. Expect pushback that the number punishes slow shifts; the honest answer is yes, and that is the point — slow shifts should not carry passenger headcount.

Week four — build the schedule in parallel, publish the old one. Run both schedules for one week without acting on the new one. Compare them shift by shift and interrogate every gap of two or more people. Sometimes the model is wrong because your COGS attribution is off. Sometimes the old schedule was wrong for three years. You need to know which before you cut anyone's hours.

Weeks five and six — place bodies against the receipt curve. The count tells you how many; hourly sales tell you when. Pull hourly rings and find the real shape: most halls show a 12:00-13:30 lunch bump, a genuine dead zone from about 14:30 to 16:30, a dinner build from 17:30 peaking near 19:30, and a late bar push after 21:00 on Friday and Saturday. Stagger start times against that curve rather than starting everyone at once. On a twelve-person Saturday, that might be four on at 15:30 for setup and early dinner, five more at 17:00 for the peak, and three bar-weighted at 20:00 who carry the late push and close. Staggering is where most of the savings actually live — a flat twelve-at-once schedule wastes roughly two hours per person on the shoulders.

Weeks seven and eight — enforce and measure. Publish the new schedule at least fourteen days out, honoring predictive-scheduling law if your city has one. Track actual gross profit per scheduled employee after every shift and post it where the crew can see it. When a shift comes in under target three weeks running, cut a body from it. When it beats target by 25% consistently, add one — the model works in both directions, and staff will only trust it if they see it add hours, not just remove them.

Compliance guardrails to build in from the start. If you operate in a predictive-scheduling jurisdiction, late changes carry premium pay, so the model has to produce a schedule you can commit to two weeks out rather than one you tune on Thursday. Watch split shifts — a person working lunch and dinner with a four-hour gap may trigger split-shift premium pay depending on jurisdiction. Watch overtime: with a large part-time roster it is easy to push a favorite past 40 hours across scattered shifts, which quietly destroys the per-employee math because those hours cost 1.5x. Set a hard cap in your scheduling tool and audit it weekly.

What breaks this and how to catch it. Three failure modes recur. First, vendor-stall staff creeping into the count — a vendor short-staffs their counter, your runners absorb it, and your gross profit per employee craters through no fault of the schedule. Track it and bill it back or renegotiate the stall agreement. Second, seasonality treated as noise — a hall with a patio can swing 20-30% between July and February, so recalculate quarterly rather than annually. Third, the target never moving — if wages rise 8% and your target stays at $200, you are slowly scheduling yourself out of profit. Reprice the target every time your loaded hourly cost moves more than about 5%.

Related questions

Should vendor stall employees count toward my per-shift headcount?

No. The formula covers only operator-controlled staff — bartenders, barbacks, runners, bussers, dish, and floor leads. Each vendor schedules their own counter and kitchen staff against their own margins. Mixing the two makes both numbers meaningless.

What if a shift's gross profit does not support even one employee?

Then that shift should not run with dedicated staff. Either fold it into an adjacent shift, close it entirely, or run a single cross-trained person handling light service and cleanup while accepting the shift as a loss leader for hall foot traffic.

How does this differ from staffing a single restaurant?

A food hall splits labor across shared space and independent vendors, so your revenue base is bars, shared service, and any hall-level fees — not food sales. Margin mix swings harder, which makes gross-profit division more valuable than in a single-concept restaurant.

How far in advance should I publish the schedule?

At least fourteen days, and legally required at that horizon in several predictive-scheduling cities. The model supports it: gross profit by shift and day of week is stable enough that a two-week-out schedule built from trailing averages rarely needs material changes.

Does the target change for different positions?

The target is a blended average across operator-controlled roles, not a per-position quota. A dishwasher does not individually generate $200 of gross profit, and should not be asked to. The number governs total headcount on a shift; role mix is a separate service decision.

FAQ

What if my food hall's per-employee target should not be $200?

Derive it rather than borrowing it. Take trailing annual operator-controlled gross profit, subtract fixed costs and required owner profit, and divide the remainder by expected annual shift-slots. Most halls land between $150 and $300 depending on wage market, margin structure, and how much shared square footage the operator staffs directly.

How do I calculate gross profit for a specific shift?

Take that shift's revenue from all operator-controlled sources — bars, shared service, any hall fees — and subtract the cost of goods sold against it. Use thirteen weeks of history to smooth spikes, strip out festival days and closures, and make sure every POS ring maps to exactly one shift bucket before you trust the average.

What happens if I schedule to the number and service quality drops?

Raise the service floor, not the method. Define the minimum bodies required to keep tables turning when all stalls are open — often three or four regardless of math — and treat it as a documented override. The model still governs everything above the floor. Undocumented overrides are how fixed-template scheduling creeps back in.

How often should I recalculate?

Recalculate shift averages quarterly, and immediately whenever loaded hourly labor cost moves more than about 5% or your vendor mix changes. Seasonal swings of 20-30% are common in halls with patio or event space, so a schedule built on annual averages will be wrong in both directions for half the year.

Can I run different targets for different shifts?

You can, but resist it. A single blended target is what makes the schedule defensible to staff and immune to favoritism. If a shift genuinely carries structurally different economics — a private event buyout, a recurring market day with hall-level fees — model that shift separately and label it, rather than quietly softening the target.

Does scheduling software replace this math?

No. Restaurant scheduling tools handle distribution, swaps, clock-in, and labor-law guardrails, and several will forecast coverage from a POS feed. But they forecast against revenue by default, not gross profit. You supply the target and the margin logic; the tool executes the schedule and enforces overtime and break rules.

Sources

• https://www.bls.gov/oes/current/oes_nat.htm
• https://www.dol.gov/agencies/whd/flsa
• https://www.sba.gov/business-guide/manage-your-business/hire-manage-employees
• https://restaurant.org/research-and-media/research/
• https://www.irs.gov/businesses/small-businesses-self-employed/business-expenses
• https://www.dol.gov/agencies/whd/overtime
• https://www.score.org/resource/business-planning-financial-statements-template-gallery
• https://www.osha.gov/restaurant-industry

Related on PULSE

• [How Do I Negotiate a Food Hall Stall Lease and Buildout?](/knowledge/q13779)
• [How Many Staff Should I Schedule Each Shift Across My Food Truck Fleet?](/knowledge/q15607)
• [How Many Sales Reps Do I Need to Hire for My Food Distribution Business?](/knowledge/q15577)
• [Should I open a food truck in 2027?](/knowledge/q15031)
• [Should I open or buy a Mo' Bettahs Hawaiian Style Food franchise in 2027?](/knowledge/q14916)