How Many Employees Should I Schedule Each Shift at My Korean BBQ Restaurant?

Why the gross-profit divisor beats the alternatives

Most Korean BBQ operators pick from four scheduling methods, and only one of them ties headcount to money. The first is habit: "we always run five on Tuesday." Habit is how you end up with a five-top crew on a Tuesday that only produced $520 in gross profit and a five-top crew on a Saturday that produced $2,600. Same labor, wildly different return. Habit doesn't fail loudly — it fails quietly, bleeding two or three unnecessary bodies on slow shifts and understaffing the shifts that would have paid for them.

The second alternative is the labor-percentage target. You set a rule — labor stays under 28% of sales — and you build the schedule to hit it. This is the restaurant industry's default and it's genuinely useful, but it has a blind spot: it works off sales, not gross profit. Korean BBQ has an unusual margin structure. Premium beef cuts move at a much lower margin than banchan, rice, soft drinks, soju, and beer. Two shifts with identical top-line sales can produce meaningfully different gross profit if one skewed toward high-end brisket and short rib and the other skewed toward pork belly combos with heavy beverage attach. A labor-percentage rule treats those two shifts as identical. The gross-profit divisor doesn't.

The third alternative is covers-based staffing — a server handles X tables, a grill attendant handles Y, so count expected covers and back into headcount. This is the most operationally intuitive method and it's the one your floor managers will reach for naturally. It's also correct as a *constraint*, and we'll come back to it, because you can't run a fourteen-table section with three servers no matter what the math says. But covers ignore check average entirely. A four-top ordering two combos and water is not the same guest as a four-top ordering premium cuts, three rounds of soju, and dessert. Covers count bodies through the door; gross profit counts what those bodies left behind.

The fourth is POS demand forecasting — the tool projects sales for a future shift and suggests coverage. This is the most sophisticated off-the-shelf option, and it's the closest cousin to what we're describing. The difference is ownership. A forecasting engine hands you a number you didn't build and can't defend to a crew member who asks why they got cut. A per-employee gross-profit target is a number you and your leadership team agreed on out loud, and every employee can be told exactly what it is: "if you show up, take care of an average number of guests, and give average service, you should produce no less than $130 a shift in gross profit." That's a floor, not a ceiling. The people who want to make real money don't coast to the floor and clock out — they hit it doing average work, then dig for the premium cut upsell, the extra banchan round, the second bottle of soju.

The honest position is that these methods stack rather than compete. Gross-profit division sets the headcount. Covers-based ratios act as a hard floor and ceiling so you never schedule a physically impossible section. Labor percentage acts as an after-the-fact audit — if the divisor produced a schedule that blew past 30% labor, either your target is wrong or your menu pricing is. And a POS forecast, if you have one, gives you a forward-looking sales estimate to run the division against instead of a trailing average. This is the same layered logic a RevOps team uses when it sizes a sales floor: a capacity model sets the number, a territory constraint bounds it, and a cost-of-sales ratio audits it.

How to choose between the methods for your specific operation

Start with data availability, because the method you can actually execute beats the method that's theoretically best. To run the gross-profit divisor you need three to six months of sales broken out by shift and day of week, plus a cost of goods number you trust. If your POS exports daily totals but not shift-level splits, you have two options: reconstruct shift splits from hourly sales data if the POS retains it, or run a four-week manual capture where a manager records the register close at the shift change. Four weeks of clean shift-level data beats six months of guessed data.

Next, decide how granular your gross profit calculation needs to be. The rigorous version pulls actual COGS by shift, which almost nobody has without theoretical-food-cost software. The practical version applies a blended food-and-beverage cost percentage to each shift's sales. If your overall cost of goods runs 32% and a shift did $3,000 in sales, call it roughly $2,040 in gross profit and divide from there. The imprecision matters less than you'd think, because you're comparing shifts against each other, and a consistent blended rate keeps the comparison honest. Where it does matter: if your lunch daypart is combo-heavy and your dinner daypart is premium-cut-heavy, run two different cost percentages rather than one blended rate, or you'll systematically overstate dinner's gross profit and overstaff it.

Then decide your target number. There's no universal figure — $130 is a worked example, not a benchmark. Build yours from the bottom: take a fully loaded hourly labor cost including payroll taxes and any benefits, multiply by average shift length, and that's what one employee costs you. A crew member costing $150 in a shift must produce meaningfully more than $150 in gross profit or the shift can't cover rent, utilities, insurance, and your own draw. Most operators land on a target somewhere between two and three times fully loaded shift labor cost. Set it, then sanity-check it against your worst realistic shift: if the target would tell you to schedule two people on a Monday lunch when you physically need four to open the doors, the target isn't wrong — you've just found a shift that structurally loses money and needs either a menu change, a marketing push, or a shorter operating window.

Finally, weigh tooling by your shape. Per-location pricing tends to win for a Korean BBQ restaurant running a large part-time and tipped roster, because you're paying once regardless of how many students and second-jobbers are on the books. Per-user pricing wins when you run a lean, stable crew of full-timers. If you want auto-suggested coverage rather than doing the division yourself, you need a tool with a live POS connection — otherwise you're supplying the headcount anyway and paying for publishing and clock-in features. And if you operate in a fair-workweek jurisdiction with predictive scheduling penalties, built-in compliance guardrails stop being a nice-to-have and become the deciding feature, because a posted-schedule violation costs real money per occurrence.

Costs, timelines, and what the method actually moves

The setup cost is mostly time, not money. Expect four to eight hours to pull and clean trailing sales data by shift and day, another one to two hours in a leadership session to argue out the target number, and two to three hours to rebuild your first schedule against the new counts. If your POS reporting is poor, add the four-week manual capture window before you can even start. Call it a month from decision to first math-driven schedule if you're starting cold, or a single week if your reporting is already clean.

Ongoing software cost varies widely by model. Free tiers exist and are real — several scheduling tools offer no-cost single-location plans with unlimited employees, which is enough to publish schedules, handle swaps, and run a mobile time clock. Paid tiers in this category generally run either a few dollars per user per month or a few tens of dollars per location per month, with restaurant-specific platforms and enterprise workforce systems sitting at the higher end and often quoting custom. Do the arithmetic on your own roster before assuming per-user is cheaper: thirty-two part-timers at $4 per user is a very different bill than one location at $35. Verify current pricing directly with each vendor — these figures move.

The impact shows up in two places. The first is elimination of the phantom body. Most operators carry one to two unnecessary people on their slowest three or four shifts per week, not out of malice but because the schedule was copied forward from a week that was busier. If a shift is overstaffed by one person for four shifts a week at, say, $16 an hour fully loaded across a six-hour shift, that's roughly $384 a week of gross profit walking out the door. Over a year that's real money that never touched the P&L.

The second and larger effect is on the busy end. Understaffing a Saturday dinner in a Korean BBQ room is uniquely expensive because the service model is labor-dense: grills need tending and swapping, banchan needs continuous refill, ventilation and tabletop turnover take a physical body, and the drink program only performs when someone is actually walking the floor. An understaffed peak doesn't just slow service — it suppresses check average, because nobody is there to suggest the premium cut or the second bottle. Adding two servers to a peak shift costs maybe $200 in labor and can protect several times that in attach revenue and turn speed.

Set expectations honestly on the timeline for results. The schedule change is immediate; the P&L signal takes six to eight weeks to separate from normal week-to-week noise. Track three numbers monthly: labor as a percentage of sales, gross profit per labor hour, and voluntary turnover. That last one matters and gets ignored — a schedule that finally staffs the peak correctly reduces burnout on your strongest crew members, and turnover is one of the largest hidden costs in a grill-forward restaurant where every new hire needs weeks of coaching before they can run a section during a rush.

Recalculate the target quarterly, and immediately after any menu price change, protein cost swing, or minimum wage adjustment. Beef pricing in particular moves enough to invalidate a target that was correct six months ago. Seasonal adjustment is also legitimate: a patio-heavy summer or a holiday-party-heavy December can justify a temporarily higher target because the traffic supports it, then a step back down in the slow months.

Implementation, station mapping, and the handoff to your managers

Getting the number is the easy part. Getting the number to survive contact with a floor manager who has favorites, a crew member who wants Fridays off, and a Saturday that inexplicably doubles is the actual work. Do it in this order.

First, publish the target openly. Don't make it a back-office metric. Every employee should be able to state it. The moment it's a secret, it becomes a weapon managers use to justify cuts, and it loses the thing that makes it work — a shared yardstick that removes bias from the schedule. No favorites, no "we've always run five," no manager scheduling their friends. Gross profit divided by target.

Second, translate raw headcount into stations, because "sixteen people" is not a schedule. In a Korean BBQ room the count typically splits across four functions: front-of-house service covering grills and banchan refills, a grill-and-prep line handling meat portioning and cut plating, a bar or beverage position when your soju and beer program justifies it, and a host running the waitlist and table turns. The exact split is yours, but write down the ratio you use so it's reproducible. If sixteen usually means nine servers, four kitchen, two bar, one host, then a twelve-person shift should scale proportionally rather than cutting entirely from one function — cutting all four from the kitchen line is how a busy Saturday collapses.

Third, place bodies against the receipt curve rather than spreading them flat. The count tells you how many; hourly sales tell you when. Pull hourly transaction data and find the real shape. In most Korean BBQ rooms the weekend evening block is dramatically denser than the shoulders — grills run continuously, parties linger past their table time, and the bar carries a large share of the margin. If your data shows a substantial majority of dinner sales landing in a two-hour window, weight your dinner staff into that window and thin the shoulders. Staggered starts are the tool here: not everyone clocks in at 4 p.m. A few open the room and set stations, the bulk land thirty to sixty minutes before the wave, and the close is covered by whoever is scheduled latest rather than by everyone waiting around.

Fourth, set hard floors and ceilings that override the math. A floor is the minimum crew required to legally and safely open — you cannot run a room with two people because the divisor said so. A ceiling is station capacity: past a certain point extra bodies don't produce, they collide. When the math exceeds capacity, that's a signal your target is too low or your prices are too low, not a signal to jam more people onto the floor.

Fifth, define the handoff protocol. The person building the schedule needs the gross-profit-by-shift table refreshed monthly, the current target, the station ratios, and the floor/ceiling constraints. Put those four things in one document that a new general manager could pick up cold. Then require a variance note: when a manager schedules off the number, they write one line explaining why — a private party booking, a holiday, a new hire shadowing. Variance isn't forbidden; unexplained variance is.

The adjacent workflows are worth wiring in while you're here. Forecasting for ordering runs off the same shift-level data — if you know Saturday dinner will do $2,080 in gross profit, you know roughly how much short rib and pork belly to thaw, which reduces both waste and the mid-service run to the walk-in. Tip pooling and distribution get cleaner too, because a schedule built on gross profit naturally puts more people on the shifts that generate more tip income, reducing the resentment that comes from a strong server being parked on a dead Monday. And hiring math falls out of the same table: total weekly headcount-hours divided by your average availability per crew member tells you whether you're structurally short-staffed or carrying too many names on the roster.

One last piece of discipline. Log actual gross profit against scheduled headcount after each shift and let that feed the next month's table. The method only compounds if the loop closes. Operators who run this well end up with a living document that shows exactly which shifts are earning their labor and which are being carried — and that's the input to the harder decisions about operating hours, daypart menus, and whether that Monday lunch service should exist at all.

Related questions

What if a shift's math says I need fewer people than I can legally open with?

That shift is structurally unprofitable at current volume. Don't schedule below the safe minimum. Instead, treat it as a business question: shorten operating hours, run a limited daypart menu, add a lunch promotion, or close that period entirely and redeploy the labor to shifts that earn it.

Should tipped and non-tipped staff use the same gross-profit target?

Generally yes for headcount math, since the divisor is about what a shift produces, not what an individual is paid. But if your fully loaded cost differs sharply between a tipped server and a kitchen position, consider separate targets by function so the kitchen line isn't judged against a front-of-house number.

How do private parties and large group bookings change the calculation?

Book them separately and staff them on top of the baseline. A confirmed twenty-person party has known revenue, so calculate its gross profit contribution directly and add the headcount it requires rather than assuming the trailing average already includes it. Log it as a documented schedule variance.

Does this method work if I'm opening a brand-new location with no history?

Not initially — you have nothing to divide. Staff the first six to eight weeks off covers-based ratios and comparable-concept benchmarks, capture shift-level data from day one, then switch to the divisor once you have a trailing eight weeks of real numbers.

Can I apply this to a second concept or a different restaurant type?

Yes. The principle — shift gross profit divided by a per-employee target — is concept-agnostic. What changes is the target, which should reflect your average check, margin structure, and labor cost. A fast-casual room with low per-guest spend sets a much lower target than a fine-dining one.

FAQ

How do I set the per-employee gross-profit target if I've never used one?

Start from cost, not aspiration. Calculate what one crew member costs you for a full shift including payroll taxes and any benefits, then set the target at a multiple of that number large enough to cover fixed costs and leave profit. Most operators land in the range of two to three times fully loaded shift labor. Write it down, sit with your leadership team, and argue about it until everyone can defend it out loud. Then treat it as a floor rather than a quota — average work should clear it, and anyone digging for upsells should beat it comfortably.

What if I don't have gross profit broken out by shift?

Reconstruct it. Most POS systems retain hourly sales even when they don't produce shift-level reports, so you can bucket hours into shifts yourself. If not, run a four-week manual capture where a manager records the register total at each shift change. Apply a blended cost-of-goods percentage to get gross profit. Use separate percentages for lunch and dinner if the menu mix differs meaningfully — combo-heavy lunches and premium-cut dinners have different margin profiles, and one blended rate will distort your dinner headcount upward.

How often should I recalculate?

Refresh the gross-profit-by-shift table monthly and revisit the target quarterly. Recalculate immediately after any menu price change, significant protein cost movement, or minimum wage adjustment. Seasonal shifts also warrant a look — a summer patio season or a heavy December party season can support a temporarily higher target, which then steps back down when traffic normalizes. The failure mode is letting a table go stale for a year and scheduling against a demand pattern that no longer exists.

Doesn't this conflict with labor-percentage targets?

They complement each other. The divisor builds the schedule; labor percentage audits it afterward. If the divisor produces a schedule that consistently pushes labor past your acceptable percentage, something upstream is wrong — either the target is set too low, your menu prices don't support your cost structure, or a particular daypart is being carried by the rest of the week. Treat a labor-percentage breach as a diagnostic signal rather than a reason to abandon the method.

How do I keep managers from ignoring the number?

Make variance visible rather than forbidden. Managers can schedule off the number, but they write one line explaining why — a booked party, a holiday, a new hire shadowing a section. Review those notes monthly. Unexplained variance is the thing to catch, because it's usually favoritism or habit reasserting itself. Publishing the target openly to the whole crew also helps, since a schedule everyone can check against a stated number is much harder to quietly bend.

Is this a restaurant-only method, or does it generalize?

It generalizes. The underlying logic — divide the revenue or margin a time block produces by what one person should produce, then place people against the demand curve — is standard capacity planning and shows up in retail floors, service counters, and RevOps sales-capacity models alike. What's concept-specific is the target number and the station ratios you use to convert raw headcount into an actual schedule.

Sources

• U.S. Bureau of Labor Statistics, Food Services and Drinking Places industry data — https://www.bls.gov/iag/tgs/iag722.htm
• U.S. Department of Labor, Wage and Hour Division — Fact Sheet on tipped employees under the FLSA — https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa
• National Restaurant Association — industry research and operations resources — https://restaurant.org/research-and-media/research/
• U.S. Small Business Administration — managing business finances and costs — https://www.sba.gov/business-guide/manage-your-business
• Cornell University School of Hotel Administration, Center for Hospitality Research — https://sha.cornell.edu/faculty-research/centers-institutes/chr/
• USDA Economic Research Service, Food Price Outlook — https://www.ers.usda.gov/data-products/food-price-outlook/
• City of San Francisco Office of Labor Standards Enforcement — Formula Retail Employee Rights (predictive scheduling) — https://www.sf.gov/information/formula-retail-employee-rights-ordinances
• U.S. Department of Labor — Overtime pay under the Fair Labor Standards Act — https://www.dol.gov/agencies/whd/overtime

Related on PULSE

• [How Many Attendants Should I Schedule Each Day at My Car Wash?](/knowledge/tl0067)
• [How Many Sales Reps Do I Need to Hire for My Logistics Company?](/knowledge/tl0058)
• [How Many Salespeople Do I Need to Hire for My Car Dealership?](/knowledge/tl0052)
• [How Many Producers Do I Need to Hire for My Insurance Agency to Grow My Book?](/knowledge/tl0015)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)