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

Staffing to margin versus the alternatives operators actually use

Most independent restaurants schedule one of four ways, and only one of them survives contact with a P&L.

Habit staffing is the default: "we've always run eight on a Saturday." It is fast, requires no data, and quietly institutionalizes every past mistake. The eight was probably correct in 2019 when your check average was lower and your menu had fewer scratch components. Nobody re-derived it. Worse, habit staffing is politically sticky — the shift lead who padded Fridays with two friends three years ago has now made that padding the baseline everyone defends.

Cover-count staffing is the next rung up: some fixed ratio of servers to tables or guests, typically one server per 12–16 covers in full service. This is genuinely better than habit because it responds to volume. Its flaw is that covers do not pay payroll — margin does. A vegan brunch room turning 90 low-check covers at $14 a head and a dinner room turning 60 covers at $38 a head produce wildly different gross profit from similar-looking rosters. Cover ratios treat those two shifts as comparable when the money says they are not.

Labor-percentage staffing — hold labor to, say, 28–32% of sales — is what most POS-connected schedulers optimize toward, and it is a real improvement. The catch is that it works on *sales*, not gross profit. A vegan kitchen's food cost varies enormously by daypart: a juice-and-smoothie-heavy brunch carries a different COGS profile than a dinner service built on house-made seitan, cashew cheeses, and long-braise prep. Two shifts that produce identical sales can leave you with meaningfully different margin, and a labor-percentage rule can't see that gap.

Gross-profit-per-employee staffing is the method on this page. You set one number on purpose — the daily gross profit an average employee working an average shift ought to generate — and one number comes off your books — trailing gross profit by day of week. Divide. The headcount falls out. It absorbs the strengths of the other three: it responds to volume like cover ratios, it constrains cost like labor percentage, and it produces a defensible standing roster like habit staffing, but one derived rather than inherited.

The honest trade-off: this method demands clean COGS data by day. If your inventory is a shoebox and you cannot separate Tuesday's food cost from Thursday's, you cannot run it faithfully yet. Fix the bookkeeping first, or run the method on a rolling weekly gross-profit figure — coarser, but still miles ahead of habit. A second limitation: the division gives you a headcount, not a skills mix. Six bodies is not the same answer as "two on hot line, one cold prep, two servers, one register," and a heavy-prep vegan menu is unusually sensitive to that distinction. The arithmetic sets the budget; your chef spends it.

Adjacent context worth noting: this is the same logic RevOps teams apply to quota capacity planning — pipeline coverage divided by per-rep productivity yields headcount. Restaurants arrived at it later, but it is the identical shape, and the same failure mode applies in both worlds. Set the per-head target too low and you overstaff into a margin leak nobody notices monthly; set it too high and you understaff into ticket times, walkouts, and one-star reviews that cost more than the labor you saved.

How to choose between the staffing methods

Pick by what your books can actually support and how much variance your room carries week to week.

Start with data quality. If you can pull gross profit split by day of week for a trailing three to six months, use the gross-profit division — it is strictly better than everything else available to you. If you have sales by day but no reliable COGS split, run labor-percentage staffing as an interim and fix inventory in parallel. If you have neither because the restaurant is under six months old, use cover ratios seeded from comparable rooms in your market, and re-derive on real data at the six-month mark.

Then weigh variance. A vegan restaurant near a university with a 40% swing between term and break needs the trailing average *plus* a deliberate seasonal overlay; a neighborhood room with steady regulars can schedule straight off the smoothed baseline. The rule of thumb: if any weekday's gross profit routinely moves more than about 20% week to week, do not trust a single-month average — extend to six months and layer manual bumps for known events.

Then weigh crew shape. A deep bench of thirty part-timers filling fifteen slots is a different scheduling problem than eight full-timers. Deep benches make availability collection the bottleneck, and they punish per-user software pricing. Small senior crews make cross-training the lever — one person who can flip from cold prep to register collapses two half-slots into one full one.

A practical sequencing note: do not shop for software before you have settled the per-employee target. Every scheduling product on the market gets sharper the instant you hand it a real number, and every one of them is a guess-amplifier if you don't. The tool executes the method; it does not supply it. Operators who buy first and derive later end up with an expensive interface wrapped around the same habit staffing they had before.

One more selection axis that catches people: POS integration. If you want a system to propose coverage off live sales rather than you re-running the math monthly, you need a scheduler that ingests your POS feed. That capability sits in the mid-tier and up. Lighter tools publish and notify beautifully but will never volunteer that Friday needs fifteen — you supply the count, they supply the logistics. That's often the right trade for a single room.

Setting the target, reading the days, and what it costs you

Setting the per-employee daily target. Get your chef and front-of-house lead in the same room and settle on one number: the gross profit one average employee should generate in one average shift. For a scratch-cooking vegan kitchen where prep is heavy and check averages run lean, a starting floor around $140 a day is sensible. Many rooms land somewhere between $100 and $175 depending on check average, prep intensity, and how much of the menu is made in-house versus bought in. Frame it as a floor you refuse to schedule below, not a target anyone should coast toward. The people who intend to earn here clear it on autopilot and then turn one more four-top or sell one more cold-pressed juice.

How to back into it if you're starting cold: take trailing gross profit for a normal month, divide by the number of employee-days you actually staffed that month. That gives you your *current* per-employee gross profit. If that number is $95 and your margin is thin, your floor is not $95 — it's what you need it to be, and the gap tells you how much overstaffing you're carrying.

Reading gross profit into days of the week. Average day by day across a trailing three to six months so one freak Tuesday can't warp the plan. Say a normal Monday runs $840 and a normal Friday runs $2,100. Divide each by the $140 floor: Monday wants 6 bodies, Friday wants 15. Six people each honestly clearing $140 exactly account for the $840 the room genuinely produces, and any table they turn faster is upside. Grind that division through all seven days and the roster essentially writes itself — no favorites, no "we've always run eight."

Anchoring to the ticket clock. Headcount is half the answer; timing is the other half. Pull hourly sales and watch where tickets cluster. A vegan room living on a brunch wave and a dinner rush should be staffed for a muscular open, thinned through the mid-afternoon dead zone, and reloaded for a heavy close — not parked in a uniform block. Fifteen Friday bodies does not mean fifteen people from 10 a.m. to 10 p.m.; it means fifteen *employee-shifts* distributed across the day, which might be four on prep at 7 a.m., nine overlapping through the dinner push, and two closing.

Timelines. Expect roughly this arc: week one to pull and clean the data; week two to settle the target and derive the grid; weeks three through six to run the new schedule and watch what breaks. Do not judge the method on one week. Give it a full month before you conclude the numbers are wrong, because the first two weeks surface scheduling *execution* problems (availability conflicts, someone who can't actually run the hot line alone) that masquerade as math problems.

Expected impact. The realistic win is not dramatic — it is the elimination of the two or three chronically overstaffed shifts nobody had a reason to question, plus the confidence to add coverage where the money justifies it. If you find you were carrying two extra employee-days a week at typical hourly rates plus payroll burden, the annualized recovery is real money for an independent room. Understaffing costs are harder to quantify but more dangerous: ticket times stretch, servers stop upselling because they're triaging, and reviews mention the wait. The floor protects you from the first failure; the division protects you from the second.

Software costs, if you decide you want them. Restaurant-specific schedulers commonly price per location per month, in the rough range of $25 to $80 depending on tier, with free single-location tiers available from several vendors. Generalist hourly-workforce apps typically price per user per month in the low single digits, climbing as you add attendance and labor-cost modules. Enterprise workforce-management platforms sell by custom quote and are architected for multi-unit groups with dedicated ops staff — overkill for one dining room. The structural point: per-location pricing wins when you carry a deep part-time bench (thirty people filling fifteen slots), per-user pricing wins when your crew is small, senior, and stable. Verify current pricing directly with each vendor before budgeting; tiers change.

Rolling it out, handing it off, and keeping it honest

The math is the easy part. Making it survive a manager change is the actual work.

Write the method down. One page: the per-employee target, the date it was set, who set it, the trailing window you average over, and the derived headcount by day. Post it where the schedule gets built. A number that lives only in the owner's head reverts to habit staffing within one manager transition — this is the single most common failure mode.

Publish two weeks out. Derived headcounts are only useful if people can plan around them. Fourteen days of visibility dramatically reduces the call-outs and last-minute swaps that force you to overstaff defensively. Many jurisdictions now have predictive-scheduling or fair-workweek rules that mandate advance notice and pay penalties for late changes — check your local requirements, because they materially change the cost of a sloppy schedule.

Separate the count from the assignment. The division tells you six. Your chef decides that six means two hot line, one cold prep, two front, one floater. Keep those two decisions distinct in the handoff, because when a shift goes badly, you need to know whether the *count* was wrong or the *mix* was. Conflating them is how good methods get abandoned after one bad Saturday.

Instrument the feedback loop. After each week, capture three things per shift: derived headcount, actual headcount, and whether the shift felt over-, under-, or correctly staffed according to the person who ran it. Three or four weeks of that log tells you more than any forecast. If Thursdays consistently read "understaffed" at the derived count, your Thursday gross-profit average is probably being dragged down by an unrepresentative stretch — or Thursday's covers are unusually prep-heavy relative to revenue.

Refresh cadence. Re-pull the day-of-week gross-profit averages monthly and re-derive the grid whenever any day moves more than roughly 10%. The per-employee target itself is stickier — revisit once or twice a year, or immediately after a menu-price change, a food-cost shock, or a real shift in check average. Moving the floor too often turns a shared standard back into a moving target nobody can plan against.

Handoff to a new manager. Walk them through one full derivation live — pull the data, do the division, build one day's shift. Do not hand over a finished grid and expect it to be maintained; hand over the *procedure*. Then have them derive the next month themselves while you watch. The method is simple enough that this takes an hour, and that hour is the difference between a discipline and a document.

Adjacent workflows this feeds. Once headcount is derived rather than guessed, several downstream things get easier. Hiring plans become arithmetic: if your derived weekly employee-days exceed what your current roster can cover at reasonable hours, you know exactly how many heads you're short and on which days. Cross-training priorities become obvious — the slot you keep struggling to fill is the skill to train for. Prep scheduling follows too, since a heavy-prep vegan menu means some of your derived headcount has to land *before* service, and the ticket clock won't show you that; your prep lists will. And when you eventually negotiate a lease renewal or model a second location, per-employee gross profit is one of the few operating numbers that transfers cleanly between rooms.

What to watch for. Two failure signals. First, if managers start rounding the derived count up "just in case" every week, the floor is set too low and people don't trust it — re-derive it together rather than fighting the rounding. Second, if a shift consistently clears far above the target with the derived headcount, that's not a win to bank quietly; it's evidence the room could support more coverage and probably more revenue. The division cuts both ways, and treating it as a cost-cutting tool only is how you leave money on the table.

Related questions

How do I handle a shift where headcount comes out as a fraction?

Round to the nearest whole body, then check the ticket clock. A derived 6.4 often means six full shifts plus one short mid-shift covering the peak two hours — that partial is usually where the extra 0.4 actually lives, and scheduling it as a full shift reintroduces the overstaffing you were removing.

Does this method work for a counter or takeout concept?

Yes, unmodified. The division never changes: that day's gross profit over your per-employee target returns the headcount. A grab-and-go counter, a full-service dining room, and a delivery-heavy kitchen run identical arithmetic. Only the menu behind it and the daily gross-profit averages you feed in differ.

What about salaried managers — do they count against the target?

Generally no. Keep the target measured against hourly, schedulable employees, since those are the bodies the division is actually deciding about. Managers are fixed overhead in this model. If a manager routinely works the line, count that portion of their time as an employee-shift.

How does this interact with tip pooling and minimum-wage rules?

The headcount math is separate from compensation structure, but the two collide in practice: adding a body to a pooled shift dilutes everyone's tips. Communicate derived increases clearly, and check local tip-credit and minimum-wage rules before assuming a schedule change is cost-neutral to staff.

Can I use the same target across two locations?

Only if the rooms are genuinely comparable in check average, menu, and prep intensity. Two vegan restaurants with different concepts should carry different floors. Derive each independently, then compare — the gap between them is one of the more useful diagnostics you'll get.

FAQ

How do I set the daily gross-profit-per-employee target for a vegan restaurant?

Start from two facts already in your books — trailing gross profit and current employee-days staffed — and back into the honest daily floor an average employee should clear. Many vegan rooms settle somewhere between $100 and $175 a day depending on check average and how prep-heavy the menu runs. Lock it in jointly with your chef and front-of-house lead so it reads as a shared standard rather than one manager's invention, and re-open the number once or twice a year as covers and costs move.

What if my gross profit swings a lot week to week?

Lean on a trailing three-to-six-month average by day of week to flatten the noise, and schedule to that smoothed baseline rather than to last week's outlier. For spikes you can see coming — a holiday, a payday, a neighborhood event, a large booking on the sheet — layer a deliberate manual bump on top of the calculated count instead of letting one wild week drag the whole average out of shape.

Why staff to gross profit instead of covers or a fixed headcount?

Because neither cover counts nor "we've always run eight" ever paid a payroll — margin does. Pinning headcount to gross profit guarantees every scheduled body is backed by real dollars of margin, and it forces the uncomfortable conversation about which shifts are actually earning the coverage you keep giving them out of habit. It also makes the reverse case: shifts that clear well above target can probably support more staff and more revenue.

How often should I recalculate my per-shift headcounts?

Refresh the underlying day-of-week gross-profit averages monthly, and re-derive the headcount grid whenever those averages move by more than roughly ten percent on any given day. The per-employee target itself is stickier — revisit it once or twice a year, or immediately after a real menu-price change, a food-cost shock, or a shift in your check average.

Do I need scheduling software to run this, or will a spreadsheet do?

A spreadsheet runs the arithmetic fine — the division is one formula. Software earns its cost on execution: publishing to phones, collecting availability, handling swaps, enforcing break and overtime rules, and tying the schedule to POS sales. Start with the spreadsheet, confirm the method holds at your room for a month, and buy tooling only once you know which execution problem you're paying to solve.

What's the fastest way to tell if I'm currently overstaffed?

Compute your actual per-employee daily gross profit for last month — trailing gross profit divided by employee-days staffed — and compare it to the floor you'd set on purpose. If the actual is well below your intended floor, the gap is your overstaffing, and it usually concentrates in two or three specific shifts rather than spreading evenly across the week.

Sources

• U.S. Bureau of Labor Statistics, Food Services and Drinking Places industry data — https://www.bls.gov/iag/tgs/iag722.htm
• U.S. Small Business Administration, manage your finances guidance — https://www.sba.gov/business-guide/manage-your-business/manage-your-finances
• National Restaurant Association — https://restaurant.org/
• U.S. Department of Labor, Wage and Hour Division (FLSA, tipped employees, overtime) — https://www.dol.gov/agencies/whd
• IRS, Cost of Goods Sold guidance (Publication 334) — https://www.irs.gov/publications/p334
• SCORE, free small-business mentoring and templates — https://www.score.org/
• Cornell University School of Hotel Administration research center — https://sha.cornell.edu/faculty-research/centers-institutes/chr/
• U.S. Census Bureau, Monthly Retail Trade and Food Services survey — https://www.census.gov/retail/index.html

Related on PULSE

• [How Many Employees Should I Schedule Each Shift at My Korean BBQ Restaurant?](/knowledge/tl0455)
• [How Many Employees Should I Schedule Each Shift at My Hot Pot Restaurant?](/knowledge/tl0454)
• [How Many Staff Should I Schedule Each Shift at My Sushi Restaurant?](/knowledge/tl0094)
• [How Many Employees Should I Schedule Each Shift at My Pottery Studio?](/knowledge/tl0481)
• [How Many Employees Should I Schedule Each Shift at My Ice Cream Shop?](/knowledge/tl0072)
• [How Do I Know How Many Cooks and Servers to Schedule Each Shift at My Pizza Restaurant?](/knowledge/tl0003)