How Many Employees Should I Schedule Each Shift at My Frozen Yogurt Shop?

The job this staffing math is hired to do

Most frozen yogurt owners schedule from memory. Four people on Saturday because there were four last Saturday. Two on Tuesday because two feels right. The schedule becomes a habit rather than a decision, and the habit outlives the conditions that created it — the shop opened next to a middle school that has since moved, the neighboring gym closed, a competitor opened three blocks over, and the grid never changed.

The job this method is hired to do is to convert an argument into arithmetic. When your shift lead says Saturday needs six and you say Saturday needs four, there is no way to settle it except by seniority or volume. When you both agree that a counter employee should generate about $150 of gross profit in a shift, and the Saturday 6-to-9 p.m. block has averaged $1,050 of gross profit over the last four months, the answer is seven and neither of you had to win an argument to get there.

Start with the per-employee number, because everything else divides by it. Sit down with whoever helps you run the shop and settle on the gross profit an average counter employee should produce on an average shift. Not your best closer on the best night. Average. In self-serve froyo, where the product carries a strong per-ounce margin but tickets run small and volume runs high, that floor tends to land in the $100 to $200 band. Call it $150 as a working figure and adjust once you see real data.

Say the number out loud to the crew, in plain language: if you show up, keep the toppings bar stocked and clean, weigh and ring at a normal pace, and give people a normal amount of attention, you should be producing no less than $150 in gross profit on your shift. That is the floor, not the ceiling. The employees who want more hours and a path to shift lead do not coast to $150 and lean on the register — they hit it during the ordinary rush and then dig for the next $150 by mentioning the loyalty punch card, suggesting the second topping, pointing at the larger cup.

Then pull the actual data. Break your operating day into dayparts — say, open-to-3 p.m., 3-to-6 p.m., 6-to-close — and average each daypart's gross profit by day of week across a trailing three to six months. Not revenue: gross profit, meaning sales minus the cost of the yogurt mix, toppings, cups, spoons, and napkins that went out the door. Most POS systems will export this or get close enough that you can back into it with a blended cost-of-goods percentage.

Divide, and the plan writes itself. Saturday 6-to-9 p.m. at $1,050 divided by $150 is seven. Tuesday 1-to-4 p.m. at $300 divided by $150 is two. Sunday afternoon at $600 is four. Do that for every daypart and every day, and you have a grid built on receipts rather than on who is friendly with the shift lead. The favoritism problem solves itself as a side effect — nobody can stack their friends onto the good weekend nights when the count is a division problem, and nobody gets stranded alone on a Friday night that clearly needed four.

Two adjustments keep it honest. Round up on any shift the math prices below one — a 0.4 result on a slow Monday open still requires a human to unlock the door, run the machines through startup, stock the bar, and serve the six people who wander in. And separate your outliers before you average. A holiday, a hot-weather Saturday, a first-day-of-summer-break Friday, the night the high school won something — those belong in their own bucket. Blending a $2,000 event night into your ordinary Saturday average inflates every normal Saturday and understaffs the actual event.

How it fits the RevOps stack

Frozen yogurt sounds like an odd place to invoke RevOps, but the underlying discipline is identical to what a revenue operations team does with a sales floor. RevOps sets a per-head productivity target, divides the pipeline requirement by it to size the team, and then holds capacity against the demand curve. Swap "quota per rep" for "gross profit per shift" and "territory coverage" for "daypart coverage" and the machinery is the same. A scoop shop is a very small revenue operation with a very short sales cycle.

That framing matters practically, because it tells you which systems have to talk to each other. The staffing number is an output, and it is only as good as the three inputs feeding it: point-of-sale data, cost-of-goods data, and the scheduling surface where the number becomes an actual published shift. Break any one of those links and you are back to guessing.

The POS is the source of truth for both sales timing and sales volume. You need it to export transactions with timestamps, not just daily totals, because the count tells you how many and the timestamps tell you when. Cost of goods can live in the POS if you have recipe-level costing, or it can be a blended percentage you calculate quarterly from invoices — either works, as long as you apply it consistently. The scheduling app is downstream of both; it publishes the grid, handles availability and swaps, and tracks whether the labor you scheduled is the labor that actually clocked in.

The feedback loop at the bottom is the part most shops skip. You publish the schedule, the week happens, and nobody goes back to check whether the plan survived contact with reality. Compare scheduled hours to clocked hours weekly. If people are consistently clocking in early and out late, your shifts are mis-sized rather than your headcount. If labor as a share of sales keeps drifting above target on specific blocks, either the demand forecast for those blocks is stale or the per-employee number is set too low for what your crew can actually do.

Downstream effects show up in places you might not connect to scheduling. Understaffed rush blocks produce longer topping-bar lines, which produce abandoned visits, which suppress the very sales data you use to size the next schedule — a self-fulfilling shrink where a block looks slow because it was miserable to shop. Overstaffed slow blocks produce bored employees and, over a season, resentment about hours being cut later. The math is not just a cost-control exercise; it protects the demand signal you rely on.

The same stack shape works in adjacent operations. An ice cream shop, a boba tea counter, a cookie or donut shop, a smoothie bar, a seasonal shaved-ice stand — all of them run small tickets, spiky demand, high margins on the base product, and a workforce heavy on students. A car wash or a small-format bakery runs a comparable curve with different peaks. The variable that changes between them is the per-employee target, not the method.

Pricing, engagement models, and typical ranges

The calculation itself is free — it is a division problem you can do in a spreadsheet. What costs money is the layer that publishes, tracks, and enforces the result. Scheduling software in this category prices two ways, and the difference matters enormously for a shop with a seasonal bench.

Per-user pricing charges by headcount. Per-location pricing charges a flat rate regardless of how many names are on the roster. A frozen yogurt shop that carries twelve people through winter and twenty-six through the summer surge pays a materially different bill under those two models, and the surge months are exactly when cash is tightest relative to the labor you are adding. As a rule, per-location wins for seasonal operations with a big rotating bench; per-user can win for a small, stable crew of six or eight who work year-round.

Several tools in this space offer a genuinely usable free tier for a single location — enough for scheduling, a mobile time clock, and basic team messaging. For one shop, that is often sufficient. Paid tiers generally start in the low-to-mid tens of dollars per location per month and climb as you add labor forecasting, POS integration, compliance tooling, and payroll export. Per-user plans commonly start in the low single digits per person per month and roughly double or triple when attendance and labor-cost features are included. Enterprise restaurant platforms are typically custom-quoted and aimed at multi-unit groups with dedicated operations staff — they are overbuilt for a single scoop shop and priced accordingly.

Vendors change pricing frequently, so verify current numbers on the vendor's own pricing page before committing. Treat any figure you read in an article, including this one, as a rough shape rather than a quote.

Beyond the software line, the real cost decision is your labor target as a percentage of sales. Quick-service dessert operations commonly run labor somewhere in the mid-to-high twenties as a share of sales, though this swings hard with local wage floors, whether the owner works the counter, and how seasonal the concept is. That percentage is the guardrail on the whole exercise: if your gross-profit division keeps producing headcounts that push labor well past your target, the target per employee is set too low, or your pricing is too low, or your slow dayparts should not be open at all.

That last option is worth taking seriously and rarely gets considered. If your Monday-through-Wednesday 11 a.m.-to-2 p.m. block averages $180 of gross profit, the math says one employee, and that one employee costs you more in wages and payroll burden than the block generates. The honest answer is not to staff it thinner — it is to open at 2 p.m. on weekdays. Trimming unprofitable hours is a staffing decision even though it looks like an operating-hours decision, and this method surfaces it clearly.

Engagement models for help are worth naming too. Some owners run this entirely in-house with a spreadsheet and a POS export. Some lean on a bookkeeper or fractional operations consultant to build the daypart model once and then maintain it themselves. Multi-unit groups sometimes hire a labor analyst or buy an enterprise platform that enforces budgets store by store. The in-house spreadsheet path costs nothing but your time and is genuinely adequate for one or two locations — the value of the paid tooling rises sharply once you cannot personally see every store's schedule.

How to evaluate and shortlist a scheduling tool

Once you have the headcount math, the tool question narrows. You are not shopping for something to tell you how many people to staff — you already know. You are shopping for the layer that publishes the grid, manages a crew of students with volatile availability, and tells you honestly whether the plan held.

Start by separating method from logistics. Most scheduling apps are logistics: drag-and-drop grids, availability collection, shift swaps, mobile clock-in, copy-last-week-forward. That is genuinely valuable and it is most of what you need day to day. A smaller set of tools adds demand awareness — they connect to your POS, project sales for upcoming blocks, and suggest staffing against a labor-percentage or sales-per-labor-hour target. Those overlap with what you are doing manually and can automate the refresh. Neither category replaces the decision you made about the per-employee floor.

Then run through the criteria that actually differ between products:

Sales and labor awareness. Does it ingest POS sales so you can see labor as a percentage of sales for each shift, live, during the shift? For a spiky dessert concept this is the difference between finding out on Monday that Saturday ran hot and cutting someone loose at 8 p.m. when the line dies.

Pricing model against your bench size. Count your peak-season roster, not your current one. Multiply it out under both models for a full year including the off-season, when you will still be paying for the seats.

Minor-labor compliance. If most of your counter staff are sixteen and seventeen, break rules, maximum daily and weekly hours, and school-night restrictions are not optional and vary substantially by state. A tool that flags a violation before you publish is worth real money compared to one that finds out after the fact. Fair-workweek and predictive-scheduling ordinances apply in some cities and carry advance-notice requirements with penalties attached — check whether yours is one before you build a habit of Thursday schedule changes.

Mobile experience for the crew. Your staff will interact with this on a phone, at school, between classes. If swaps and availability updates are painful, you will end up managing the schedule by text message anyway and the tool is dead weight.

Data export. You want to be able to pull scheduled versus actual hours and reconcile against sales without retyping anything. If the export is locked behind a higher tier, factor that into the price.

Shortlist three, run each on a real week in parallel with your current process, and judge them on one question: at the end of that week, could you tell whether the schedule was right? A tool that publishes beautifully but cannot answer that leaves you exactly where you started.

One practical sequencing note — build the daypart model in a spreadsheet first, before you buy anything. Two hours of POS exports and a pivot table will tell you your real demand curve, and that curve is what you use to evaluate whether a tool's forecasting is actually smarter than your own arithmetic. Buying first means you inherit whatever the vendor's default assumptions are and never learn your own shop's shape.

Buyer decision framework

The decision has fewer branches than the vendor landscape suggests. Most of it comes down to how many locations you run, how big your seasonal bench gets, and whether you want the tool to do the demand math or just publish what you decided.

Two failure patterns are worth naming because they cost more than the wrong software choice.

The first is treating the calculated number as a rigid rule. The formula sets floors and exposes bad habits; it does not know that your machine needs a forty-minute clean-and-refill that has to happen before the evening wave, that a new hire's first three shifts are half-speed, or that your Saturday lead can genuinely run a rush that would take two of anyone else. Adjust for those knowingly. What you should never do is adjust for them unknowingly, which is what scheduling from memory amounts to.

The second is raising the per-employee target on paper to make the labor line look better. If you move the floor from $150 to $200 without any change in pace, station layout, or ticket size, you have not improved productivity — you have just published a schedule that will understaff every rush and produce the long lines that suppress next quarter's sales data. Raise the target only when you have changed something real: a second register, a better topping-bar layout, a genuine upsell habit, a price increase that lifts margin per cup.

Where the target legitimately moves is with the shop. A location with a $2 higher average ticket, or a drive-through window, or a wholesale side business supports a higher floor. A tiny mall kiosk with one register and a four-foot counter physically cannot clear $150 per person during a crush no matter who is working, because the constraint is the space, not the staff. Set the number against your actual geometry.

New shops without trailing data should estimate and correct fast. Use the first three or four weeks of real receipts, accept that the numbers will be noisy, and recalculate every two weeks until the curve stabilizes — usually within a season. An imperfect model built from your own receipts beats a confident guess, and it improves on a schedule while the guess never does.

Related questions

How do I set the per-employee gross-profit target if I have no history?

Use your first three to four weeks of receipts, apply a blended cost-of-goods percentage from your invoices, and divide by the average bodies on the floor. Recalculate every two weeks. Expect the number to move until you have a full season.

Should the formula include the shift lead or manager?

Count anyone whose hours hit the labor line and who serves guests. A working shift lead counts. A salaried owner who only opens and closes generally does not — track that cost separately so it does not distort your per-employee floor.

Does this work for a mall kiosk or a seasonal trailer?

Yes, with a lower target. Physical constraints — one register, a short counter — cap throughput regardless of staffing. Set the floor against what your layout can actually produce, then keep the same division method for each block.

How often should I rebuild the daypart averages?

Quarterly for a mature shop, and always before a seasonal turn. Weather-driven concepts shift demand curves fast, so a summer grid built on spring data will understaff June evenings within two weeks of school letting out.

What if the math says fewer people than I need to open?

Round up to one and treat the block as a service floor, not a math result. If that block loses money every week, the real answer is trimming operating hours rather than staffing an unprofitable window thinner.

FAQ

What if my gross profit per shift comes in well below $150 per employee?

That is a signal to look upstream rather than at the schedule. Check pricing per ounce, portion and cup sizing, waste from mix left in the machine at close, and whether traffic in that block is genuinely thin. The $150 figure is a working floor for quick-service frozen yogurt, and the honest range runs roughly $100 to $200 depending on your market, rent, and season. The method works at any target — it only breaks when the target is aspirational rather than real.

Can I use the same formula for part-time and seasonal staff?

Yes. The division does not care about employment classification. A four-hour shift simply carries a smaller slice of the day's gross profit, so it produces a smaller headcount. Where you do need to adjust is speed: a brand-new seasonal hire in week one is not producing at the same rate as someone who has worked two summers, so plan an overlap during the ramp and expect the first few weeks of a seasonal wave to run slightly heavier than the math suggests.

How should I handle holidays, heat waves, and local events?

Average them separately and never blend them into your ordinary day-of-week numbers. A hot-weather Saturday or a post-game rush can multiply a block's gross profit several times over, which would call for double or more the usual headcount. Pull last year's comparable date if you have it. Blending an event night into your standard Saturday average does double damage — it inflates every normal Saturday and still leaves you short on the night that actually spikes.

What if the formula asks for more people than I have on the roster?

Prioritize the highest-gross-profit blocks first and let the thin blocks run lean, since that is where a missing body costs you least. Cross-train so anyone can cover register, topping-bar restock, and machine duty. Do not simply raise the per-employee target to make the shortage disappear on paper — that relocates the understaffing instead of solving it. A genuine capacity gap is a hiring problem, and the math has just told you how many people to hire.

Does this method apply outside frozen yogurt?

It transfers cleanly to any small-ticket, high-margin, demand-spiky counter business — ice cream, boba, smoothies, cookies, donuts, coffee, shaved ice. The structure holds: set a per-employee gross-profit floor, divide each daypart by it, place the bodies where the receipts ring. The per-employee number is what changes, because average ticket and throughput differ substantially between a $6 froyo cup and a $5 coffee with a two-minute service time.

How do I keep the schedule from drifting back to habit?

Put a recurring review on the calendar — monthly is enough for most shops — where you pull the trailing daypart averages fresh and recompute the counts before publishing the next cycle. Habit creeps back through small overrides that never get revisited. Comparing scheduled hours to clocked hours each week, and labor percentage to target by block, catches drift while it is still one or two shifts wide.

Sources

• https://www.bls.gov/oes/current/oes353023.htm — Bureau of Labor Statistics wage data for fast food and counter workers
• https://www.dol.gov/agencies/whd/youthrules — U.S. Department of Labor youth employment rules and hour restrictions for minors
• https://www.sba.gov/business-guide/manage-your-business/hire-manage-employees — Small Business Administration guidance on hiring and managing employees
• https://restaurant.org/research-and-media/research/ — National Restaurant Association industry research and operations data
• https://www.7shifts.com/ — restaurant and quick-service scheduling platform, pricing and POS integrations
• https://www.joinhomebase.com/ — scheduling and time-clock platform with per-location pricing
• https://www.deputy.com/ — demand-based scheduling and compliance tooling
• https://wheniwork.com/ — shift scheduling and mobile clock-in for hourly teams
• https://www.irs.gov/businesses/small-businesses-self-employed/depositing-and-reporting-employment-taxes — IRS employment tax obligations for employers
• https://www.fourth.com/ — enterprise restaurant labor forecasting and scheduling

Related on PULSE

• [How do I forecast demand by daypart for a small retail shop?](/knowledge/tl21653)
• [What labor percentage should a quick-service dessert shop target?](/knowledge/tl21652)
• [How do I cross-train a seasonal counter crew?](/knowledge/tl21651)
• [What POS data do I need to schedule against sales?](/knowledge/tl21650)
• [When should I cut operating hours instead of cutting staff?](/knowledge/tl21649)