Designing a catering menu for preparation and delivery capacity
Operational worked example · Small off-premise caterer from an existing DBPR-licensed shared commercial kitchen
Design the menu around the paid preparation window and the delivery promise, then test guest counts against each separate limit. A menu can have a positive food margin while exceeding room-hours, finishing labor, hot/cold storage or one van’s loading schedule. This worked buffet uses 14.0 preparation person-hours and 6.5 kitchen reservation hours before service. Those are worksheet assumptions to measure, not a tested production rate.
Read the complete operating guide set · Browse operating articles
Start with the promised event and guest guarantee
Define who buys the meal, when it must reach the venue and what the team does after arrival. A drop-off lunch, staffed buffet and made-to-order station have different loading and labor demands even when they serve the same guests. Record the final headcount date, dietary requirements, portion specifications, serving time, venue access and return procedure. Price the actual promise instead of assuming every guest simply adds one identical meal.
The linked reference case combines 60-guest drop-offs at $28.00 and 100-guest buffets at $48.00 per guest. It assumes shared licensed kitchen infrastructure and owned carriers, with event tableware rented separately. No real venue diary or timed menu trial establishes those assumptions. Florida’s catering guidance requires preparation and storage within the licensed food-service scope; individual retail made-to-order sales are a different operation. DBPR boundary
Choose dishes that can share measured preparation components and survive the actual holding and journey. Dietary variants need labeled processes, ingredients and packing capacity of their own. Do not describe a common sauce or utensil as automatically suitable for every dietary request. The case specifies no universal transport time, temperature guarantee or allergen certification; the actual food process must be reviewed with the responsible kitchen and authority.
Measure purchased quantities and usable portions
The staffed menu’s assumed ingredient value is $14.00 per guest before usable yield. At 90.0% yield, purchased food becomes $15.56 per served guest. Moving yield to 84.0% raises that to $16.67 and adds $111.11 across the same 100-person event before any price change. This arithmetic isolates preparation loss; it is not a supplier ingredient quote.
Keep a recipe register with purchased weights, batch output, usable yield, portion size and actual discard. Separate trimming, cooking shrinkage, remakes and unsold prepared food, then choose the factor that matches the model’s single yield denominator. Counting the same waste inside recipe cost and again as a generic surcharge exaggerates the expense. Counting ideal yield while omitting leftovers creates the opposite error.
Test a mixed menu, not only an isolated tray. Shared ingredients can reduce receiving and setup tasks, but more options may require extra cold space, pans and labeling. A food-cost saving that adds a finishing person or blocks an oven cycle can reduce contribution after labor. The buffet’s $1,752.04 event contribution pays fixed work and overhead; it cannot all be allocated to expanding the menu.
Keep person-hours separate from kitchen hours
| Constraint | Simple menu | Menu with added hand finishing | Meaning |
|---|---|---|---|
| Prep person-hours | 14.0 | 18.2 | All active work added across people |
| Reserved kitchen hours | 6.5 | 8.6 | Added finishing shared by two prep people in this comparison |
| Available trial reservation | 8.0 | 8.0 | One hypothetical day window |
| Incremental paid flex prep | No extra menu task | $95.83 | Only if fixed productive time is already full |
The simple menu uses four setup person-hours plus the case’s per-guest prep work. An article-local finishing step takes 2.5 minutes per guest, adding 4.2 person-hours. With two people sharing that step, the comparison extends room occupancy beyond the 8.0-hour reservation. Two paid people reduce some elapsed work, but they do not create another reserved kitchen station or unlimited refrigerated storage.
This is a local worksheet, not a replacement input revision. The executable monthly engine keeps its original per-guest preparation coefficient. A real slow item requires updating both prep-person and kitchen-room coefficients, checking the owner’s usable time and applying the current wage/roster. The assumed flex wage is $20.00 plus 15.0% burden; historical Orlando BLS wages are comparison evidence rather than hiring offers. Local wage scope
Count receiving, washup, loading, breaks, returns and cleaning inside the paid roster. The monthly model already pays the lead cook and owner during quiet periods. Do not charge their prep twice as a variable hourly invoice while also retaining the full fixed salaries. Flexible help is an added cost when productive fixed time has been used, and a staffing step requires a real person available on the promised date.
Fit batch size to carriers and separate cold products
For an exact packing example, assume three hot components, each with 100 portions, and 25.0 usable portions per shallow full-size pan. That requires 12 hot pans, or 2 selected carriers. The manufacturer specifies six such pan positions per UPC400; the chosen four units provide 24 positions. The portion yield is our assumption, while the carrier positions are the stated manufacturer scope. Carrier specification
A cold salad, dessert or special portion does not automatically fit into the unused hot positions with the same temperature control. The startup budget has separate cold carriers, but their exact capacity and cooling method still need confirmation. Two events may fill every nominal hot position while also competing for the same return, washing and reload window. Count clean backup pans and loaded weight rather than treating every slot as reusable immediately.
Track equipment occupancy as well as hand work. In an illustrative oven test with 40.0 portions per batch and 30.0 minutes per cycle, three batches are needed for 100 portions. That occupies the selected oven for 90 minutes before handling and interruptions, even when the cook works elsewhere during a cycle. The case makes no manufacturer throughput claim; verify rack fit, recipe cycle, holding and shared users at the actual host kitchen.
Test a busy day before accepting the whole month
The model’s monthly calendar assumes 60.0% of staffed requests use 4 prime slots. Accepted staffed bookings therefore stop at six per month before forecast cancellations. It also limits owner hours, total events, preparation and kitchen occupancy. Those monthly rules protect the projection, but they cannot prove that particular venue times and kitchen reservations are compatible.
A single drop-off plus one staffed buffet consumes 11.6 separate reservation hours under the unchanged task coefficients. That exceeds the article’s eight-hour day even though both events can fit a month’s unused capacity. Shared batching might change the work, but its saving must be measured and reflected in the reservation, sanitation and staff plan. A second van removes only a transport constraint; it does not repair the prep window.
Put every booking on a dated sequence: receiving, preparation, holding, loading, arrival, service, clearing, return and washing. Record who owns the delivery vehicle and who remains at the kitchen. The owner cannot drive to one venue and simultaneously supervise another with the same budgeted hours. Test missed-access and late-return cases before selling a second overlapping date.
Run the mixed-menu trial that changes the forecast
| Measure | Record | Model revision if it differs |
|---|---|---|
| Demand promise | Guest guarantee, variants and timing | Event mix, price and scope |
| Recipe yield | Purchased batch, usable portions and discard | Ingredient/yield coefficient |
| Work | Prep person-hours, paid roster and lost time | Fixed/flex capacity and wages |
| Kitchen | Reserved room time and station occupancy | Available hours and booking cap |
| Packing | Pans, hot/cold separation, loaded weight | Portable kit and transport specification |
| Diary | Loading/arrival/service/return times | Peak slots, owner time and vehicle needs |
Keep failed trial periods as well as successful ones. The useful menu is one the paid team can make and deliver with the promised food process and a positive contribution after added work. A higher guest price cannot physically shorten a late finishing step or create another reserved slot.
Use the unit-economics guide to reconnect the measurements to fixed cost and event contribution, then the startup-cost guide to price any kit or vehicle change. Deposit and supplier timing belongs in the cash model after the menu scope is settled. A larger reserve funds uncertainty temporarily; measured capacity determines what can actually be promised.
Sources and scope
- Educational catering case assumptions · Checked 2026-10-05 · Authored case inputs, not observed market averages or a real operator.
- Florida DBPR: Guide to Catering · Checked 2026-10-05 · Florida catering from an existing DBPR-licensed shared kitchen; no individual made-to-order retail sales or home preparation.
- BLS: Orlando employment and wages, May 2025 · Checked 2026-10-05 · Orlando-Kissimmee-Sanford metro, May 2025 occupation means; not 2026 recruitment quotes.
- Cambro UPC400 insulated carrier · Checked 2026-10-05 · UPC400, non-electric insulated transport; US manufacturer list price, distributor price may differ.