How do you start a corporate catering business in 2027?

The Wednesday that decides whether you have a business

Picture a 180-person software company eight minutes from your kitchen. Their office is a ghost town Monday and Friday, but Tuesday through Thursday it fills, because leadership tied in-office attendance to the roadmap and to lunch. Their office manager — title is probably Office Experience Manager or People Operations Coordinator — has a budget line for food and a standing anxiety: the last caterer showed up at 12:20 on a day the VP of Sales had a client in the lobby, and two of the boxes labeled "gluten-free" clearly were not.

That single failure is your entire opening. It is the most common trigger by far in this market, because most independent caterers are operationally sloppy — they are cooks who took on logistics rather than logisticians who happen to cook. When you deliver at 11:40 for a 11:45 setup, label every allergen, send a photo of the finished spread to the office manager's phone before you leave, and email a clean invoice the same afternoon, you have not just fulfilled an order. You have demonstrated the only product this buyer is actually shopping for: the absence of failure.

Now do the arithmetic that reframes the whole venture. That 180-person account, eating on one anchor day at $18 a head with an 85% attendance rate, is roughly $2,750 a week. Fifty weeks is about $137,000 from one logo, with no cold start, no tasting, no proposal cycle after the first one. Land twelve to twenty of those inside a 15-minute radius and you have a business with the revenue predictability of a subscription company and gross margins closer to a food operation. Land the same twelve spread across a 45-minute metro and you have a fleet problem, burnt-out drivers, and a margin that quietly goes negative on fuel and windshield time.

The contrast case is the founder who opens "a catering business" because they love to cook. That founder takes the wedding, the bar mitzvah, the milestone birthday, the one-off corporate holiday party. Every job is a fresh sale, a fresh menu, a fresh logistics plan, a 45–60 day cycle, and zero reorder. Their calendar spikes in May–June, September–October, and December, and starves in January and July. They have no purchasing leverage because nothing repeats, no labor standardization because every job resets, and nothing sellable at the end because the entire operation lives in their head.

So the framing to hold from day one: this is a recurring-revenue B2B business that happens to produce food. The corporate catering slice of the US market runs roughly $22B–$28B of a total catering industry in the $60B–$72B range, and it is the fastest-growing slice — mid-single-digit to high-single-digit annual growth — precisely because the three-day hybrid anchor week converted sporadic event feeding into standing weekly programs. The demand pattern changed in your favor. Most operators have not changed their playbook to match, and that gap is the opportunity.

How the recurring account machine actually works

The mechanism has five stages, and each one has a specific failure mode you can engineer against.

Stage one is discovery. In 2027 the corporate lunch buyer shops on software. ezCater, Sharebite, Fooda, ZeroCater, Forkable, CaterCow, and Hungry are where an office manager goes when they need a caterer, the same way they go to a vendor portal for anything else. You need to be listed on at least two, with professional food photography — this is not vanity, the photo grid is your storefront and it is the single highest-leverage $1,500–$4,000 you will spend — and you need to hit every service-level commitment perfectly, because marketplace visibility is rating-weighted. Marketplace commission runs roughly 8–18% depending on platform and terms.

Stage two is the trial order. Corporate buyers rarely sign anything first. They test you with one lunch, often 40–120 people, and they watch three things: did it arrive inside the window, were dietary needs handled and clearly labeled, and did the setup look presentable without them having to fix it. Your job on a trial is to over-execute on the boring parts. Send the setup confirmation photo. Leave the labeling immaculate. Follow up the next morning with a short note and a reorder link.

Stage three is the reorder pattern. Watch for the third order. Two orders is a coincidence; three is a habit forming, and it is the moment to make the direct pitch. The conversation is not a hard sell — it is a convenience offer: a standing weekly slot, a dedicated contact, a rotating menu built for their dietary profile, and a per-head rate $1–$3 below what they pay on the marketplace.

Stage four is the standing-order agreement. Keep it to one page so the office manager can sign it without routing to legal — that friction kills more conversions than pricing does. The page specifies cadence (every Wednesday), an estimated headcount with a defined flex band, a final-count lock 24–48 hours before delivery, a per-head rate, a 3–6 month term, net-15 or net-30 terms, cancellation notice, and allergen-handling language your attorney reviewed. That headcount lock clause is not administrative trivia; it is the single line item that protects your food cost from a client's loose guessing.

Stage five is expansion. A converted account grows three ways: additional anchor days (Wednesday becomes Tuesday and Wednesday), add-on attach (coffee service at $3–$7 a head, breakfast on the same delivery at $8–$14, individually packaged allergen-safe meals at a $4–$9 upcharge, dessert platters), and referral — office managers move companies every 18–30 months and they take their vendor list with them.

The reason this machine works is that it inverts the normal catering sales burden. In event catering, revenue decays to zero the moment a job ends and you must resell to survive. In corporate recurring, revenue persists by default and your sales effort compounds on top of a base instead of replacing it. That is the difference between a treadmill and an asset.

Real numbers: startup capital, per-order margin, and the growth curve

Startup cost is dominated by one decision — how you get access to a legal commercial kitchen. There are three paths.

Path A, shared commissary. You rent time in a licensed shared kitchen, typically $18–$35 an hour or $900–$2,400 a month for a recurring block. All-in to open: roughly $12,000–$45,000. That covers smallwares, hot-holding transport (insulated front-loaders, sheet-pan carriers, catering bags), a used cargo van at $8,000–$28,000, packaging inventory, insurance, branding and photography, a website, and software. This is the correct path for the large majority of founders, because it converts what would be fixed cost into variable cost while you are still proving demand. The ceiling: shared kitchens limit hours and storage, so you typically outgrow one somewhere around $400,000–$800,000 in annual revenue.

Path B, small leased production kitchen. A 1,200–2,500 square foot space built out for catering production — hood and fire suppression, walk-in cooler and freezer, prep tables, combi oven, delivery staging area. Total to open: roughly $85,000–$280,000, with lease running $2,500–$8,000 a month. This is a Year-2 decision made against a proven book, not a Year-1 leap of faith. Buildout is where founders overspend most predictably; sourcing used equipment through restaurant liquidators and auctions commonly cuts equipment cost 30–50%.

Path C, ghost kitchen or cloud-catering stall. Roughly $25,000–$70,000 to open, with monthly occupancy in the $2,800–$6,500 range plus whatever revenue-share terms the operator imposes. Faster than a buildout, more capacity than a shared commissary, and a defensible middle option in dense metros where commissary time is scarce.

Now the per-order math you must be able to run in your head. Take a representative recurring drop-off: 150 people at $18 a head, $2,700 in revenue.

• Food at 30% — $810
• Direct labor, prep plus pack plus allocated driver, at 24% — $648
• Packaging and disposables at 4.5% — $122
• Delivery, fuel, vehicle allocation at 6% — $162
• Kitchen rent allocation at 8% — $216
• Software and payment processing at 5% — $135
• Insurance and overhead allocation at 4% — $108

Net contribution lands near $499, roughly 18.5%. Run the identical order through a marketplace at 15% commission and you subtract another $405, collapsing net contribution to about $94 — roughly 3.5%. That one comparison is the strategic spine of the entire business, and it explains why marketplaces are an acquisition channel and a capacity-fill tool rather than a place to live.

The second lever is batch density. If that 150-person order is one of four deliveries on the same route inside a 15-minute radius, your per-order delivery and vehicle cost roughly halves and effective margin climbs toward the low-to-mid 20s. One isolated order across the metro does the reverse. Density, not reach, is the margin variable.

Pricing tiers to anchor against. Drop-off lunch — boxed individual meals or buffet trays, delivered and set up in ten minutes, no staff stays — runs $14–$22 a head with minimums of 10–25 people, and this tier is 70–80% of your recurring revenue. Staffed buffet with one or two attendants runs $26–$48 a head with a 15–22% service fee, minimums of 25–50, typically for all-hands, sales kickoffs, and client-facing days. Full-service plated or stationed events run $55–$130 a head with a 20–28% service charge — holiday parties, board dinners, executive offsites — lower frequency, higher per-job margin, operationally heavy.

Cost-structure targets to run the business against: food 28–34% of revenue, labor 22–30%, packaging 3–6%, delivery and vehicle 4–8%, kitchen occupancy 5–12%, software and marketplace commission 3–8%, insurance 1–3%, marketing 3–6%, and net margin 12–20% once route density exists. Labor below 22% almost always means the owner is absorbing unpaid hours that are not sustainable; above 30% means a scheduling or productivity problem, usually flex-pool mismanagement during the Tuesday-through-Thursday surge that carries 60–70% of weekly volume.

The realistic growth curve. Year 1 with the owner cooking, selling, and delivering: $180,000–$420,000, typically 8–16 recurring mid-size accounts plus a marketplace drip, with modest owner draw for the first six to nine months and 55–70 hour weeks. Year 3 with a chef de cuisine, a dedicated driver, and an ops coordinator: $900,000–$1.8M across 35–70 recurring accounts, 40–55 hour weeks, and the owner functioning as a sales-and-accounts CEO. Year 5 ceiling before a structural decision: $3M–$7M. At that point a systematized operator with documented processes has real exit optionality — regional roll-ups and strategic buyers price businesses like this in the neighborhood of 0.5–0.9x revenue or 4–7x seller's discretionary earnings, usually structured with a cash majority plus a seller note and an earn-out tied to account retention.

Sizing your actual market is a drive-time exercise, not a national one. Count establishments in the 50–600 headcount band inside your radius using county business-pattern data or a list provider, multiply by a plausible annual catering spend per establishment, and you have a defensible number. A single production kitchen with one delivery team realistically serves $2.5M–$5M on one shift and $4M–$8M on split shifts. Your Year-1 constraint is never market size — it is your own two hands and your pipeline.

Trade-offs: kitchen path, channel mix, and menu scope

Three decisions carry almost all the strategic weight, and each is a genuine trade-off rather than a right answer.

Kitchen path trades fixed cost against capacity ceiling. Path A minimizes downside — if the book never materializes, you walk away having lost tens of thousands rather than hundreds. The cost is a hard capacity ceiling and scheduling friction: shared kitchens allocate hours, and if your commissary has three other tenants competing for the 5am–10am prep window, your surge days get tight. Path B removes the ceiling and gives you control of your own schedule, cold storage, and staging area — but converts your risk profile entirely. A $5,000 monthly lease plus equipment debt means you now need roughly $700,000–$900,000 of annual revenue just to carry occupancy at a healthy ratio. The disciplined sequence is A first, B when recurring revenue makes the lease boring rather than terrifying.

Channel mix trades margin against acquisition speed. Marketplace-heavy launch gets you orders in weeks rather than months and builds a public reputation you can point to, at 3–5% net. Direct-outbound-heavy launch preserves 18%+ margin but has a slower ramp and demands sales work most food founders find uncomfortable — list-building, cold email to office managers, complimentary sample drops. The mature answer is a blend: 60–80% direct recurring, 20–40% marketplace for acquisition and capacity fill. The marketplace share should never reach zero, because it is genuinely useful for smoothing production gaps, and it should never be the majority, because of the margin math above. If you launch marketplace-native, put a hard date on hiring or assigning someone whose only job is converting the top reorderers to direct.

Menu scope trades flexibility against margin. A caterer who will cook anything has no purchasing leverage, no prep standardization, and labor that resets every job — food cost drifts toward 40% and labor toward 34%, which is exactly how a business doing $520,000 in revenue can still burn cash. The disciplined alternative is 5–8 buildable lunch formats — a build-your-own bowl, a handheld or taco bar, a Mediterranean spread, a sandwich-and-salad executive format, a hot-entrée-plus-sides format, a rotating global format — engineered so 70–80% of prep uses shared ingredients across the rotation. That shared base is what creates bulk purchasing leverage, standardized labor, and a 28–32% food cost. Rotate on a 4–6 week cycle so a client eating your food every Wednesday for a year does not get bored, and build a clearly labeled vegan, vegetarian, gluten-free, and nut-free path into every format structurally rather than scrambling per order. The cost of this discipline is that you will decline requests. That is the point.

Two smaller trade-offs worth naming. Delivery staffing: gig drivers are cheap and available, trained employee drivers cost more and are the actual product, since a driver who sets up to your standard and represents your brand is the last person the client sees. And segment choice: micro offices under 50 people are easy to win but order sporadically at low value, mid-size 50–199 offices are the primary wedge with a 7–21 day sales cycle, 200–600 offices sign real contracts but take 30–90 days through a procurement or finance gate, and 600+ campuses are largely owned by managed-foodservice players you should treat as a relationship for overflow work, not a strategy.

Pitfalls that end new corporate caterers, and the systems that prevent them

A public delivery failure. The order that arrives late on the day the CEO has a client in the office does not just cost you that account — office managers in a metro talk to each other in local Slack and community groups, and the story travels. Prevention is unglamorous: a banquet event order checklist verified before anything leaves the building (headcount, menu, dietary breakdown, delivery window, setup instructions, on-site contact, address, parking and loading dock notes), buffer time built into every route, a named backup driver, and a same-day recovery protocol you have rehearsed.

An allergen incident. This is the catastrophic one — a cross-contact failure that hurts someone is an existential legal and reputational event, not a bad review. The system is a documented allergen-control protocol, per-item labeling on every single item you deliver, individually packaged meals for severe allergies rather than buffet-line trust, trained staff, attorney-reviewed disclaimer language, and product liability coverage. Budget $800–$2,500 of attorney time on your contract templates and allergen language. It is not optional spend.

Food cost blowout. Volatility plus an undisciplined menu plus over-portioning is the standard death spiral. The counter-system is costed standardized recipe cards with live pricing, the engineered rotating menu that lets you flex toward chicken or plant-forward formats when beef spikes, purchasing against confirmed orders rather than forecasts (your recurring demand is known 24–48 hours out — a structural advantage restaurants do not have), a weekly food-cost review against actuals, and audited portion specs. The difference between a four-ounce and five-ounce protein portion across 30,000 meals a year is real money. Waste should run under 4–6%; anything higher in a known-demand kitchen is a process failure, not bad luck.

Marketplace dependence. Living permanently at 3–5% net while a platform owns your customer relationship. The counter is a named owner for conversion, a target direct mix of 60–80%, and knowing each platform's terms on off-platform conversion — some are permissive, some are not, and you should read rather than assume.

Over-built fixed costs too early. A $200,000 buildout before recurring demand exists is the most common way a talented chef ends up closing at month 20 with respectable revenue and negative margin. The lease and the equipment debt are fixed; lumpy seasonal revenue is not. Start Path A.

Route sprawl. Saying yes to the tempting 200-person account 40 minutes away feels like growth and is usually margin destruction plus driver churn. Define the radius before you sell, then decline outside it. If a distant account is genuinely worth it, it should justify a second route or a second kitchen — not an exception.

Account concentration. One client at 25–30% of revenue means their budget cut is your crisis. Cap any single account at 12–18% of revenue as the book matures.

Accounts receivable drift. Net-30 corporate clients routinely pay net-50. Invoice the day of delivery, automate reminders, keep a card on file or take a deposit from newer accounts, and treat collections as a weekly rhythm rather than a quarterly panic.

Compliance lapse. A failed health inspection or a lapsed certificate of insurance loses corporate accounts outright, particularly in tech, finance, and healthcare where clients audit vendors. Carry general liability at $1M–$2M per occurrence (corporate clients frequently require higher aggregates and additional-insured status), product liability, commercial auto, workers' compensation once you have employees, and property coverage — figure roughly $3,500–$12,000 annually for a small operator. Hold a certified food protection manager credential, keep temperature and sanitation logs, and treat your documented food-safety program as a sales asset you volunteer rather than a cost you hide.

Owner burnout. Doing everything yourself past the sustainable point is the quiet killer. Hire the kitchen lead ($18–$26/hour) and the dedicated driver ($17–$24/hour) earlier than feels comfortable, because the owner's job by Year 2 is selling and managing accounts, not packing boxes at 6am. The Year-2/3 hires that change the business are a chef de cuisine ($55,000–$85,000) who owns production and food cost, and an account manager ($45,000–$70,000 base plus commission) who owns the pipeline and the marketplace-to-direct conversion.

One closing discipline that spans all of these: run a lean full-time core with a trained part-time flex pool for the mid-week surge, rather than carrying peak-day headcount all week. Combined with route density, that is the single largest controllable margin lever in the operation — and it is the same operational instinct a good RevOps team applies to pipeline coverage, matching capacity to a known demand curve instead of to a worst case.

Related questions

How long before the business is profitable?

Path A operators commonly reach breakeven around month 4–8, once three to five recurring accounts cover kitchen rent, insurance, and vehicle costs. Meaningful owner income usually arrives in the second half of Year 1. Path B pushes breakeven later because occupancy and equipment costs start immediately.

Can I run corporate catering from a home kitchen?

No. Nearly every US jurisdiction requires a licensed commercial kitchen for catering, and cottage-food laws do not cover it. Corporate clients also request health-department documentation. Rent a shared commissary or ghost-kitchen stall — it is the legal and practical entry point.

Should I take weddings and social events too?

Only opportunistically, and never as strategy. Social work is seasonal, every job is a cold start, and it fragments your menu and labor. If you take an occasional high-margin holiday party inside your radius, price it at the full-service tier and treat it as overflow revenue.

What is the fastest way to get the first ten accounts?

Marketplace listings for discovery plus complimentary sample drops to office managers inside your radius. The sample drop works because their current caterer will eventually fail, and you want to be the name already sitting in their desk drawer when it happens.

How many accounts can one kitchen realistically handle?

One production kitchen with a single delivery team supports roughly $2.5M–$5M annually on one shift — typically 35–70 recurring accounts depending on order size. Split shifts push that to $4M–$8M before a second kitchen becomes the constraint-breaking move.

FAQ

What licenses and permits do I need to start a corporate catering business?

You need a business entity (LLC is standard, with an S-corp election common once profitable), an EIN, a food service or caterer's license from your local health department, access to a kitchen that passes health inspection, food handler certification for staff, and a certified food protection manager credential such as ServSafe Manager. Some jurisdictions add a separate catering endorsement or event-specific permits. Requirements and fees vary widely by state and county, so verify with your local health department directly rather than relying on national summaries.

How much money do I need to start?

The realistic range depends entirely on kitchen path: roughly $12,000–$45,000 for a shared commissary launch, $25,000–$70,000 for a ghost-kitchen stall, and $85,000–$280,000 for a leased production kitchen buildout. Beyond the kitchen, budget for a reliable delivery vehicle, hot and cold transport equipment, packaging inventory, insurance, professional food photography, and enough working capital to survive six to nine months of modest owner draw while the recurring book builds.

Do I have to use catering marketplaces like ezCater?

Not strictly, but going without them slows your ramp considerably, because that is where corporate buyers now shop. The practical approach is to list on at least two platforms for discovery and capacity fill, accept the 8–18% commission as a customer-acquisition cost, and systematically convert high-frequency reorderers into direct standing agreements where you keep the full margin. Check each platform's terms regarding off-platform conversion before you build a process around it.

What should I charge per person?

Drop-off lunch typically lands at $14–$22 a head, staffed buffet at $26–$48 plus a 15–22% service fee, and full-service events at $55–$130 plus a 20–28% service charge. Price against your target 28–34% food cost rather than against local competitors. On recurring agreements, a $1–$3 per-head discount versus spot pricing is a reasonable trade for a multi-month commitment and a firm headcount lock.

How do I handle dietary restrictions and allergies safely?

Build vegan, vegetarian, gluten-free, and nut-free paths structurally into every menu format rather than treating them as special orders. Label every item individually with its allergen profile, package severe-allergy meals separately rather than serving them from a shared buffet line, train all staff on cross-contact procedures, keep a written allergen-control protocol, and have an attorney review your disclaimer language. Carry product liability insurance regardless.

Is corporate catering better than opening a restaurant?

They are different businesses with different lifestyles. Corporate catering offers weekday hours, predictable recurring revenue, known demand you can purchase against, and a book of accounts that is genuinely sellable. Restaurants carry nights, weekends, walk-in demand uncertainty, and thinner margins. The trade is that catering is deadline-driven in a way that never relaxes — an 11:45 delivery window is never a suggestion — and it demands B2B sales work that many food founders would rather avoid.

Sources

• https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs
• https://www.fda.gov/food/retail-food-protection/fda-food-code
• https://www.servsafe.com/
• https://www.census.gov/programs-surveys/susb.html
• https://www.bls.gov/ooh/food-preparation-and-serving/food-and-beverage-serving-and-related-workers.htm
• https://www.ezcater.com/
• https://www.nace.net/
• https://www.restaurant.org/research-and-media/research/
• https://www.irs.gov/businesses/small-businesses-self-employed/employer-id-numbers
• https://www.foodsafety.gov/

Related on PULSE

• How do you price a B2B service contract for recurring delivery?
• How do you build a route-density model for a local delivery business?
• How do you convert marketplace customers into direct accounts?
• How do you structure a one-page recurring service agreement?
• How do you calculate unit economics for a food service business?
• How do you decide between renting a commissary and leasing your own space?