A corporate client at a tech campus does not see the kitchen. They see two people in clean uniforms, a stack of chafers, and hot food on a table 15 minutes before their meeting starts. What they do not see is the four hours of work that had to sequence perfectly for those 15 minutes to look effortless.
I spent years delivering hot lunches into Stanford, Google, Apple, Meta, LinkedIn, Salesforce, Cisco, Adobe, and Nvidia out of a $30M Michelin-recognized restaurant group. Some campuses are easy. Some have loading docks with 20 minute turnaround windows, badge desks that add 10 minutes, elevators that only take one chafer at a time, and setup rooms two buildings away from where the vehicle can park. Every one of those steps eats holding time, and holding time is the currency of hot food catering.
Here is the sequence that lands a 400 person drop on time and hot, and the traps I learned to plan around after missing them at least once.
T minus 24 hours: the confirmation is a contract
The catering coordinator sends a written re-confirmation to the client contact exactly 24 hours before the drop. Not a note. Not a text. A formal message that mirrors the original order and asks the client to reply confirming five things:
- Final headcount and any dietary changes.
- Delivery window down to a 15 minute band.
- Loading dock number and vehicle instructions.
- Setup location inside the building, including floor and room.
- Client contact who will be onsite at delivery, with mobile number.
If the client replies without confirming all five, the coordinator asks again. If it is 4 p.m. the day before and we still do not have all five, the coordinator calls. Nothing goes into production without a locked confirmation. Every miss I have investigated in this channel traced back to a verbal change made between confirmation and delivery that never got written down.
The reverse timeline: build backward from drop
Once the confirmation is locked, the executive chef and catering coordinator build the operational timeline from the drop time backward. Everything else on that day, every prep station, every vehicle load window, every cook clock in, is oriented against that reverse clock.
Fig. 1 · The reverse clock for a 400 person hot lunch, 12:00 p.m. drop.
Cold prep first, hot cook last
Anything that plates well cold and holds in the walk-in for hours goes on the earliest prep station. Salads, dressings, cold sides, fresh sliced accompaniments. Anything hot that starts losing quality after 45 minutes goes on the latest possible cook window. Proteins, grains, sauces held at temperature. The prep sheet is written in reverse and read in forward order by the cooks, and the sequencing is the executive chef's job, not the coordinator's.
Pack out is a station, not an afterthought
The single biggest logistical error in catering is treating pack out as something that happens after the food is done. Pack out is a station, staffed by a named person, with a printed pack list on the wall and a physical checklist for every chafer, sterno, tray, platter, utensil, sign, napkin pack, label, and delivery bag.
The pack out station opens 75 minutes before departure and closes 15 minutes before. Every item gets checked off as it goes into the vehicle. If the pack list is not complete when the vehicle needs to roll, the vehicle waits and the coordinator gets a call.
Pack the vehicle in reverse setup order
Load the vehicle so that the last item packed is the first item out at the client site. That means chafers and sternos load first because they set up first. Hot food goes in second because it comes out second, but they load last because they are the last thing prepped and they set up second. Beverages load in the position that lets them come out last.
The vehicle diagram is fixed. Same layout every drop. It is taped inside the back door of the vehicle. Drivers who have run 200 drops still glance at it before pulling away, because a wrong pack order costs 10 minutes at the client site and there is no way to make up 10 minutes when the meeting starts at noon.
The client sees 15 minutes of setup. The kitchen ran the play against a reverse clock for four hours to make those 15 minutes look effortless. Every second of holding time was accounted for before the vehicle left the building.
Depart with a 15 minute buffer, not more
The vehicle should arrive at the loading dock 15 minutes before setup starts. Earlier is worse than later. Loading docks at large tech campuses have security windows and queues. A vehicle sitting for 45 minutes waiting to unload is a vehicle taking up a spot another vendor needs, and it starts to burn the goodwill you need to keep at the badge desk.
Fifteen minutes is the sweet spot. Enough time to park, badge in, get the freight elevator, and walk the first load to the setup room. Not enough time to eat into other vendors' windows or make anyone at the security desk annoyed to see you.
Setup is a choreography, not a scramble
A well-trained two person catering team can set up 400 covers in 15 minutes. Not because they are fast. Because they never stop to think. The setup sequence is identical every time, every client, every menu.
- Chafers and sternos first. Assemble every chafer, light every sterno, wait for the water pan to heat. This is the only step that takes real time and it happens before food comes out.
- Hot food into chafers. Top to tail, one pass through the buffet line, lids on immediately.
- Cold platters. One pass, same direction as the hot food, all cold plating done together.
- Beverages. At the end of the line, or on a separate table, so they do not become a bottleneck at the front of the queue.
- Signage, utensils, napkins. Front of the line, staged for the first guest. These get set as the first two steps happen, by the second person.
Same sequence every time. New drivers learn it on their first shadow drop and can run it independently by the third. The consistency is what allows the coordinator to schedule two person teams instead of three person teams, which is where the labor economics of catering actually work.
The debrief that becomes the operational asset
The moment the vehicle leaves the client parking lot, the driver and the delivery team run a two question debrief on the road back.
- What almost went wrong that the client did not see?
- What did we do that we should do every time?
Both answers get typed into the client file before the driver returns to the kitchen. Over 12 months, that file becomes a document that is worth more than the account itself. It says: at this campus, use the north loading dock, not the east one. The elevator to floor 4 is out on Fridays. The client contact goes to lunch at 11:30, so all confirmations must be done by 11:00. There is no signage room, bring extra tape.
The kitchen gets 1 percent better on every drop. The client experiences it as a team that seems to know their building better than their own facilities staff do. That is not a personality trait. That is a discipline. It is the same reason a Michelin-recognized restaurant kitchen can plate 200 covers in a night without a plate leaving the pass wrong.
The things that will go wrong on the day
The playbook above assumes an ordinary day. Some percentage of drops are not ordinary. Traffic on the 101 turns a 25 minute delivery into 55. The loading dock the client promised is closed for a fire drill. The setup room the client reserved was double booked with a leadership offsite and got moved two floors up. A client updates the headcount from 300 to 380 at 9:15 a.m., an hour before the vehicle is supposed to leave.
None of these are excuses. They are the operating environment. The playbook has to handle them without a miss showing up on the client's side.
Traffic and route buffer
Every delivery time gets a route buffer built in, sized to the specific address and time of day. A drop into a downtown San Francisco office at 11:00 a.m. gets a 25 minute buffer against the base drive time. A drop into a suburban campus at 10:00 a.m. gets an 8 minute buffer. The coordinator maintains the buffer table off actual delivery times, not off Google Maps' estimate, because the vehicle also has to load, exit the parking area, get through a security gate, park, and roll a cart. Google Maps does not know about the security gate.
Late headcount changes
An 80 person increase at T minus 105 minutes is manageable if the operation is designed for it. Every large drop carries a 10 percent overproduction cushion in the prep sheet, and every catering menu is designed with at least two items that can be scaled in the last hour without a new prep cycle. The coordinator has the client's late change protocol in writing: any change above 15 percent inside 12 hours incurs a rush fee. The rush fee is not to make money. It is to make sure the client's team pushes back on their own last minute changes.
The catering coordinator is the load bearing role
The single most under-invested position in most catering operations is the catering coordinator. A great one turns a chaotic channel into a boring one, which is exactly what you want catering to be. A weak one turns a good channel into a series of near misses that eventually convert into a lost account.
The role sits at the intersection of sales, operations, and client management. On a Tuesday morning the coordinator is confirming Wednesday's drops with clients, sequencing the prep sheets with the executive chef, scheduling drivers, updating the CRM with new leads from Monday's inbound, and pulling Monday's debrief notes into the client files. On a Wednesday morning they are on the phone with a client whose meeting got moved from 12:00 to 12:30 and coordinating with the kitchen to hold hot food an extra 30 minutes without dropping quality.
Coordinators pay for themselves at a fraction of what they cost. In a channel doing $60K a month, a $75K coordinator is 10 percent of revenue. In a channel doing $200K a month, that same coordinator is 3 percent. Hiring the coordinator is what allows the channel to scale from $60K to $200K, so the ratio moves in the right direction if the growth is real.
What the client actually notices
After delivering into every major tech campus in the Bay Area at scale, one thing became clear. Clients do not remember the drops that go well. They remember the drops that go badly, and they remember the drops where something almost went badly and the vendor caught it in a way the client noticed.
The client asks for a last minute vegan option and the driver produces one from the vehicle because the coordinator had already built the possibility into the pack list. That is the moment the vendor becomes irreplaceable.
You cannot build every possibility into the pack list. What you can do is build the coordinator role, the reverse timeline, the setup sequence, and the debrief loop such that the odds of the small catch happening are high, and the odds of the miss happening are low. Both matter to the account.
The point
Large corporate catering is not a bigger version of a small delivery. It is a logistics operation with a hard clock, a demanding client, and no room to recover from a single missed sequence. The teams that do this well are not more talented than the ones that miss. They run the same play every time. Confirm in writing. Reverse the timeline. Pack in setup order. Roll with a fifteen minute buffer. Set up in a fixed sequence. Debrief on the way back.
Do it that way and the client comes back next week. Miss any one step and you spend three weeks earning back a relationship that took two years to build.