Most of your Thanksgiving catering revenue is committed before the middle of November, by people making one phone call and going with whoever picks up and sounds organized. The food is not the differentiator in that call. The answer is.
This is the clearest case in the calendar of phone performance directly converting to revenue, because the orders are large, the decision window is short, and the caller has three or four restaurants on a list. A missed call in the first week of November is not a lost twenty-dollar ticket. It is a four-hundred-dollar order that goes to the place that answered.
The shape of the two weeks
Volume does not build gradually. It steps.
The first weekend of November is when household planning happens, and the calls land Monday through Wednesday after it. That block is the largest single concentration of catering inquiries most restaurants see all year, and it arrives during regular lunch and dinner service with no extra staff on the floor. Your host is taking a fifteen-minute catering call while four people wait to be seated.
The second wave is smaller and comes about a week out, mostly from people confirming, changing headcount, or adding a side. Those calls are short but there are a lot of them, and each one interrupts someone.
The third wave arrives in the final forty-eight hours. Those callers have a problem, they are calling everyone, and they will buy anything you can make. It is the highest-margin wave and the one most restaurants miss entirely because everyone is cooking.
What a good catering call captures the first time
A Thanksgiving order is not a phone order with a bigger number on it. It has a different set of required fields, and the cost of missing one is a callback during the busiest two weeks of your year.
Headcount and package. Pickup or delivery, with the address if delivery. The specific pickup window, not just the day. Hot or cold, which decides your entire morning staging plan. Sides and substitutions itemized rather than "the usual". Dietary constraints, which on this holiday are more common and more consequential than usual. A callback number that someone will actually answer on Thanksgiving morning. And whether a deposit was taken.
That is the checklist a person forgets on the fourth call of a lunch rush and a configured agent does not. The general principles of taking a large order over the phone are in voice AI for large catering orders, and they apply here with less margin for error.
Where the orders leak
Four places, in rough order of how much they cost.
Unanswered calls during the first-week block are the biggest loss and the hardest to see, because a caller who reached voicemail and booked elsewhere leaves no trace in your system at all. If you want to know how big this is for you, pull your carrier's call log for the first two weeks of last November and compare answered against total. Most operators are surprised, and the method is in how many calls does your restaurant miss.
Callbacks that never happen are the second. A message taken on a ticket rail during a rush is a message that gets found at ten that night, when the caller has already ordered somewhere else.
Overselling a pickup window is third. It does not lose the order, it loses the customer, because a family standing in your vestibule for thirty-five minutes on Thanksgiving morning remembers it for years.
Fourth, and least discussed, is the large inquiry that reached a person who was too busy to sell. A twenty-five-person order deserves a conversation about upgrades and add-ons. It rarely gets one when the person answering has a room to seat. That qualification step is what catering lead qualification exists to protect.
Splitting the volume so the big calls get a human
The useful arrangement is not "automate the phone" or "hire a phone person". It is triage.
Let the agent take the standard packages, answer the hours and menu and pickup-window questions, confirm existing orders, and handle the third-wave desperation calls that come in at eight at night. Those are high-volume and low-variance. Route anything above your threshold, anything with a dietary conversation, and anything that sounds like a corporate account to a named person with a phone that is not the host stand phone.
The reason to split this way is that the two kinds of call reward opposite things. A standard package order rewards speed and consistency: the caller wants to book, get a confirmation, and go. A twenty-five-person order rewards attention, because the person on the other end is making a decision they will be judged on by their family, and they are choosing partly on whether you sounded like you had done this before.
Put a person in front of the first kind and they get bored, rushed, and inconsistent by the fortieth call. Put software in front of the second kind and it takes a competent order while leaving four hundred dollars of upgrades unsold. Neither is a failure of the tool, it is a failure of routing.
The threshold is yours to pick. Twenty guests is a reasonable default for a full-service room; a deli doing office trays might set it at forty. What matters is that it is written down before November, so it is not being decided by whoever happens to be holding the receiver.
The second half of that arrangement is that the person receiving handoffs has to be reachable. A handoff into an unstaffed extension is worse than no handoff, because the caller has now been transferred and abandoned. Decide who carries that phone on which days, and put it on the schedule like a shift.
Capacity, deposits, and the morning itself
Cap by slot. Thanksgiving morning is a staging problem more than a cooking problem, and your real limit is how many complete orders can sit boxed and labeled at once without someone handing out the wrong one. Count your staging space, divide by thirty-minute windows, and let the phone stop offering a full window rather than telling someone at pickup.
Take deposits on the large ones. Not on everything, because the friction costs bookings, but above a threshold where a no-show is genuinely painful. Card capture over the phone has real handling requirements and they are covered in collecting payment over the phone safely and in more depth in deposits for large catering.
Then send confirmations in writing. A text with the date, time window, items, and total kills most of the day-of confusion by itself, and the mechanics are ordinary. Details in SMS order confirmations.
On Thanksgiving morning your phone should be doing exactly one job: telling people what time their order is ready and where to park. Everything else should have been settled two weeks earlier. If your kitchen is answering questions about whether the gravy is included at 9am on the day, the failure happened in the first week of November, and the fix is on next year's calendar rather than this morning's. The general holiday version of this planning sits in the holiday rush phone ordering playbook.