Most of the conversation about missed calls is about the lunch rush. That is where the volume is, and it is where operators can see the problem with their own eyes: the line out the door, the phone ringing behind the counter, nobody free to pick it up. We wrote about that version of the leak in why roughly 30% of our calls went unanswered.
The call I want to talk about here is the one nobody sees at all. It comes in at 8pm, after the dinner crew has started closing, or at 9:30 when the lights are already off. Nobody watches it ring. It just rings out, or it goes to a voicemail box that gets checked sometime the next afternoon, if at all. In my experience running stores with my brother Rohit, that call is disproportionately catering, and it is the most expensive call a store misses.
When catering calls actually come in
A walk-in customer decides what they want when they are hungry. A catering customer decides when they are planning. Those are different moments in the day.
The person ordering a tray for tomorrow's 11am meeting is usually an office manager, a teacher, a coach or a parent. The catering order is a task on their list, and a lot of them get to it at the end of the day, when they finally sit down and realize lunch for fifteen people is their problem tomorrow morning.
So they call. In our stores we saw two clusters. One sits in the lunch and dinner rush, when people are calling to plan the next day while the store is slammed. The other is after the rush and after close, which is the one this post is about. Catering customers often call at 8pm to book for the next morning. Before we fixed it, those calls went to voicemail and got lost.
You do not have to take my word for the pattern in your stores. Pull one week of call logs and sort the unanswered calls by hour. Look at everything after your dinner peak and everything after posted close. If your stores look like ours did, that tail is not empty, and it is not mostly wrong numbers.
What a lost 8pm call actually costs
A missed sandwich order at noon is an $18 problem. A missed catering call at 8pm is a different category of loss, for three reasons.
The ticket is bigger. The average catering order is $500 and up. That is the figure we use across everything we publish, and it is the one in our full missed-call math.
The caller does not wait for you. The person calling at 8pm needs food at 11am. If nobody answers, they are not going to call you back at 10am the next day and hope you can still fit them in. They call the next place on the list that night, or they place an online order somewhere else before they go to bed. By the time anyone at your store hears the voicemail, the order is already gone.
You lose the account, not just the order. Catering repeats. The office that orders for one monthly meeting orders for the next one too, and they order wherever it went smoothly last time. When the 8pm call goes to a competitor, the competitor gets the relationship. That is why we spend so much effort winning lapsed catering customers back: it is much cheaper not to lose them in the first place.
Here is the arithmetic, labelled as what it is: an illustration, not a measured result. Say one store misses one after-hours catering call a week. At a $500 average, that is about $2,000 a month in orders that never happened, before you count the repeat business that went with them. Change the inputs to match your stores. If you miss one a month, it is still $500 or more every month, and one recovered catering order a month more than covers a month of our Growth plan on its own (the numbers are on our pricing page).
The after-hours call is the only missed call where the customer has already decided to spend $500 and simply needs someone to say yes.
Why the usual fixes do not work after close
We tried the obvious things before we built anything.
Voicemail. A catering customer at 8pm is not going to leave a detailed message and wait. Even when they do, the message is missing half of what you need, and by the time someone calls back the order may already be elsewhere.
Forwarding to a manager's cell. This works for about two weeks. Then the manager is driving home, or putting kids to bed, or asleep, and the calls stop getting answered again. It also burns out your best people.
An online catering form. Worth having, but a lot of catering buyers want to ask a question or two and hear that it is handled. A form does not catch the ones who called.
Extending phone coverage with staff. Paying someone to sit by the phone from close until midnight for a handful of calls does not make sense at one store, let alone ten. We compared that option in detail in AI answering versus hiring a receptionist.
24/7 capture, next-day confirm
The fix that worked for us has two halves: capture the call completely when it comes in, whenever it comes in, and confirm it with a human first thing the next day.
Capture. Maya, the AI voice agent we built for our own stores, picks up the store's line on the first ring, at any hour. The store keeps its number; calls forward to her. She knows that store's menu, pricing and catering options, and she works in English and Spanish, switching mid-call if the caller does. On a catering call she takes the details a good manager would take: what they want, how many people, the date and time, pickup or delivery, the contact name and number. She can answer the questions that usually decide the order, like what comes in a platter, what the lead time is, and whether there are vegetarian options.
Hand-off. When the call ends, the captured order goes to the store team by text, with a full transcript and recording of the call on file. It does not go into your POS automatically. Nisra does not inject orders into any POS today, and I would rather say that plainly than have you find out later. A person on your team sees the order, makes the call on anything unusual, and puts it in.
Confirm. The next morning, the opening manager has the order waiting, with everything needed to confirm it in one short call or text back to the customer. That call is worth making in person: it catches anything that needs a human decision, and it is the first step in turning a one-time order into an account.
During open hours the hand-off can be live instead. If a caller asks for a person or a catering lead needs one, Maya warm-transfers to the manager and introduces the situation before she bridges the call.
The result is that the 8pm caller gets what they actually wanted that night: someone picked up, took the order properly and told them the store would confirm in the morning. They go to bed with lunch handled, and they do not call the next place.
Where to start this week
Before you change anything, measure it. Pull one week of call logs per store, filter to calls after your dinner peak and after close, and count how many went unanswered. Then pick one catering customer you know orders regularly and ask them when they usually call. The answer is often "the night before."
If the tail is real, the case is simple arithmetic: a handful of after-hours catering calls a month at $500 and up each, against what it costs to answer them. You can see the full feature set on our product page. We have answered 5,200 real customer calls at our own stores as of October 2026, and own-store sales are up around 40% on this model. Those are our stores, not a promise for yours, which is exactly why I would start with your own call logs.