Most operators I talk to think about catering as a marketing problem. They want more of it, so they look at ads, flyers, a banner in the window. I understand the instinct. But when Rahul and I went through our own catering numbers store by store, the problem was rarely that nobody knew we did catering. It was that we were losing orders we already had a shot at, and we weren’t keeping the customers who had already ordered once.
Catering is the highest-value order a quick-service store takes. The average catering order is $500 or more. One of those is worth dozens of walk-in tickets, and it usually comes from an office, a school or a hospital that will need lunch again next month. So before you spend a dollar finding new catering customers, it is worth fixing the four places where most stores leak them. This is the playbook we run in our own stores, in the order I would fix things.
1. Capture every inbound catering call
The first leak is the phone. Roughly 30% of inbound restaurant calls go unanswered, and the calls that get missed are not random. They cluster at the lunch rush, when the person who would answer is on the line making sandwiches, and after close, when nobody is there at all. That is exactly when an office manager is planning tomorrow’s meeting. A catering buyer who gets voicemail does not leave a message and wait. They call the next place on the list. We wrote up what those missed calls actually cost if you want the full math.
This is the reason we built Maya in the first place. She answers every call to the store after one ring, 24 hours a day, in English or Spanish. When someone says “I need lunch for 25 people tomorrow at noon,” she takes the headcount, the date and time, the items and the contact details, and texts the order straight to the store so the team only has to make it. If the caller wants a person, or the order needs a judgment call, she warm-transfers to the manager and introduces the situation first. You can see how that works on the product page.
Whatever you use, the test is simple. Call your own store at 12:15 on a weekday and again at 9 at night, and ask to order catering for fifty. If either call goes unanswered, that is the first thing to fix, because nothing else in this playbook matters if the order never reaches you.
2. Build a real catering list
The second leak is that most stores do not know who their catering customers are. The order gets made, the tray goes out the door, and the name and number live on a ticket that ends up in the trash. A month later, when that customer has another meeting, you have no way to reach them.
Every catering order should leave behind a record: the person’s name, their phone number and email, the organization, what they ordered, how many people, and the date. That is your catering list, and it is the most valuable list your store owns. If your POS or online ordering already holds this history, export it. If it is scattered across tickets and inboxes, start capturing it properly today and backfill what you can.
When Maya takes a catering call, that record is created as part of the call, with a transcript and recording behind it. For the list you don’t have yet, our Prospecting Engine sources new business accounts around each store, the offices, schools and hospitals within reach that order lunch for groups. But the list you already have comes first. It is full of people who already trust you.
3. Run the reactivation loop
The third leak is the quiet one. Catering customers do not churn because they were unhappy. They churn because they only order when they have an event, and when the next event comes up, they call whoever comes to mind first. If that is not you, you lose the order without ever knowing it existed. I wrote about this in more depth in our post on catering reactivation, but the short version is that a past catering customer who has gone quiet is a retention problem, not a marketing problem.
The loop we run starts after 60 days of silence from an account worth $100 to $500 or more per order. It is four touches in one attributed sequence:
A text from the store. A short, personal SMS built from the customer’s own order history. It names the last order and asks if they have something coming up. No blast, no generic promo.
A phone call. If the text gets no answer, Maya calls them. She references their last order, asks about upcoming events and can book on the spot.
An email. The final touch in the sequence, with an offer.
New accounts. Alongside the win-back, the Prospecting Engine keeps adding new business accounts, so the list grows while you work it.
Two details matter more than the copy. First, it only goes to catering accounts, not every retail customer who once bought a sandwich. Second, it is timed. Office managers rarely plan catering weeks ahead. A reminder that lands at the right moment converts; the same message sent to everyone on a Tuesday does not. And once an account orders again, it should leave the sequence immediately, so nobody gets a win-back text the day after they placed an order.
4. Upsell at the moment of booking
The fourth leak is smaller per order but it happens on every order. When a catering buyer calls, they have already decided to spend. That is the best moment to suggest the things that make the event go better: drinks for the group, a cookie or dessert tray, an extra platter for the people who always show up unannounced. A busy cashier on the lunch rush rarely asks. They are trying to get the caller off the phone and back to the line.
Maya asks every time, in the same natural way, because she is not trying to get back to anything. I don’t have a measured average-order-value lift to give you, so here is the arithmetic instead, as an illustration only. If a $500 catering order picks up a $40 drinks-and-dessert add-on, that order is worth $540, or 8% more. Run the same math with your own order sizes and add-on prices, because yours are the numbers that matter.
Then close the loop after the event. We send a review request by text 24 hours after the order. A happy catering customer who leaves a Google review is also telling the next office manager searching for catering near your store that you can be trusted with a big order. We covered that side in our post on Google reviews.
What this looks like across several stores
For a single store, you can run some of this by hand if you are disciplined. Across several stores, you can’t. Each location has its own phone, its own lunch rush and its own list of offices nearby, and the manager at each one has a store to run. That is where it has to become a system: every call answered, every catering order recorded, every quiet account followed up on schedule, every booking offered the add-on.
We built it for our own stores first. Maya has answered 5,200 real customer calls at our stores as of October 2026, and our own stores’ sales are up about 40% on this model. Those are our results, not a promise for yours. What I can say is that the math on catering is unusually forgiving. Because the average catering order is $500 or more, one recovered catering order a month more than covers the Growth plan on our pricing page. Do that math with your own numbers before you decide anything.
If you only do one thing this week, make the 12:15 test call. Then pull your catering orders from the last six months and count how many of those customers have not ordered since. That number is your starting list.