Facebook ads for restaurants work when every campaign is tied to one measurable action: an order on your own ordering site, a booked table, or an enquiry for catering or a private event. Boosted posts that collect likes and reach rarely prove anything, because nobody can tell whether the person who liked the photo ever ate there.
Restaurants are unlike most local businesses we work with. The decision is fast, often made within the hour, and the ticket is small, so the margin per order is thin and the ad has to be cheap per sale to make sense. At the same time, the most valuable customer is the regular who orders every other week, not the one-off. Below is how we set up Facebook and Instagram for independent restaurants, small groups and delivery-first brands. If you want to talk about your restaurant, apply here.
Three different goals, three campaigns
Direct orders (delivery or collection through your own website or app). The ad drives hungry people to a menu that loads fast on a phone and checks out in a few taps. This is the easiest goal to measure, if tracking is set up properly.
Table bookings. The ad sells an occasion: date night, the weekend brunch, a set menu, a new season's dishes. The booking happens on your reservation system, and the visit happens days later.
Catering, private dining and events. Fewer people, much larger orders, a longer decision. This is the one place in a restaurant account where a lead form makes sense.
Mixing these goals in one campaign makes Meta chase whichever action is cheapest, usually clicks to the menu, and leaves bookings and events underfunded.
What a "qualified" result means for a restaurant
For orders, the only result that counts is a paid order, ideally from a new customer or a lapsed one. For bookings, it's a table that was actually seated, not just reserved. For catering, a qualified enquiry is someone with a date, a headcount and a budget that fits your minimum, inside the area you can deliver to.
Repeat behaviour matters more here than in almost any other niche. A campaign that brings first orders at a loss can still be the right campaign if those customers come back. That only shows up if you can see repeat orders per customer, which your own ordering platform can usually tell you and a delivery marketplace usually cannot.
Questions for catering and event enquiries
When you advertise catering or private dining, a short form saves your events manager time:
- What's the occasion? Office lunch, private party, wedding or celebration, regular corporate order.
- Date (or how soon)?
- How many guests? Ranges starting at your minimum.
- Delivered, or at the restaurant?
- Rough budget per person, if you're comfortable asking.
For orders and bookings there is no form. The menu and the reservation widget are the form, so their speed on a phone matters more than any ad setting.
Creative angles that sell food
- The dish, shot close, in motion: steam, the sauce going on, the cheese pull, the knife through the crust. Phone video in good light usually beats a polished studio shot.
- The kitchen and the people: the chef plating, the pizza going into the oven, the team before service.
- A specific reason to order today: the new menu item, the weekday lunch deal, the dish that's only on this month.
- Occasion framing for bookings: "Friday table for two", "brunch with the group", "a private room for 20".
- Social proof you actually have: your real rating and review count on the platform customers know, quoted accurately.
Food creative wears out fast because people see a lot of it. We plan for regular new shoots, often a short phone session every couple of weeks, rather than one big shoot a year.
Tracking from ad to order, table and event
This is where most restaurant ad spend becomes guesswork. What we set up depends on the channel:
| Sales channel | What we can measure | |---|---| | Your own ordering site or app | Purchase event via pixel plus Conversions API, deduplicated, with order value | | Reservation platform | Bookings, if the platform supports Meta tracking or sends booking data we can match | | Third-party delivery marketplaces | Usually no pixel; we rely on promo codes, dedicated links and before/after comparisons | | Walk-ins | No direct tracking; location-level trends and offer codes only | | Catering forms | Lead, then quote sent, then order confirmed in your records |
Wherever possible we push orders to your own ordering channel, because that's where tracking works, margins are better and you keep the customer's details for repeat orders. The full technical setup for purchases is in our article on tracking Meta Ads purchases for food delivery.
Meta and your POS will never match exactly. We expect a gap and agree up front which number decides budget.
Dayparts, weekdays and seasons
Restaurant demand moves by the hour. Delivery spikes around dinner and on weekend evenings; lunch ads matter near offices on weekdays; brunch is a weekend story. We schedule creative and budget around your own order history, so the ad that sells a lunch deal doesn't run at 9 pm.
Across the year, holidays, school breaks, big sporting events and weather all shift demand. A rainy evening is usually good for delivery and bad for a terrace. The calendar for catering is different again: office parties cluster at the end of the year, weddings and graduations in their own seasons.
Common mistakes in restaurant advertising
- Boosting posts and calling it advertising. Reach and likes are not orders.
- Sending ad traffic to a delivery marketplace where you can't track, can't retarget and pay a commission on every order.
- A menu that's a PDF or a site that takes seconds to load on a phone.
- A radius that's too wide. People won't travel far for a weeknight dinner, and delivery has a hard edge.
- Discounts with no end. A permanent 20% off trains customers to wait for it; a time-limited offer gives the ad urgency without eroding the price.
An illustration: the first order vs the regular
The numbers below are invented to show the arithmetic. An ad brings 100 first orders at $12 of ad spend each, $1,200 in total. If each order leaves $8 of margin after food and packaging, the campaign loses $400 on the first order. If 30 of those customers order four more times over the next few months at the same margin, they add $960. The campaign only looks good if you can see those repeat orders, which is why owning the ordering channel matters.
How we work
- Application and brief. You tell us your locations, channels (own ordering, marketplaces, reservations, catering), delivery area and what you want more of.
- Written proposal. What we would do and the scope, before you pay anything. We work with businesses that have a website or an active business page and an ad budget from $1,000 a month, paid to Meta separately.
- Setup. Separate campaigns per goal, purchase tracking on your ordering site, a fast mobile landing page or menu if needed (landing pages and conversion), and promo codes for channels we can't track directly.
- Ongoing management. New food creative on a regular rhythm, dayparted budgets and reports on orders, bookings and events, not reach. More about how we run accounts on our Meta Ads page.
We don't promise a number of orders, covers or events. That depends on your food, prices, location, reviews and delivery times. We promise measurement you can check against your own POS.