TL;DR: Restaurant Facebook marketing works best when a paid ad has one job: fill one quiet service you can actually serve. Pick the night from your POS data, give guests one reason to come, show a real dish in a simple format, target people nearby, and set a spending ceiling from your own margin per cover. Run it for four weeks and count that night's covers. If nothing moves, change one thing or stop.
If ads feel like the hardest part of marketing, there are reasons. The dashboards are full of terms, the budget feels like money thrown into a hole, and nobody can say whether last month's boosted post brought a single guest. Meanwhile, Tuesday night still has twelve empty tables.
This article is only about paid Facebook ads, and only about one use: filling quiet weekday services. What to post organically, across platforms, is in how to promote a restaurant on social media. Building the full campaign around the ad is in restaurant marketing campaign examples. Both belong to our AI for restaurant marketing guide.
When does a restaurant need paid Facebook ads?
When there is a specific service you want to fill, you can staff it well, and your organic posts are not reaching enough people nearby. Paid ads are a tool for a target, not a replacement for a Facebook page that is kept up. If the page has no recent posts or menu, fix that before spending.
Restaurants are already on Facebook. In TouchBistro's survey of more than 600 US independent full-service operators, 99% had at least one social media profile, and Facebook was the most commonly used platform for restaurant marketing (Silicon UK). Being there is the baseline; paying to reach more nearby people on a particular night is the next step.
Before the first ad, check the page a new guest will land on:
- Opening hours, address and phone are correct.
- The current menu or a link to it is there.
- A booking link works on a phone.
- The last few posts are recent and show real food.
Example (illustrative, not a real case): before its first ad, a trattoria finds that its Facebook page still shows summer opening hours and links to the spring menu. The owner fixes both, adds the booking link and posts two fresh dish photos from that week's prep, then starts the ad.
Which quiet night should the ad fill?
The service that is consistently quieter than the rest and that you can staff without strain, usually one or two weeknights. Find it in your POS system: covers by day and service over the last eight weeks. Do not choose a night you are thinking of closing; an ad cannot fix a service the team does not want to run.
Traffic is under pressure for many restaurants. In the National Restaurant Association's 2026 outlook, 60% of US operators reported softer customer traffic and 42% said their restaurant was not profitable the previous year (National Restaurant Association). That makes it more important to spend only where you have empty seats and the capacity to fill them.
Three questions with the head chef and the floor manager:
- How many extra covers could the kitchen handle on that night without another cook?
- Does front of house need an extra person above a certain number?
- Which dishes can be produced from regular stock for an offer that night?
The answers set the size of the target. If the kitchen can take 15 more covers, an ad aiming for 50 is the wrong ad.
Which ad formats suit a slow weeknight?
Simple formats that show the dish and the night: a single strong photo, a short vertical clip of the dish being made, a carousel of a set menu, or an event for a one-off evening. For a small restaurant, one photo or one clip is usually enough to start. The format matters less than a clear reason to come that night.
| Format | What to show | When it suits |
|---|---|---|
| Single photo | The plated dish, the night and the offer in the text | A repeating weekly special, such as pasta night |
| Short vertical video | 10–20 seconds of the dish being made or served at the pass | When the process looks good: grilling, rolling, flambéing |
| Carousel | The courses of a set menu, one per card | A fixed-price weeknight menu |
| Event | Date, time, seats, booking link | A one-off wine evening or supplier dinner |
Use photos and clips from your own kitchen, taken during prep. Guests decide on a restaurant partly from what the food looks like; a stock image or an AI-generated plate sets an expectation the kitchen cannot meet.
How do you write a Facebook ad for a slow night?
Lead with the night or the dish, then the offer, then where and how to book, in a few short lines. Write for someone nearby who has not been in yet. AI tools are useful here for producing several versions of the first line; you choose the one that sounds like your restaurant and names something real.
A four-line structure:
- Hook: the night or the dish ("Tuesday is pasta night on Király utca").
- Offer: what exactly is on ("Fresh pasta, a starter and a glass of house wine at a fixed price").
- Proof: one concrete detail ("Pasta rolled by hand every afternoon").
- Action: how to book, and until when ("Book for Tuesday, tables from 6 p.m.").
Example (illustrative, not a real case): a 50-seat trattoria asks an AI tool for ten first lines for its Tuesday pasta night, giving it the dish, the fixed price, the street and the chef's sentence ("we roll it at 4 p.m. every day"). The owner crosses out seven that could be for any restaurant ("Treat yourself this Tuesday!") and tests two: "Tuesday is pasta night on Király utca" and "Pasta rolled at 4, on your plate by 7". Both ads use the same 15-second clip of the pasta being cut, and target people within a short distance of the restaurant.
Avoid promises in the text: no "the tastiest pasta in town", no claims you cannot back. A specific day, dish and street do more work than adjectives.
How much should a restaurant spend on Facebook ads?
There is no reliable published figure that fits every restaurant, so set a ceiling from your own numbers and test with a small part of it. The ceiling is the most an ad could cost before the extra guests cost more than they bring. Start well below it, and only raise the spend if covers on the target night move.
Weekly ceiling = extra covers you can serve × (average spend per cover − food and drink cost per cover)
Example (illustrative, not a real case): a café-bistro with 40 seats fills about 14 on Wednesday nights. The kitchen can handle 12 more without extra staff. Average spend per cover is 28 and food and drink cost per cover is about 9, so the ceiling is (28 − 9) × 12 = 228 a week in the café's currency. The owner tests with 40 a week for four weeks, running the ad from Monday to Wednesday afternoon only, so it reaches people while they are still deciding.
The ceiling is not a forecast. Some weeks the ad will bring nobody; the test is there to find out whether this night, this offer and this audience respond at all.
How do you know whether the ad filled tables?
Count covers on the target night, week by week, against the four weeks before the ad, and ask new guests how they heard about you. Platform numbers such as reach and clicks are signals, not covers. If the night's covers have not moved after four weeks, change one thing (the offer, the hook or the audience), or stop.
Marketing is one of owners' biggest headaches: in Toast's 2025 survey of 712 US operators, it was the joint second-biggest pain point (16%), after inflation (20%) (Modern Restaurant Management). A simple counting method makes ads less of a guess.
A weekly log with five columns is enough. Example (illustrative, not a real case): the café-bistro's Wednesday log, filled in.
| Week | Spend | Covers on Wednesday | Bookings mentioning the ad | Note |
|---|---|---|---|---|
| Baseline (average of 4 weeks) | 0 | 14 | — | — |
| 1 | 40 | 15 | 1 | Rain all evening |
| 2 | 40 | 19 | 3 | New photo of the risotto |
| 3 | 40 | 18 | 2 | — |
| 4 | 40 | 21 | 4 | Two tables of four from the office block |
Here the owner keeps the ad for another four weeks with the risotto photo and checks again; a flat log would have meant changing the offer or stopping.
Servers can help by asking new faces, "Is this your first time here? How did you find us?" and noting the answer on the booking. After four weeks, the log answers the question better than any dashboard. The wider picture is in the AI for restaurant marketing guide.
If you want a method for writing hooks and choosing ad formats from your own guests' words, and for reading why an ad did not land, we are building the Ad Creative Hooks & Formats Pack. Get notified when it launches.
FAQ
Do Facebook ads work for restaurants?
They can help fill a specific quiet service when the offer is clear, the audience lives or works nearby, and the kitchen can handle more covers that night. They work poorly as general awareness for a small restaurant. Run one ad for one night for four weeks, count that night's covers against the weeks before, and stop if nothing moves.
How much should a restaurant spend on Facebook ads?
No published figure fits every restaurant, so work out a ceiling: the extra covers you can serve that night multiplied by the average spend per cover minus the food and drink cost per cover. That is the most an ad could cost before it loses money. Test with a small part of that ceiling first and raise it only if covers move.
What should a restaurant Facebook ad say?
One reason to come on one night, the offer, where you are and how to book, in a few short lines. Lead with the dish or the night, not with your restaurant's name. Use a real photo or short clip from your kitchen. Leave out superlatives and vague claims; a specific dish and a specific day do more.
Should a restaurant boost posts or create ads in the ads manager?
Boosting a post is a quick start and fine for a first test. Building the ad in the ads manager gives you more control over the audience, the placement and the schedule, which matters when you want the ad to run only before a particular night. Start with whichever you can manage, and keep the same counting method either way.
Sources
- TouchBistro's 2025 State of Restaurants Report — Silicon UK (Business Wire release), 2024.
- Persistent cost increases and enduring demand will shape the restaurant industry in 2026 — National Restaurant Association, 2026.
- Operators double down on profitability, tech and guest experience — Modern Restaurant Management, reporting Toast's survey, 2025.