Every restaurant marketing guide says the same five things. Post your food. Use video. Run a discount. Target a radius. Boost your best post.
None of that is wrong. It is just not information. It does not tell you whether the venue two streets away has been running the same ad since spring, or what offer sits in it.
That question is answerable, because the Meta Ad Library is public by law and free. This is a method for reading it as an operator rather than a marketer.
Why generic restaurant Facebook ads advice falls apart locally
Restaurants break most of the assumptions the advice is built on.
Your audience is tiny. A national brand can spend for months before anyone sees the same creative twice. You are buying a few miles of map, so frequency climbs fast and creative wears out sooner.
Your margins cannot absorb the standard offer advice. "Lead with a strong discount" is written by people whose product costs almost nothing to reproduce. On a covered table, food cost and labour are spent before the discount comes off, and a deep one can turn a busy Friday into a night that made less than an empty one.
Demand is shaped by the clock and the calendar. You have a lunch window, a Friday, a Sunday and a dead Tuesday. An ad that works at 4pm Thursday is waste at 10am Monday.
The product is appetite, not argument. Almost nobody reads the body copy. They see a plate and decide whether they want it in twenty minutes. Creative quality is not a polish step, it is the mechanism.
The first visit is not the prize. Most of a new guest's value sits in visits three to ten, so judging an ad on the cost of one first visit makes nearly everything look like a failure. Delivery apps bid for that same hungry person too, with more budget than you.
And much restaurant advertising is not direct response at all. It is presence, with no click to measure, because the outcome happens in your dining room days later. Look outward instead: the operators near you spend their own money, and they stop when something stops earning.
What the Meta Ad Library shows you about restaurants, free
Open the Ad Library, choose your country, search a venue or group name, and you get every ad that page currently has live. The useful parts for a restaurant:
- The creative and copy, image or video.
- The date the ad started running.
- How many near-identical variants of one creative are live at once.
- Where it sends people: a booking tool, a menu, a delivery app, or nowhere.
- Whether it is a produced shoot or a boosted post from the page feed.
What it never shows, for any advertiser: spend, radius, day-parting, click through rate, cost per result or return on ad spend. Those live inside the advertiser's account, so any tool showing a rival restaurant's exact ROAS invented the number. That leaves one signal you can verify, and it happens to be the good one.
Why days running is the only performance signal you can verify
An advertiser watches numbers you cannot see. When an ad loses money they kill it. When it earns, they leave it on. So how long an ad has stayed live is a performance proxy you can check yourself. Restaurants are strict judges, because the cash cycle is weekly and the budget is small and visible. An ad live through a season has survived many Monday morning conversations about whether it is worth the money.
The catch is that the Ad Library will not sort by run time. There is no longevity column and no filter for "live over 90 days", just a start date buried in each ad's panel, and Meta's API returns no history either. A single visit is a photograph, not a record. That grind is walked through in how to find a brand's longest-running Meta ad.
AdWhispr ingests a page's entire library and re-snapshots it daily, so the history exists as data and the question becomes a sentence:
"Show me every ad this pub group has had live over 90 days, and tell me what the offer is in each one."
Then read the shape, not just the winner:
"How many of their ads are under 30 days, 30 to 90, and over 90?"
A local group with several ads past 90 days has found something durable. An independent whose oldest ad is 18 days is still experimenting, which tells you the category is open.
One caveat for this trade. A long-running restaurant ad is sometimes a small always-on brand ad rather than a scaled winner, because at a few pounds a day the bar for "not worth switching off" is low. So read longevity next to variant count, as covered in days running as a performance proxy.
The patterns that repeat in durable restaurant ads
Describe the survivors structurally rather than admiring them. Four questions do the work: hook shape, format, offer mechanic, proof type.
Hooks that persist name something concrete in the first few words: the dish, the occasion, the neighbourhood or the time window. "Sunday roast, from noon" survives in a way "great food, great vibes" does not.
Formats split into the plate, the room and the person: close motion on food, wide shots selling atmosphere, and a face doing something real. Long-runners are often less produced than new ads, not more.
Offer mechanics matter most, because each one hits the economics of the visit differently.
| Offer mechanic | What it does to the visit | When it tends to survive |
|---|---|---|
| No offer, appetite only | Protects margin fully | Strong dish photography, known venue, presence budgets |
| Off-peak window only | Fills hours that were empty anyway | Lunch, early week, late night trade |
| Bundle at a set price | Fixes the ticket, controls the damage | Set menus, sharing plates, two courses |
| Free item above a spend threshold | Lifts the ticket to pay for the giveaway | Drinks, sides, desserts |
| Straight percentage off | Discounts guests who were coming anyway | Rarely, usually at a launch |
| Loyalty or list signup | Buys the repeat visit, not the first | Groups with a working email or SMS programme |
Proof types repeat too: a review screenshot, a local award line, a full room, a chef credential, and guest footage that clearly was not shot by an agency.
Turning a proven pattern into your own ad
The output of research is a structure, not an asset to copy.
- Build a watch list of five to ten pages in your catchment: direct rivals, the local group that clearly spends, and one venue from a nearby city worth being compared to.
- Pull each one's survivors and note the four attributes above. AdWhispr classifies every ad by hook, format, tone and offer, and see every ad a competitor runs covers this step.
- Find the mechanic appearing in more than one venue's long-runners. Repetition across independent advertisers is the closest thing to a local verdict.
- Rebuild it with your dish, your room, your face. AdWhispr can clone a proven image ad, or turn a video winner into a shot list you can film on a phone before service.
"Which restaurants and bars near me are running ads right now, and which have ads still live from before summer?"
Then decide what "working" means before you spend, because here it is usually not a click. Presence ads are judged on covers over a period, direct response ads on a redeemable mechanic the dining room can count.
Check the watch list weekly, not once
The value is not the first look. It is noticing that the group down the road quietly retired its spring discount ad and replaced it with a set menu bundle. So stop asking what restaurants should put in their ads, and start reading what the venues near you cannot bring themselves to switch off. For the guided version, start with your first competitor research run.
FAQ
Can I see how much a competing restaurant spends on Facebook ads?
No. Spend is only published for political and social issue advertising. You can see the ads, creative, copy and start date, but never the budget, radius or results. Anyone selling you a rival's exact spend is estimating.
Are Facebook ads still worth it for a single location restaurant?
They can be, but the test differs from a national brand's. A small radius saturates quickly, so plan to refresh creative rather than leave it running forever, and judge results over a repeat visit window, not a single day.
Should a restaurant Facebook ad include a discount?
Not automatically. A percentage off usually subsidises guests who were already coming. Bundles, off-peak windows and spend thresholds protect the economics of the visit better, which is part of why they keep appearing in ads that stay live for months.
Why do my restaurant ads get engagement but no bookings?
Often because the ad is doing an awareness job while being measured as a direct response one. If you need countable results, give it a redeemable mechanic or a booking link, and check whether the long-running ads near you ask for a booking.