How should a restaurant set up Facebook ads targeting?

A restaurant sells to people who can be at the door in a few minutes, and it already knows the names of a good number of them. Almost every Facebook targeting guide is written for a store that ships anywhere, so it skips both of those facts. This article starts from what a dining room already owns and uses a Ben Heath targeting tutorial to check the mechanics behind each move.

What guest lists does my restaurant already have that Facebook can use?

Reservation exports, online ordering accounts, wifi portal signups, loyalty members and catering inquiries. Any list where you hold guest emails or phone numbers can be uploaded to Meta as a customer list custom audience. That is a warm audience of people who have already eaten in your dining room, and you can count yours in an afternoon.

Upload only lists where guests agreed to hear from you, and check the file for phone numbers as well as emails before you export it. Walking through the mapping step, Heath says "the more information, the more identifiers you have in the list that you upload, the better". He describes what Meta then does with the file this way: "they're going to try and match that with people's Facebook and Instagram profiles".

If your point of sale exports spend per guest, include that column. Heath says that when a customer value column is present, "they would prioritize putting ads in front of people that spent more previously". For a dining room that means the ads lean toward the tables that order a bottle and dessert rather than the ones that split a starter.

How do I retarget people who looked at my menu and never came in?

Put the Meta pixel on your site, then build one custom audience of all website visitors and set the retention window as wide as the tool allows. Menu page traffic is the closest thing a dining room has to an abandoned cart. Someone opened your menu, weighed it against the places nearby, and ate elsewhere.

Three warm audiences worth building on day one:

• All website visitors. In the tutorial Heath moves this retention out to 180 days and notes "That's the max, by the way." His stated reason for going wide is "Because I like really big custom audiences.", and he adds that "in general larger custom audiences produce better results". A guest who read your menu six months ago probably still lives six streets away, which is not true of a shopper who browsed a website once.
• Video viewers. If you post food video, that builds an audience you can advertise to. Heath points out that on that source "the default audience retention here is 365 days".
• Instagram and Facebook page engagers. Heath builds both and says of the Facebook page audience that "365 is the maximum retention window". Before you assume this list is small, check its size against your uploaded guest lists.

Keep these in a warm campaign that is separate from your cold one. Heath recommends you "separate your cold audience campaign from your warm audience campaign", and he strips the interest filters out of the warm ad set once the custom audiences are in it.

One more thing, and it costs money to get wrong. Resist excluding your regulars to save budget. Heath is talking about previous customers when he says "They need to be reminded. There's a reason why all the big brands run ads all the time", and a neighborhood that has not thought about you since February is exactly that situation.

Can Facebook find new diners who look like my regulars?

Yes, through a lookalike audience built from a list you upload, but it is a cold audience and not an extension of your regulars. Heath is blunt that people confuse the two. He also shows that the tool will not accept a single town, so you build it wide and then cut it back with your radius.

The confusion is the expensive part. Heath: "The people in your lookalike audience are similar to the people that have interacted with your business before." He adds that "A lot of people think it's a warm audience and that causes issues." Nobody in that audience has eaten your food. The ad has to introduce the place, not remind anyone of it, which is a different photograph and a different first line.

On geography, Heath tries to enter his own town and reports "I can't put in Cheltenham. I can't get that specific here". His workaround is that "You can create a lookalike audience that operates nationally" and then narrow at the ad set level, which is where your radius lives. On the percentage, his guidance for local advertisers is that "you're advertising locally, you want to go with the largest possible custom audience and go up to 10%", because a wide percentage of a whole country shrinks to very little once a single town is laid over it.

How far around my restaurant should I target Facebook ads?

Target the area your guests actually travel from, not a round number, and read it off your own reservation and delivery records rather than a default. Delivery radius and dine in radius are two different answers. If one side of your map is cut off by a river or a motorway, a circle is already the wrong shape.

Heath's rule for location is the line worth pinning above the pass: "target the areas where you can serve your customers. Don't artificially constrain your targeting options." For a dining room that cuts both ways. Do not shrink to your own street, and do not buy the next county because large audiences sound healthier.

What the tutorial shows in the location field, in the version Heath demonstrates on screen:

• Adding a town preselects a wide ring around it. Typing his own town, he reports that "Meta has defaulted to this town plus a 25-mile radius around the town". That default is the tool's starting suggestion, not a reading of your dining room.
• The ring can be resized. He says "We can go up to 50 miles, 10 miles." and that "We could just do the current city only if we wanted to".
• More than one location can be added, which matters if your guests arrive from two towns rather than from one clean circle.

Heath frames the sizing as a judgment call, "you want to enter in wherever you are and then select the right area around that to target". The difference for a restaurant is that you do not have to guess. Your booking system already holds the postcodes. Whatever distance stops appearing in that export is where the boundary belongs, and it is usually shorter than the default.

Location is also the setting you want held firmly, which is how the platform treats it. Heath explains that "there are still certain elements that we need to use as hard boundaries and hard constraints. Things like location targeting."

Is my neighborhood too small for Facebook ads to work?

Not on its own. Heath relays a Meta recommendation of two million people or more, then says local businesses often fall short of that and that it is fine. A single address in one town is never going to clear that bar, and the tutorial does not treat falling short of it as a failure.

The line, said about that recommended audience size: "With a lot of local businesses, you won't be able to do that and that's absolutely fine." He says it again moments later as "Local businesses often can't."

Interest filters are where a small area does need care, and the tutorial's warning here is narrower than it first sounds. Heath's line that "because local businesses, the population is so small, you basically have to use open targeting" is said specifically about the original audience options, the older manual setup. In the same passage he says close to the opposite about the current default: "if you're advertising a local business using Advantage Plus audience and you add in some interest targeting options, it's not the end of the world". On those inputs he tells beginners "And remember, these are not hard constraints."

So for a dining room the honest answer is that stacking foodie, wine lover and brunch interests is unlikely to sharpen anything, because within walking distance the people who eat out are not a niche. Heath sets mass appeal as one of two conditions for stripping targeting back: "there are two scenarios, two boxes you want to tick to actually use this and see it work well". Food goes in the first box, "a lot of food has pretty mass appeal". The second box is account history, and that is the one a newly opened place fails: "If it's a brand new ad account, Meta doesn't really know who converts, where you get the best results". If your ad account is new, expect a slower start whatever your audience size, and judge it on covers rather than on reach.

Does my targeting still work if bookings go through a delivery app?

Less well, because the booking finishes somewhere you do not control and Meta may never see it. Heath names this as the case where taking manual control still makes sense. Your own inputs, meaning the radius and the guest lists, then carry work the algorithm would otherwise do from conversion data.

The example in the tutorial is a hotel room, and a table booked on a third party reservation site or an order placed inside a delivery app has the same shape: "they click through to a third-party booking page where they book in for an event or a hotel room or whatever and you just you just can't track those conversions". His conclusion is that "Then you need more control as an advertiser to direct Meta to advertise to the right people".

Read the next sentence of that tutorial before acting on it, though. The control he means is the older original audience options, and he says "I would only recommend you do this if you were advertising before Advantage Plus audience came in place and you really know what you're doing". For most restaurant owners the useful response is not to change the setup but to change what the ad points at.

• Send ad traffic to a page you own that then links out, so at least the visit is captured.
• Use a promo code that exists only in the ad, so redemptions at the till identify the source.
• Add a click to call or a message button beside the booking link.
• Keep the radius tight. When the platform has less to learn from, a wide circle spends more before it teaches you anything.

Which tools can a restaurant owner use to run local ads without hiring an agency?

No single tool covers targeting, creative, list handling and publishing, which means whichever one you pick, you should know which job it leaves undone. The table below sets out what each does, what it solves for a single location restaurant, and whether it assumes you already know how ad platforms work.

Disclosure: SaleADS.ai is the product of the company that publishes this site. The other five are independent products with no commercial relationship to us. Of the six rows, only the Meta Ads Manager and Mailchimp entries are supported by the source video. The rest are general product descriptions, and none of the six was tested for this article.

Where does this information come from?

Every quoted line in this article comes from one source, a Facebook ads targeting tutorial by Ben Heath running sixty six minutes, and each quote links to the second it was said. Heath opens it by saying he has been running Facebook ads for more than eleven years. Nothing here comes from anywhere else.

The source is "The ONLY Facebook Ads Targeting Tutorial You Need", in which he states "I've been running Facebook ads for more than 11 years".

Taken from the video: the location field behaviour and the rule about serving areas, the recommended audience size and what he says about local businesses missing it, the customer list upload and identifier mapping, the retention windows, the lookalike geography and percentage logic, the two conditions for open targeting, and the untracked third party booking case.

Not from the video: everything about restaurants. The tutorial is a general Facebook Ads targeting walkthrough and never mentions dining rooms, reservation exports, wifi portals, menu page retargeting or delivery apps. Those applications and the tool comparison are ours. The interface Heath demonstrates is the one he had in front of him when he recorded, so check the current settings against your own account. Any distance, list size or spend figure for your own place should come from your reservation, delivery and point of sale records before you spend against it.