Structured data for restaurants, done once

A search engine reading a restaurant's website sees words: an address in the footer, some times on a contact page, a menu, a phone number. It has to guess which numbers are the opening hours, which line is the address and whether the menu is current. Structured data for restaurants removes the guessing. It is a small block of code that tells search engines, in a standard format, exactly what the venue is, where it is, when it is open and how to book.

It sounds technical, and the code itself is, but the idea is simple and the work is mostly done once. A venue sets out its facts in the agreed format, puts them on the website and updates them when something changes, such as the hours. Search engines can then read those facts with confidence, which helps them show the venue accurately in results and maps.

This guide covers what structured data is, the fields that matter for a restaurant, where the markup goes, the two fields people most often get wrong, what it can and cannot do for review stars in search and how to check it. Every venue and figure in it is an example.

Structured data is information about a page written in a format search engines agree to understand. For restaurants and most other businesses, that format comes from Schema.org, a shared vocabulary created by the major search engines. It has a type for a restaurant and fields for its address, hours, menu, cuisine and more.

The most common way to add it is a small block of code called JSON-LD, placed in the page's code. Visitors never see it. Search engines read it alongside the visible page, so it should always describe the same facts the page shows.

Structured data helps search engines understand a venue's facts without guessing, which makes it more likely that the right hours, address and menu appear when someone searches. It can also make a page eligible for richer search results, such as details shown directly under the link.

It is not a ranking trick. Structured data does not push a venue to the top of results on its own. What it does is remove doubt: the search engine knows exactly what the venue is and when it is open, which supports everything else a venue does for its visibility.

The best thing about structured data is that most of it is done once. The venue's name, address, phone number, cuisine and menu link rarely change. Set them out correctly, put them on the website and they keep working.

The parts that do change, mainly opening hours and the menu, need updating when they change, just as the visible website does. The simplest approach is to treat structured data as part of the website: whenever the hours or the menu change on the page, check that the markup changes too.

Schema.org has a type called Restaurant, which belongs to a wider family of food businesses and local businesses. Using it tells search engines that the page is about a place where people eat, with a location and opening hours, rather than about a shop, an article or a recipe.

There are related types for cafés, bars and other food businesses. Choose the one that best describes the venue. For most venues serving meals, Restaurant is the right choice.

For a restaurant, a handful of fields do most of the work: the name, the address, the phone number, the website address, the opening hours, a link to the menu, the cuisine, the price range, an image and whether the venue accepts reservations, with a link to book.

There are many more fields available, and it is tempting to fill them all. Start with these. They answer the questions guests ask most, and they are the ones search engines are most likely to use.

The venue's name, address and phone number should match exactly what appears on the website, on the Google listing and on directory sites. Consistency matters: if the website says "12 High Street" and the listing says "12 High St, Unit 2", search engines may treat them as uncertain.

Pick one version of each and use it everywhere: the same spelling of the name, the same format of address and the same phone number. Structured data is a good moment to settle on the version you will use from now on.

Opening hours in structured data are set out day by day, with an opening and closing time for each. They should match the hours shown on the website, which should match the hours on the Google listing and on the door. Opening hours online that disagree confuse search engines and send guests to a locked door.

If the venue has different hours for lunch and dinner, or closes between services, the markup can say so, with separate opening times for each part of the day. Getting this right means search results can show whether the venue is open when someone is searching.

Opening hours are the field most often wrong. The common mistakes are hours that close after midnight entered as closing "before" they open, split lunch and dinner hours entered as one long day, a day the venue now closes still marked as open and holiday hours never updated.

Each mistake can show guests the wrong hours in search. Check the hours carefully when the markup is written, and again whenever the venue's hours change. Our guide to opening hours online: the detail that loses guests covers keeping hours right everywhere.

Structured data can point to the menu with a link, and it can describe the menu itself: its sections, its dishes, their descriptions and their prices. A menu on your website as a web page makes this possible; a PDF does not, because the dishes inside a file are much harder for search engines to read.

At minimum, include a link to the menu page. If your website tools can mark up the menu itself, it gives search engines a clear picture of what the venue serves. Our guide to the menu on your website, not in a PDF covers why the web page matters.

Prices in structured data should include the currency, and match the prices on the visible menu. A price range field, often written as a short symbol such as "££" or a range of amounts, tells search engines roughly how expensive the venue is.

Keep prices current. When the menu's prices change, the structured data should change with them. Out-of-date prices in search results create the same disappointment as out-of-date prices on the website.

Dishes in a marked-up menu can carry dietary information, such as vegetarian, vegan, gluten-free or halal, using the vocabulary's recognised diets. It helps search engines, and the guests using them, understand which dishes suit which needs.

Only mark dishes with diets they genuinely meet, and keep allergen information on the visible page as well. Structured data supports the menu; it does not replace the clear allergen information guests need before they order.

Structured data can say whether a venue accepts reservations and where to make them, with a link to the booking page. That tells search engines that booking is possible and where it happens, which supports the booking links shown alongside a venue in some results.

Point the reservation link at the page where guests actually book, not the home page. If the booking page moves, update the link, so search engines do not send guests to a page that no longer exists.

Many venues want review stars in search results, and hope structured data will produce them. For a restaurant's own website, it usually will not. Since 2019, Google has not shown review stars for reviews that a local business publishes about itself on its own site, which it calls self-serving reviews.

The stars many people see next to restaurants in search usually come from reviews collected by the search engine itself, through the business's listing, or from independent review sites. Review stars in search are earned through those reviews, not through code on your own website.

That makes reviews the second field people most often get wrong. Marking up a selection of the venue's best reviews, or an average rating the venue calculated itself, does not produce stars and can count against the page if it misleads.

Leave reviews out of your own structured data unless you have a clear, legitimate reason and know the rules. Put the effort into getting genuine reviews on your Google listing and on review sites instead. Our guide to online reviews: the ones that move bookings covers where reviews matter.

Structured data is usually placed in the code of the page it describes. For a restaurant, the main block describing the venue often sits on the home page or the contact page, and a menu block on the menu page. Each page's markup should describe what that page shows.

If your website is built with a common website builder, look for a setting for business details or structured data; many add it automatically from information you enter. If a developer built the site, ask them to add a single, accurate block, and to show you where it is so you can check it later.

Many website builders and booking or menu tools add structured data for you, from the details you enter in their settings. That is convenient, but check what they produce. Two tools adding two different versions of the venue's details, one with old hours, is worse than one accurate block.

Look at the page with a testing tool, see which structured data it carries and where it comes from, and make sure there is one consistent version of the venue's facts.

Test structured data with the free tools the search engines provide. Google's Rich Results Test shows whether a page's markup is valid and which rich results it may be eligible for. The Schema Markup Validator checks the markup against the Schema.org vocabulary.

Run the tests after adding the markup and after any change to the website. They catch most mistakes, such as a missing field, a wrong format for the hours or a broken link, before search engines do.

Structured data is done once, but it is not finished once. When the hours change, the menu changes, the phone number changes or the venue moves, the markup must change too. Old structured data can keep showing old facts in search long after the visible website has been updated.

Add it to the monthly website check: the menu, the hours, the booking link and the structured data. A quick test with one of the free tools confirms that what search engines read matches what guests see.

Structured data is one small part of restaurant SEO, the work of making a venue easy to find in search. The larger parts are a complete and accurate Google listing, genuine reviews, a fast website with a real menu and clear information, and links from local sites and publications.

Treat structured data as supporting those, not replacing them. Our guide to restaurant SEO: what actually moves the needle covers the wider picture.

For local searches, such as "restaurants near me", search engines rely heavily on the business's own listing, such as its Google Business Profile, alongside the website. The listing and the structured data should tell the same story: the same name, address, phone, hours and booking link.

When they agree, search engines can be confident. When they disagree, they have to choose, and may choose wrong. Our guide to your Google Business Profile is the real homepage covers the listing.

In a group, each venue should have its own page and its own structured data: its own address, phone, hours and menu. A single block describing "the group" with one address tells search engines nothing useful about each venue.

If the group has one website with a page per venue, put each venue's markup on its own page. That lets search engines show the right venue, with the right hours and booking link, to someone searching nearby.

Structured data can also describe events, such as a wine dinner or a festive menu, with their dates, prices and booking links. For venues that run regular events, it can help those events appear in search.

It is optional, and only worth doing if events are an important part of the trade and are kept up to date. An event in the markup that has already passed is a small but avoidable error.

Structured data cannot guarantee rich results; search engines decide whether to show them. It cannot make a slow or confusing website rank well. It cannot produce review stars from your own reviews. And it cannot fix facts that are wrong on the page itself.

What it can do is make the venue's facts clear and consistent for search engines. Done once, correctly, and kept current, that is a small investment with a lasting benefit.

When adding structured data, check:

Then add the markup to the monthly website check, so it stays as current as the page it describes.

An invented bistro's website has no structured data, and search results sometimes show the wrong hours. Its developer adds one block of markup to the contact page: the Restaurant type, the name, address and phone exactly as on its Google listing, the hours for lunch and dinner on each day, a link to the menu page, the cuisine, the price range and the booking page.

The bistro had considered marking up its best reviews to get stars, but leaves them out. It tests the page with the free tools, fixes a late-closing Saturday that was entered wrongly and adds the markup to its monthly checks. Search results now show its hours correctly, and the booking link appears alongside its listing.

In servearound, the venue's booking page on servearound.com already carries structured data for a restaurant: its name, its address and map position, its phone number, opening hours taken from the venue's shifts, a link to its menu, its logo and social links where they are set, and a reservation action pointing at the booking page. The menu page carries menu structured data, with each section and dish, its description, its price and currency, and its dietary tags where they match a recognised diet.

Those pages are open to search engines only when the venue has published them: the booking page when public booking is on and the venue has shifts, the menu page when a menu is live, while the table's QR menu and the staff screens stay out of search. Review stars are not included, because servearound collects no online reviews. Your own website's pages need their own markup, if you add any. See the reservations module and the menus module.

Information about a page, written in a standard format search engines understand, describing the venue's facts.

Not directly. It helps search engines understand the venue's facts, which supports everything else.

Name, address, phone, website, opening hours, menu link, cuisine, price range, image and reservations.

Usually not. Google does not show stars for reviews a local business publishes about itself.

In the code of the page it describes, often the home or contact page for the venue and the menu page for the menu.

With Google's Rich Results Test and the Schema Markup Validator.

Yes, whenever the hours, the menu or the venue's details change.

Opening hours, especially late closing and split services, and self-made review ratings.

Structured data for restaurants tells search engines, in a standard format, what the venue is, where it is, when it is open and how to book. Use the Restaurant type and start with the fields that matter: name, address and phone matching the listing exactly, opening hours for every day, a link to a menu on your website rather than a PDF, cuisine, price range, an image and a reservation link.

It is mostly done once, but the hours and the menu must be kept current. Watch the two fields people get wrong: opening hours, especially late closing and split services, and reviews, because self-made ratings do not bring review stars in search. Test the markup with the free tools, keep it consistent with the Google listing and add it to the monthly website check.