What structured data a university listing can carry

Universities DB is a directory of 198 universities in 65 countries. Each university has its own page with tuition, programmes, rankings and an enquiry form. We read the page source of a few listings in October 2026 to see how much of that information is also given to search engines as structured data. The Chulalongkorn University listing is a good one to go through, as most of its fields are filled in.

The CollegeOrUniversity object

The main block of JSON-LD on the page is a CollegeOrUniversity. It has the name, the page URL, a one-line description, the street address (254 Phayathai Road, Pathumwan, Bangkok 10330), latitude and longitude, a logo, a campus photo and a founding date of 1917.

The sameAs list is the part we would point a client to first. It links the listing to the university's official site, its Facebook, Instagram, LinkedIn, X and YouTube accounts, and its Wikidata entry (Q855730). Search engines already have an entity for Chulalongkorn. sameAs links tell them that this page is about that same university and not another one with a similar name. For a directory with 198 universities, many of them with long or translated names, that removes a lot of guesswork.

Below that is an additionalProperty list made of PropertyValue items. On Chulalongkorn's page there are three: 34,173 students, 12 percent international students and a 10 percent acceptance rate, the last two with "percent" as the unit. Schema.org has no ready-made property for acceptance rate, and PropertyValue is the usual way to add a figure like this without inventing a property.

FAQ markup built from fields

The page also has a FAQPage block with nine questions. Each one is written from a database field. Tuition gives "How much is tuition at Chulalongkorn University?", answered with THB 69,000 to 1,730,000 a year for 2025 to 2026 and a reminder to check the official fees page. There are questions on the acceptance rate, teaching languages, application deadlines (December to February for early applications, March to May for regular ones), scholarships, accommodation, the share of international students, the founding year and the QS rankings (221 in the world, 51 in Asia).

Since August 2023 Google shows FAQ rich results only for a small group of government and health sites, so a directory should not expect the questions to appear as drop-downs in search results. The markup is still worth having for another reason. The same questions and answers are printed on the page as plain text, in the form people type into a search box or ask a chatbot, and an answer that names the university, the figure and the year can be quoted on its own.

A template has one weakness, and this page shows it. The language answer reads "Courses at Chulalongkorn University are taught in Thai; English." The database field holds two values separated by a semicolon, and the template drops them into the sentence as they are. On a site with 198 listings, one fix to the template ("taught in Thai and English") corrects every page that has the same problem. When we review a templated site we read a handful of generated answers aloud for exactly this reason.

Country pages and the homepage

The country pages use a CollectionPage with an ItemList of the universities on them. The Thailand page lists three: Chulalongkorn, Thammasat and Mahidol. The title is generated with the count in it, "Universities in Thailand (3): Compare Programs & Tuition", and so is the meta description, which names the cities covered ("across Bangkok and Nakhon Pathom"). A searcher can see from the result how many options the page has before clicking.

The university pages use a similar pattern for their meta descriptions. Chulalongkorn's starts with the short description, adds the tuition range in baht and ends with the city and country. The homepage has Organization and WebSite markup, and the WebSite block includes a SearchAction that points at the site's own search.

Where the markup comes from

All of this markup is produced from the same database that prints the visible page. Nobody writes JSON-LD by hand for each university. That is the only way it stays correct on a directory, because a fee or a ranking that is edited in one place and left alone in another will make the page contradict its own markup.

If you are planning a listings site, list the facts you hold for every entry before choosing the schema types. A university directory that holds a founding date, an address, coordinates, a Wikidata ID and three or four numbers per entry can mark up most of what Universities DB does. One that holds only a name and a website will have little to mark up beyond those two fields.