- Author:
- Patric Wirth
- Published:
Which details belong to a business, and which profiles are official? Structured data makes these connections clear. A salon example shows how it works and how to check your own website.
A hair salon’s website lists the business name, the owner and the product brands it uses. It also includes an address, opening hours and links to social media. Visitors use the wording and layout to understand which details belong together. Structured data makes those relationships explicit: this is the business, this is where it is located, and these profiles belong to it.
This is relevant to a salon, a trades business or a beauty studio alike. It concerns the information customers use to find, recognise and contact a business.
What is structured data?
Structured data gives information a defined meaning. For a business, it can say: this is the business name, this address belongs to its premises, and these are its opening hours.
Two other terms often come up alongside it:
Schema.org is a shared vocabulary. It describes businesses, people and products, among other things, along with their properties and relationships.
JSON-LD is a format for writing down and connecting these details. It is based on the JSON data format; “LD” stands for “Linked Data”.
Schema.org supplies the terms, while JSON-LD provides a way to write them in code. Google supports several formats and recommends JSON-LD. The data block sits in the page’s HTML without appearing as additional text for visitors.
How is this information used?
When a search system retrieves a web page, it can process the structured data alongside the visible content. This gives it explicit clues about what the page describes and how the things mentioned relate to one another.
On a business website, we can distinguish between the web page, the business and its location. The page can refer to the business it presents. The business description, in turn, can provide its address and opening hours. This makes it clear which information belongs to which business.
To connect separate entries, the things being described can be given an identifier. In JSON-LD, this is called @id. Reusing the same identifier means the reference points to the same thing. A URL can serve as an identifier without representing a separate, additional web page.
The website provides this description. It does not automatically register a record in a central database. Each search system decides which information it processes and uses for its results.
Google uses structured data to help understand content and enable certain enhanced search appearances, known as rich results. For local businesses, Google documents details such as addresses and opening hours. These provide additional information about the business. An existing Google Business Profile should also be kept up to date. Neither a particular search appearance nor higher rankings are guaranteed.
A small JSON-LD example for a business
Consider a fictional German salon called “Salon Beispiel”. Its website says: “Visit us at Beispielstraße 12 in Musterstadt. We are open Tuesday to Friday, from 9 am to 6 pm.” The structured description assigns these details to the salon and connects it to the web page.
The following snippet is a simplified teaching example. The name, address and opening hours are fictional; example.org is a placeholder for the business’s own domain. It is not a complete implementation template.
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@graph": [
{
"@type": "WebPage",
"url": "https://example.org/",
"mainEntity": {
"@id": "https://example.org/#salon"
}
},
{
"@type": "HairSalon",
"@id": "https://example.org/#salon",
"name": "Salon Beispiel",
"address": {
"@type": "PostalAddress",
"streetAddress": "Beispielstraße 12",
"addressLocality": "Musterstadt"
},
"openingHours": "Tu-Fr 09:00-18:00"
}
]
}
</script>Four connections are enough to get started:
@contextpoints to the vocabulary. Here,@graphgroups the descriptions of the web page and the business.WebPagedescribes the page. ThroughmainEntity, it refers to the salon as the page’s main subject.The
HairSalonentry contains the same@id, identifying the business being referred to. Its address is listed underaddress.openingHoursdescribes the opening hours.Tu-Frmeans Tuesday to Friday, followed by the opening and closing times.
The website still contains readable text for visitors. The additional description makes explicit which details describe the business. Understanding the principle and checking a website does not require being able to write this code.
sameAs: which profiles belong to the business?
A business may have a presence on its own website, LinkedIn and other platforms. sameAs adds references to other pages that unambiguously identify the same business. In effect, it says: “The business on this profile page is the same one described here.”
For our salon, its official business profiles are appropriate references. The owner’s personal profile would belong in a description of that person instead. A supplier or a similarly named business is not a suitable match.
Google explicitly includes sameAs in its documentation for organisations, giving profiles on social networks or review platforms as examples. Organisation data can help distinguish one business from another.
My recommendation: linking existing official profiles with sameAs is a useful addition. Its practical value lies in making their shared identity clear. Simply adding more links does not improve the description; choosing the right ones matters.
HGbudd.net’s documented implementation automatically associates configured social profiles with an existing organisation entry. @id and sameAs serve different purposes: the identifier connects descriptions, while sameAs points to other reference pages for the same identity.
Which terms and rules apply?
The technical names come from the Schema.org vocabulary. A type describes what something is, such as HairSalon for a hair salon or the more general LocalBusiness for a local business. A property describes a detail about it, such as name, address or openingHours.
These technical names stay the same regardless of the website’s language. Their values can be in German or another language. For each property, the Schema.org documentation explains which types it applies to and what values are expected, such as text, a URL or an address.
The requirements of the system using the data also apply. For certain search appearances, Google specifies required and recommended properties. Not every Schema.org type is used for a special Google search appearance.
Under Google’s guidelines, the information must match the visible content and be up to date. In our example, the address and opening hours must genuinely belong to the business being described and agree with the website. Fabricated reviews or details about unrelated content do not meet that requirement.
What role does this play in AI answers?
GEO, short for “Generative Engine Optimization”, refers to efforts to improve visibility in AI-generated answers. Structured data can add to a website’s description; it does not replace helpful content.
Google’s AI Overviews and AI Mode require no special Schema.org data. JSON-LD is therefore not a mandatory switch for AI visibility, nor a promise that ChatGPT, Claude or other systems will recommend a business.
How to check your own website
No coding knowledge is needed for an initial check. The tools test the specific page entered. Besides the homepage, it is worth checking the contact page or a dedicated location page.
1. Look at the existing data
Open the Schema Markup Validator and choose the option to test a URL.
Enter the full page address, including
https://, and start the test.Open the entries it finds, such as
LocalBusiness,HairSalonorOrganization. Some details are nested inside another entry.Check the business name, address, opening hours and any
sameAslinks. Do they match the business and its actual profiles?
If no data appears, first check the address and whether the page is publicly accessible. The page may not contain structured data, or a retrieval problem may explain the empty result. This test is not an inventory of the entire website.
2. Check Google’s requirements
Next, test the same address with the Rich Results Test. It shows which supported data types it recognises and where there are issues. Fix critical errors first, then review suggestions about optional properties.
The tools have different purposes: the Schema Markup Validator checks Schema.org markup generally. Google’s tool checks the enhanced search appearances it supports. If only the first tool finds data, that does not automatically mean something is wrong. Passing a test also does not confirm that Google has already added the page to its search index.
3. Make one concrete improvement
My recommendation is to start with one important page. Correct inaccurate information first, then consider useful additions. If someone manages the website for you, send them the test report with a specific question: “Are our business, contact details and profiles described correctly here?”
If the data is already correct, there is no need to install another plugin just in case. What matters more is keeping the visible content and its structured description consistent as information changes.