Refract Group2026
Glossary
Schema markup
Schema markup is structured data added to a page, usually as a JSON-LD block, that uses the schema.org vocabulary to tell search engines what the page is and what it describes: an organisation, an article, a service, its breadcrumbs.
refract.group/
<script type="application/ld+json">
{"@context": "https://schema.org",
"@type": "Organization",
"@id": "https://refract.group/#organization",
"name": "Refract",
"alternateName": ["Refract Group"],
"url": "https://refract.group/"}
</script>
Schema markup in detail
- Also called
- Structured data, JSON-LD, Schema.org markup
- Updated
What it is
Schema.org is a shared vocabulary of types, such as Organization, Article, Service and BreadcrumbList, and the properties each type can have. Schema markup is that vocabulary written into a page. Google accepts three formats and recommends JSON-LD, a script block in the HTML that sits apart from the visible text, so it can be written and tested without touching the layout.
Each block makes statements about the page: this is an article, published by this organisation, last changed on this date, filed under these breadcrumbs. Nodes can refer to each other by an @id, so fifty pages can name one organisation without describing it fifty times. The markup should say only what the visible page says.
Why it matters
Google says it uses the structured data it finds to understand what a page is about, and some types make a page eligible for a richer listing, such as a breadcrumb trail in place of a bare URL. Eligible is the right word: no markup guarantees a rich result, and Google stopped showing FAQ rich results on 7 May 2026. What markup does reliably is remove guesswork. A company that shares its name with other businesses can state which one it is, where it works and which profiles belong to it.
How we do it
Our organisation is declared once, on the home page: the name Refract, with Refract Group as an alternate, the Sydney studio’s address, Flatiron in New York, the subjects we work on and the profiles we control. Every other page refers to it by the same @id, so search engines read the same Refract on every page.
A few small builders in our code write the rest, so each page type gets the type that fits it: Service on the industry pages, ProfessionalService on the location pages and Article on the comparisons. Each also carries its breadcrumb and its questions, word for word as the page prints them. We left founders, a phone number and a New York street address out, because nobody had confirmed them.
An example
Say a scheduling product has a page comparing its two plans. The page gets an Article node with a headline, a modified date and the company as author, and a BreadcrumbList from Home to Pricing to the comparison. It ends with five questions, so it also gets an FAQPage node that repeats each question and answer as printed. When a price changes, the table on the page and the markup change in the same edit.
Common mistakes
- Marking up what the page does not say. A rating or a price that appears only in the markup misleads every system that reads it, and the markup should describe the visible page.
- Adding FAQ markup for the rich result. Google stopped showing FAQ rich results in May 2026, so FAQPage markup now earns its place only by describing questions the page answers.
- Describing the company afresh on every page. Declare the organisation once with an @id and point to it, so its name, address and profiles cannot drift apart.
- Writing it before the copy is settled. Markup added to a draft goes stale as soon as the headline and answers change, so it is the last thing to add.
- Treating it as a ticket into AI answers. Google’s guide to its generative AI features says no special schema.org markup is needed to appear in them.
Schema markup vs meta tags
Schema markup describes what a page is and the things on it in a shared vocabulary, while meta tags give that one page a title, a description and instructions for crawlers.
| Schema markup | Meta tags | |
|---|---|---|
| Says | What the page is and who made it | Title, description, robots instructions |
| Format | JSON-LD using schema.org types | Single tags in the page head |
| Shapes | How the page is understood, rich results | The search snippet and the share card |
| Links pages | Yes, by @id | No |
Questions
What people ask about schema markup.
Google has not said that schema markup raises rankings. It says it uses structured data to understand a page’s content, and that certain types make a page eligible for rich results, which change how a listing looks. Treat schema markup as a way to describe the page accurately, and add it once the copy is settled.
FAQPage is still a valid schema.org type, but Google stopped showing FAQ rich results on 7 May 2026 and removed their documentation in June. FAQ schema now does what other structured data does: it describes the questions and answers the page prints. Add it where the questions help readers, and keep it word for word with the page.
JSON-LD is one format for writing schema markup. Schema.org supplies the vocabulary of types and properties, and JSON-LD, Microdata or RDFa carry it in the page. Google accepts all three and recommends JSON-LD, which sits in one script block apart from the visible HTML and is easier to keep correct.
Where this shows up in our work
Related terms
Working on this now?
Tell us what you’re building and where it’s stuck. We reply within 12 hours.
Most teams start with a call, then a sprint or an audit.
The people you’d work with, from the first call to launch