Article Structured Data: Writing BlogPosting JSON-LD Step by Step

Article structured data is a JSON-LD block that tells search engines the headline, author, images and dates of a piece of writing. Google's documentation lists no strictly required properties for Article markup, but it recommends five: headline, image, datePublished, dateModified and author. Fill those in accurately, use ISO 8601 dates and full image URLs, and you have solid markup. The Article Schema Generator builds it from a simple form and only includes the fields you actually fill in.

Article, BlogPosting or NewsArticle?

All three are Schema.org types, and the last two are more specific kinds of Article. Pick the one that matches the content honestly:

• BlogPosting for posts on a blog. This is the generator's default.
• NewsArticle for news reporting.
• Article for everything else: guides, essays, documentation pages.

Search engines treat the three much alike, so getting the values right matters more. Choose with the Type menu.

Which properties are worth filling in?

Here's what each field in the generator produces and why it's there:

Field	JSON-LD property	Notes
Headline	headline	The only field the tool requires. Match the visible title.
Description	description	A one-sentence summary.
Page URL	mainEntityOfPage	Written as a WebPage with this URL as its @id.
Image URLs	image	One per line. A single URL is a string, several become an array.
Author name, type, URL	author	A Person or an Organization, with an optional profile link.
Publisher name, logo	publisher	An Organization, with the logo as an ImageObject.
Date published, modified	datePublished, dateModified	ISO 8601 only.

The headline is the only field you can't leave empty, because markup without it doesn't describe an article at all. Everything else is optional, and empty fields are simply left out rather than written as empty strings, which validators would complain about.

Getting dates and images right

Dates are where most hand-written Article markup goes wrong. JSON-LD expects ISO 8601, which means one of these shapes:

• A plain date: 2026-10-14
• A date and time with a time zone: 2026-10-14T09:30:00+02:00, or with Z for UTC

Formats such as 14/10/2026 or Oct 14, 2026 are ambiguous, and the generator rejects them with a clear error. It also checks that the date exists, so 2026-02-30 fails. Date published starts out filled with today's date, which is handy when you mark up a post on the day it goes live, but change it if you're tagging an older article. If the modified date is earlier than the published date, you'll get a warning, since that's almost always a typo.

Images must be full http or https URLs, not relative paths like /images/cover.jpg. Google suggests offering large images in several aspect ratios (16:9, 4:3 and 1:1), so if you have three crops of the same cover, paste all three on separate lines.

A worked example: marking up a coffee blog post

Say you published a post called "How to Brew Better Coffee at Home" on 2 October and updated it on 14 October. Fill in the form like this:

• Type: BlogPosting
• Headline: How to Brew Better Coffee at Home
• Page URL: https://example.com/blog/better-coffee
• Image URLs: https://example.com/img/coffee-16x9.jpg and https://example.com/img/coffee-1x1.jpg
• Author name: Sam Rivera, Author is a: Person, Author URL: https://example.com/about/sam
• Publisher name: Example Coffee Co.
• Date published: 2026-10-02, Date modified: 2026-10-14T08:00:00+01:00

The output box shows a BlogPosting object with mainEntityOfPage, an image array of two URLs, an author Person with a url, and a publisher Organization. The URL path itself, better-coffee, is the kind of short slug you can create with the Slug Generator.

With Wrap in a <script> tag ticked, the result is ready to paste into your page template. Untick it for raw JSON. Copy and Download sit above the output; the file is article-schema.html or article-schema.json depending on the checkbox.

What the generator checks for you

Errors stop the output until you fix them:

• A missing headline.
• A page, author, logo or image URL that isn't a full http(s) address.
• A date that isn't valid ISO 8601.

Warnings still produce output but point at likely gaps:

• No image URL, because article results use one.
• No author name, or no publication date.
• A headline over 110 characters, which is probably not the short visible title.
• A modified date before the published date.

Common mistakes

Marking up a different headline from the one readers see. The headline should match the title on the page, not a keyword-stuffed variant.

Forgetting dateModified after an update. If you rewrite a post, update the modified date in the markup too. Don't bump it for trivial edits just to look fresh.

Putting the author in as a plain string. The generator always writes author as a Person or Organization object with a name, which is the form Google's documentation describes.

Assuming markup brings special results. Valid Article data helps search engines understand the page, but it doesn't guarantee rankings or any particular display.

FAQ

Do I also need Open Graph tags?

They do a different job. JSON-LD is for search engines; Open Graph tags control link cards on social platforms. The Meta Tag Generator writes those.

How can I double-check the output?

Paste it into the JSON-LD Validator. It confirms the syntax, flags dates that aren't ISO 8601 and lists the types it found.

Open the Article Schema Generator next time you publish, and add the markup while the details are fresh.