JSON-LD Schema for Rich Results - What to Add and What Not to Claim

Structured data tells search engines what a page is - a free web tool, an article, an organization, a breadcrumb trail - in a format machines can parse without guessing from the HTML. JSON-LD is the format Google recommends for most sites. It does not guarantee rich results, rankings, or AI citations. Valid markup makes a page eligible. Google decides whether to show anything. This guide covers what to add on a tool website, what not to invent, and how to build clean blocks with the JSON-LD Schema Generator.

JSON-LD vs microdata

Schema.org can be expressed three ways: JSON-LD, microdata (itemscope / itemprop on HTML), and RDFa. For marketing and tool sites, JSON-LD wins on maintainability.

• JSON-LD - one <script type="application/ld+json"> block (often in the head or before </body>). Markup stays separate from layout classes and React trees.
• Microdata - properties live on the same elements users see. Easy to drift when a designer renames a class or a component stops rendering a field.
• RDFa - rare on modern marketing stacks; skip unless you already maintain it.
• Google documents all three, but JSON-LD is the path Search Central examples use and the path this site's generator emits.

Pair structured data with the social tags people actually share. Open Graph controls the preview when a URL is pasted into Slack or LinkedIn - see the Open Graph Tag Generator and Open Graph Tags Complete Guide. Schema and OG solve different jobs; you usually want both.

The six types this site's generator ships

A 50-type mega generator invites empty optional fields and invented ratings. The ToolsMinify JSON-LD Schema Generator ships six types tool and content sites actually need:

• SoftwareApplication - free (or paid) web tools and apps: name, url, applicationCategory, operatingSystem, Offer with price.
• WebSite - site-level identity, optional SearchAction when you have a real on-site search URL template.
• Organization - About / homepage publisher identity (name, url, logo).
• Article - blog posts and guides with headline, dates, and author or publisher.
• BreadcrumbList - the trail you already show in the UI.
• FAQPage - only when the page renders a real question-and-answer section users can read.

Pick the type that matches the dominant content on the URL. Do not paste Organization on every tool page "for completeness." One accurate type beats three noisy ones.

SoftwareApplication for tool pages

A free online tool is software. SoftwareApplication is the natural type for a page whose job is to run a utility in the browser - the same way Article fits a blog post. For a free tool, set Offer price to 0, pick an applicationCategory (for example UtilitiesApplication or DeveloperApplication), and set operatingSystem to something honest like Any or WebBrowser.

• Required spirit: name, url, and a clear free (or paid) Offer. Empty offers and fake aggregate ratings are how pages get ignored or flagged.
• Match the visible H1 and product name. Do not invent a different brand string in JSON-LD.
• Do not add review stars, author credentials, or download counts you do not show on the page.
• Keep the script in sync when you rename the tool or change the canonical URL.

FAQPage pitfalls (especially after May 2026)

As of May 7, 2026, FAQ rich results no longer appear in Google Search. FAQPage remains valid schema.org markup, and other systems may still parse it, but you should not add FAQ JSON-LD just to chase SERP accordions. Those accordions are gone.

• Only emit FAQPage when the questions and answers are visible on the page.
• Do not hide FAQ copy in JSON-LD that users never see - that is a classic mismatch.
• Do not rename FAQPage to QAPage with search-and-replace. QAPage is for a different page model (user-submitted answers), not a publisher-written FAQ list.
• If the FAQ exists only for schema theater, remove the markup (or the section). Keep real FAQs that help users - for example steps on a tool page.

The generator marks FAQPage as markup-only for Google rich results so the UI stays honest about 2026 behaviour.

What not to claim

• Fake AggregateRating or Review blocks with invented stars.
• Author or medical/financial credentials you do not display (YMYL surfaces need real E-E-A-T, not JSON theater).
• HowTo markup for a one-line tip that is not a real step-by-step guide.
• Product schema for a free browser tool that is not a purchasable SKU - use SoftwareApplication with Offer price 0 instead.
• Duplicate conflicting types on one URL that describe different primary entities.

Markup must match visible content. If a field is not on the page, leave it out. The generator omits empty optionals for that reason.

Validate with Google Rich Results Test

After you paste a script tag into your template, verify it:

• Google Rich Results Test - paste the URL or code snippet and fix critical errors before deploy.
• Search Console - Enhancement / rich result reports for types you still care about (FAQ reporting was scheduled for removal after the May 2026 retirement).
• View source - confirm one coherent JSON-LD block and that @type matches the page.
• Crawlability - schema on a Disallow'd URL helps nobody. Keep indexable paths allowed in robots.txt; see Robots.txt Best Practices for SEO.

A minimal SoftwareApplication example

Conceptually, a free tool page needs something like: @context schema.org, @type SoftwareApplication, name and url matching the page, applicationCategory, operatingSystem, and offers with @type Offer, price 0, and priceCurrency. Build that form in the JSON-LD Schema Generator, copy the script tag, and paste it into your layout - nothing is uploaded.

Frequently asked questions

Does JSON-LD improve rankings by itself?

No. Structured data helps machines understand the page and can enable rich results when Google chooses to show them. Rankings still depend on relevance, quality, crawlability, and links - not on how clever your @type is.

Will schema get my tool cited in AI Overviews or ChatGPT?

Do not count on it. Schema does not guarantee AI citations. Eligibility for rich results is not a promise of answer-engine inclusion. Clear, useful pages that are crawlable matter more than inventing extra types.

Should every page have Organization markup?

No. Put Organization on the homepage or About page. Use SoftwareApplication, Article, FAQPage, or BreadcrumbList on the pages those types describe.

Is FAQPage still worth adding in 2026?

Only when you already show a useful FAQ. Google no longer shows FAQ rich results. Keep markup that matches visible Q&A for clarity and other consumers; do not add it solely for Google accordions.

Do I still need Open Graph if I have JSON-LD?

Yes. Open Graph (and Twitter cards) control social previews. JSON-LD is mainly for search and other structured consumers. Use both - generate OG tags with the Open Graph Tag Generator.

Build JSON-LD on ToolsMinify

Every tool below runs in your browser - free, no signup, nothing uploaded:

• JSON-LD Schema Generator - build FAQPage, Article, Organization, BreadcrumbList, SoftwareApplication, and WebSite script tags.
• Open Graph Tag Generator - create og:title, og:description, and og:image tags for social previews.
• Robots.txt Best Practices for SEO - keep schema-bearing URLs crawlable while you block only what must stay private.
• All SEO tools - the full category index.