# Schema markup for B2B and enterprise sites: what to add, test and skip (2026)

> Schema markup won't lift rankings alone. See the types B2B, SaaS, manufacturing and public-sector sites need, how to test them, and what AI answers read.

Canonical: https://www.searchmaxxed.com/research/schema-markup-guide
Last updated: 2026-10-27

HOW-TO GUIDE

Schema markup won't lift rankings alone. See the types B2B, SaaS, manufacturing and public-sector sites need, how to test them, and what AI answers read.

BH

Brenden House

Founder of Searchmaxxed. We build agentic SEO and GEO workflows that help businesses rank on Google and get named in ChatGPT, then turn that search into more customers and revenue.

PUBLISHED 27 OCTOBER 2026 UPDATED 27 OCTOBER 2026 17 MIN READ

Schema markup is code, usually JSON-LD, that tells search engines what a page is about using the shared Schema.org vocabulary. For B2B, SaaS, manufacturing and public-sector sites, the set worth building is small: Organization, WebSite, BreadcrumbList and Article with a named author. Add Product, SoftwareApplication, JobPosting, Event or VideoObject only where the page carries that thing. Schema won't lift rankings on its own. Google uses it for rich results and to understand pages, and it says AI Overviews need no special markup. The hard part is keeping it true.

I'm Brenden House. I run Searchmaxxed, an SEO and AI SEO agency, and schema is part of the technical SEO we build and maintain for clients. Every Google rule below comes from Google's own documentation, read on 27 October 2026 (AEST).

## In summary

Most B2B schema value now comes from clean entity data, and star ratings and FAQ dropdowns matter far less than they did.

| What you're deciding | Short answer |
| --- | --- |
| Format | JSON-LD, rendered in the HTML your server sends |
| Every site | Organization, WebSite, BreadcrumbList, Article with a Person author |
| B2B SaaS adds | SoftwareApplication, VideoObject, JobPosting |
| Manufacturers add | Product, VideoObject, Event for trade shows |
| Public sector adds | GovernmentOrganization, JobPosting, Dataset for open data |
| Skip for Google display | FAQPage, HowTo and the features Google retired in 2025 |

- Schema doesn't raise rankings by itself. A structured data manual action costs you rich results, and Google says it doesn't touch how the page ranks.
- Google stopped showing FAQ rich results on 7 May 2026. That includes government sites.
- Google's May 2026 AI guide says there's "no special schema.org markup you need to add" for its AI features. Microsoft says schema helps Copilot understand pages.
- Mismatched markup is the expensive mistake. Generate schema from the data that renders the page, and test it on every single release before it ships.

## What is schema markup?

Schema markup is a block of code that labels the things on a page: the company, the product, the author, the job ad. Search engines read those labels to understand the page and to decide whether it qualifies for a rich result. People also call it structured data.

The terms you'll meet:

- **Structured data:**  facts in a fixed, machine-readable format.
- **Schema.org:**  the shared vocabulary for describing things on the web. Google, Microsoft and Yahoo founded it in 2011, and Yandex joined later.
- **JSON-LD:**  a script block of structured data that sits apart from your visible HTML, so you can change one without touching the other.
- **Microdata and RDFa:**  older formats that add attributes to HTML tags.
- **Rich result:**  a search listing with extras, such as a product price, a video thumbnail or a breadcrumb trail.
- **Entity:**  one real thing, like your company or your founder, that a search engine can recognise across many pages.

Google accepts all three formats. [Its own advice](https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data) is to pick the one that's easiest to implement and maintain, and "in most cases, that's JSON-LD". For an enterprise site with templates, a CMS and a development team, JSON-LD wins. It lives in one script block per page. Engineers can generate it from data, test it in a build and change it without touching the layout.

| Format | Where it lives | Fit for a large B2B site |
| --- | --- | --- |
| JSON-LD | A separate script block | Best: generated from data, easy to test |
| Microdata | Attributes on visible HTML tags | Works, but breaks when templates change |
| RDFa | Attributes on visible HTML tags | Works, mostly seen on older CMSs |

Here's a minimal Organization block for a made-up industrial software company:

```
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://www.example.com/#organization",
  "name": "Example Systems",
  "legalName": "Example Systems Pty Ltd",
  "url": "https://www.example.com/",
  "logo": "https://www.example.com/logo.png",
  "sameAs": [
    "https://www.linkedin.com/company/example-systems",
    "https://www.wikidata.org/wiki/Q00000000"
  ]
}
```

Most sites skip the `@id`, which is a mistake. It gives the company one permanent address in your markup, so every other page can point at it instead of repeating it.

## Does schema markup help SEO rankings?

No. Schema markup doesn't raise rankings by itself. [Google's structured data guidelines](https://developers.google.com/search/docs/appearance/structured-data/sd-policies) say a manual action for bad markup means the page "loses eligibility for appearance as a rich result". They add that "it doesn't affect how the page ranks in Google web search."

Where schema pays is the listing. [Google publishes case studies](https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data). Rotten Tomatoes added structured data to 100,000 pages and measured a 25% higher click-through rate on them. Nestlé measured an 82% higher click-through rate on pages that showed as rich results. Google doesn't date those studies. They're also consumer sites, built on films, recipes and reviews.

That's the catch for B2B. Most eye-catching rich results belong to recipes, products, events, films and reviews. A SaaS pricing page, a manufacturer's capability page or a council's service page has far fewer to win. I treat schema on these sites as two jobs:

1. **Win the few rich results you qualify for.**  Think breadcrumbs on desktop, product details, video, job postings and events at a physical venue.
2. **Make your entities unambiguous.**  One company, one set of people, products and locations, described the same way everywhere and linked to the profiles that confirm them.

That second job is the big one. [Google says Organization markup](https://developers.google.com/search/docs/appearance/structured-data/organization) helps it "disambiguate your organization in search results", and it can shape the logo shown in your knowledge panel too.

## Which schema types matter for B2B, SaaS, manufacturing and public-sector sites?

Every B2B site needs the same four base types. Each sector adds two or three. The table is the short version, and the sections after it explain the calls that trip teams up.

| Schema type | Put it on | Does Google show anything for it? |
| --- | --- | --- |
| Organization | Home and about pages, referenced everywhere | Logo and knowledge panel details |
| WebSite | Home page | Site name in results |
| BreadcrumbList | Every page below the home page | Breadcrumb trail, desktop only since January 2025 |
| Article plus Person | Research, guides, news | Better title, image and date handling |
| SoftwareApplication | A product page for the app | Rating rich result, if it meets the rules |
| Product | A page for one physical product | Product snippet details |
| VideoObject | Pages with demo or product videos | Video details in results |
| JobPosting | Each job ad | Google's job search experience |
| Event | Trade shows, conferences, open days | Event listing, physical venues only |
| GovernmentOrganization | Agency home and about pages | No Google feature; describes the entity |
| Dataset | Open data and research data pages | Google Dataset Search, not Google Search |

### B2B SaaS

A SaaS site's core schema is Organization, WebSite, BreadcrumbList, Article for the blog and SoftwareApplication for the product. SoftwareApplication needs care. [Google requires](https://developers.google.com/search/docs/appearance/structured-data/software-app) a `name`, an `offers.price` (zero for a free app) and either an `aggregateRating` or a `review`. A product page with no price and no visible reviews won't qualify. You can still describe the app. You just won't get the rating listing.

Then there are webinars, which catch a lot of SaaS teams out. [Google's event rules](https://developers.google.com/search/docs/appearance/structured-data/event) say "virtual experiences that have no real-world component aren't supported," so a webinar series earns no event rich result. Demo videos can still win through VideoObject. If you sell software, our [SaaS SEO work](https://www.searchmaxxed.com/industries/saas-seo) covers how we build and mark up those pages.

### Manufacturing

A manufacturer's best schema is Product on each real product page, filled with the brand, model, images and specifications that page shows. [Merchant listing markup](https://developers.google.com/search/docs/appearance/structured-data/product) is for pages that sell the item online, so a spec page with a "request a quote" button fits product snippets instead.

Organization does heavy lifting here. A manufacturer with several brands or divisions should define one parent Organization with its own `@id`, then connect the rest through `brand`, `subOrganization` or `parentOrganization`. Event suits trade shows with a physical stand. VideoObject suits product demonstrations. Our [manufacturing SEO](https://www.searchmaxxed.com/industries/manufacturing-seo) page goes further into the product and spec pages these sites need.

### Public sector

Mark the agency up as a GovernmentOrganization. Then add BreadcrumbList, Article, JobPosting and Event where each page fits. Two changes hit government teams harder than anyone:

- **FAQ rich results are gone.**  [From August 2023](https://developers.google.com/search/blog/2023/08/howto-faq-changes), Google showed them only for "well-known, authoritative government and health websites". On 7 May 2026 it [stopped showing them](https://developers.google.com/search/updates) for every site.
- **Special announcements are gone.**  Google deprecated that feature from 31 July 2025.

GovernmentService is a valid Schema.org type for a service page, but Google shows no feature for it. Use it to describe the service, not to chase a display. Dataset markup still matters for open data portals because Google Dataset Search reads it, even though Google Search itself doesn't (clarified November 2025).

### Enterprise and multi-brand companies

Large companies need a schema graph more than a longer list of types. In a graph, each entity gets one `@id` and every page points at it, from the Organization down to each author's Person node. Pages then say "this Article, published by that Organization, written by that Person". Running several brands or markets? Our [enterprise SEO](https://www.searchmaxxed.com/services/enterprise-seo) work maps how the entities connect before anyone writes markup.

## What schema do real B2B and government homepages use?

We checked. On 27 October 2026 (AEST), we fetched the rendered homepages of 14 well-known software, manufacturing and government sites and read their JSON-LD. Twelve carried some. Eight gave their nodes an `@id`, and seven linked to outside profiles with `sameAs`.

| Homepage | JSON-LD types found (main ones) | @id | sameAs |
| --- | --- | --- | --- |
| Atlassian | Organization, WebSite, WebPage, SoftwareApplication | Yes | Yes |
| Canva | Organization | No | Yes |
| Xero (Australia) | Organization, LocalBusiness, WebSite, WebPage, OfferCatalog | Yes | Yes |
| HubSpot | Organization, WebSite, WebPage, Product with AggregateRating | Yes | Yes |
| Salesforce | Organization, WebPage, FAQPage | Yes | No |
| SafetyCulture | Organization, WebSite, WebPage | Yes | Yes |
| ResMed (US) | Corporation, WebSite, WebPage, BreadcrumbList | Yes | Yes |
| BlueScope | Organization, WebSite | Yes | No |
| Caterpillar (US) | Corporation | No | Yes |
| Cochlear (Australia) | None found | No | No |
| ATO | GovernmentOrganization, WebSite, WebPage, Article | Yes | No |
| Service NSW | GovernmentOrganization, WebPage | No | No |
| USA.gov | Organization, WebSite | No | No |
| australia.gov.au | None found | No | No |

Two details stood out. HubSpot's `sameAs` list starts with its Wikidata and Wikipedia entries, which tells an engine exactly which company it means. Two of the four government sites, the ATO and Service NSW, use GovernmentOrganization.

Treat this as a snapshot. Homepages change often, and markup says nothing about how well a site performs. If your own homepage lacks an `@id` or `sameAs`, both take minutes to add.

## How do you keep schema markup matching the page?

Generate the markup from the same data that renders the page. Then make the release fail if they disagree. [Google's rule](https://developers.google.com/search/docs/appearance/structured-data/sd-policies) is short. Don't mark up content that isn't visible to readers, because a mismatch can cost the page its rich results through a manual action.

AI tools raise the stakes. SEO consultant Mark Williams-Cook put a fake address only in a test page's schema, and ChatGPT and Perplexity both repeated it, [according to Ahrefs](https://ahrefs.com/blog/schema-markup/). His reading was that the models treated the JSON-LD as plain page text. Either way, a wrong fact in your markup can end up in an answer about you.

Here's how we keep markup honest on searchmaxxed.com, using the same approach we bring to client sites.

- **One source per fact.**  Our build script defines the Organization, WebSite and founder Person once, each with a stable `@id`, and every page points to those nodes.
- **FAQ markup comes from the rendered page.**  The script reads question and answer pairs out of the finished HTML, so a question that isn't on the page can't reach the markup.
- **The build checks the result.**  A check fails the page if its FAQPage markup doesn't exactly match the visible FAQ, or if an article's author isn't the founder's Person node.
- **Citations match the visible sources.**  Article markup lists the outside sources the page shows in its own source list, and no others.
- **Markup ships in the HTML.**  We prerender JSON-LD into the page the server sends, so a crawler that doesn't run JavaScript still gets it.

You can copy this without our setup:

1. List every page template: product, solution, case study, article, job, event, location.
2. For each template, write down its schema type and which page field fills each property.
3. Build the markup in template code.
4. Add an automated test that compares key schema values with the rendered text on the same page, so drift shows up before release.
5. Re-test a sample of URLs after every CMS or template release.

This is our [technical SEO](https://www.searchmaxxed.com/services/technical-seo) work. Schema sits beside crawl, indexing, canonicals and site speed, and we check each fix on the live URL.

## How do you test schema markup?

Use Google's Rich Results Test to see which Google features a page qualifies for. The Schema Markup Validator checks the code against the full Schema.org vocabulary, which covers types Google ignores. Then watch Search Console. Each tool answers a different question.

| Tool | What it tells you | What it misses |
| --- | --- | --- |
| Rich Results Test | Which Google rich results the page qualifies for, with errors | Types Google has no feature for |
| Schema Markup Validator (validator.schema.org) | Whether the code is valid Schema.org | Whether Google will show anything |
| Search Console URL Inspection | What Google saw on its last crawl | Pages it hasn't crawled yet |
| Search Console rich result reports | Errors and valid items across the site, by type | Types without a Google report |
| Search Console Manual Actions report | Whether Google has penalised your markup | Mismatches nobody has flagged yet |
| View source or a crawler | Whether markup is in the HTML the server sends | Whether the facts are true |

To check whether a site has schema at all, open a page, view the source and search for `application/ld+json`. Nothing there, but the Rich Results Test finds markup? Then JavaScript adds it after load. Google copes with that, because it renders pages before reading them. Ahrefs reports that GPTBot, ClaudeBot and PerplexityBot don't run JavaScript, so those crawlers would miss it.

At enterprise scale, one URL at a time doesn't hold up. Crawl the site with a tool that extracts structured data, group the results by template, and compare each template's schema with its visible fields. One broken template breaks thousands of pages.

## Does AI use schema markup?

Some AI systems read it, some don't, and none requires it. Google [said in May 2026](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide) that structured data "isn't required for generative AI search". At SMX Munich in March 2025, Microsoft's Fabrice Canel [said schema markup helps](https://www.seroundtable.com/schema-llms-copilot-bing-microsoft-39093.html) Microsoft's LLMs understand content.

| Engine or system | What's on record about schema | Source and date |
| --- | --- | --- |
| Google AI Overviews and AI Mode | Not required; no special markup; keep structured data matching visible text | Google Search Central AI guides, May 2026 |
| Microsoft Bing and Copilot | Schema helps Microsoft's LLMs understand content | Fabrice Canel at SMX Munich, reported March 2025 |
| ChatGPT and Perplexity | Repeated a fake address that sat only in schema, read as plain text | Mark Williams-Cook test, reported by Ahrefs, 2026 |
| Gemini | Evidence it uses structured data while grounding answers | Dan Petrovic research, reported by Ahrefs, 2026 |

The Gemini and ChatGPT rows come from outside testing. They're evidence, not policy from Google or OpenAI.

My view? Schema helps an AI answer get your facts right. It doesn't earn you the citation. We put 300 Australian agency-buyer questions to ChatGPT three times each, with search on. It pulled 17,870 search results into those 900 answers and kept 4,389 as sources. Being found is the easy part. About three in four results it reads never make the answer.

The page decides the cut. A direct answer, specific facts and outside sources that agree with you all count for more than any markup you could add. So the schema job for AI search matches the one for Google. Mark up true facts, keep them consistent across your site and profiles, and spend the real effort on content. Our [AEO service](https://www.searchmaxxed.com/services/aeo) covers that wider work of getting named in ChatGPT, Gemini, Perplexity and Google's AI answers.

## Which schema markup should you stop adding in 2026?

Stop building for retired features. FAQ rich results ended on 7 May 2026, and HowTo rich results went in 2023. In June 2025, Google [announced it would retire](https://developers.google.com/search/blog/2025/06/simplifying-search-results) seven more features to simplify the results page, and it has since dropped others.

| Markup | What happened in Google Search | Date |
| --- | --- | --- |
| FAQPage | FAQ rich result no longer shown for any site | 7 May 2026 |
| HowTo | Rich result removed on mobile and desktop | August to September 2023 |
| SpecialAnnouncement | Feature deprecated | 31 July 2025 |
| Course info, estimated salary, learning video, vehicle listing | Features retired; documentation removed | Announced June 2025; docs removed September 2025 |
| ClaimReview | Fact check feature retired | Announced June 2025 |
| Practice problem | Feature phased out | January 2026 |
| Dataset | Used only by Dataset Search, not Google Search | Clarified November 2025 |
| Review markup about yourself | [Organization and LocalBusiness reviews](https://developers.google.com/search/docs/appearance/structured-data/review-snippet) only count on sites reviewing other businesses | Current guideline |

Book actions sat on the June 2025 list too, but Google lifted that warning in November 2025 because a feature still uses the markup.

Should you rip out FAQPage markup that already exists? Not if it matches the visible FAQ. It does no harm, and engines other than Google can still read it when they parse your page. We keep it on searchmaxxed.com because our build generates it straight from the visible FAQ. I wouldn't write a single FAQ for a rich result that no longer exists.

## How do you add schema markup to an enterprise site?

Build it into the page templates, generate it from the CMS fields that also render the page, and test it before every release. Plugins and tag managers suit small sites. On a site with thousands of URLs, several teams and a release process, schema belongs in code.

1. **Audit what exists.**  Crawl the site, export the JSON-LD by template, and note duplicates, conflicts and retired types.
2. **Map your entities.**  Write down the legal and trading names, brands, divisions, people, products and locations, plus the profiles that confirm each one.
3. **Assign types to templates.**  One table: template, schema type, required properties, the CMS field behind each.
4. **Build the graph.**  Define each entity once with an `@id`.
5. **Render it server-side.**  Put the JSON-LD in the HTML the server returns, so crawlers that skip JavaScript still see it.
6. **Gate the release.**  Fail the build when markup and visible content disagree.
7. **Monitor.**  Check Search Console's rich result and Manual Actions reports after every release.

A worked example. Say a B2B SaaS company has 40 integration pages, 12 product pages and 300 articles. Its integration and product templates get BreadcrumbList plus a WebPage node, and each WebPage points at the Organization's `@id`. The flagship product page also gets SoftwareApplication, but only once it shows a price and real reviews. Articles get Article with a Person author. That's three templates to change, not 352 pages to edit by hand.

## Where does Searchmaxxed fit?

Searchmaxxed is an SEO and AI SEO agency, and schema markup sits inside our technical SEO work. We map your entities, build the markup into your templates, test it on every release and check the live pages. It comes with every managed plan.

| Plan | Who it suits | How schema fits |
| --- | --- | --- |
| Growth System | A company growing in one or two markets | 30 page releases a month; schema, technical and internal-link work doesn't reduce that count |
| Agentic Search Engineering | A crowded or national category | Everything in Growth System, plus comparison, industry and integration page families, digital PR and AI answer tracking |
| Custom & Enterprise | Several brands, several markets or a large site | Dedicated engineering, custom reporting and original first-party research |

Here's how we do it. We build on a website you own, the Agentic Website. AI agents pull your Search Console and GA4 data, read the Google results and check what AI answers say about you. That data feeds the Managed Search Loop, which shows which page, fix or link should bring in the most enquiries next. We build it, verify the live URL and measure what changed.

Searchmaxxed's work has generated more than A$10 million in revenue through search for its clients.

## Is Searchmaxxed right for you?

We're built for marketing and growth leaders at enterprise, B2B SaaS, venture-backed, manufacturing and public-sector organisations. We suit teams who want schema done properly inside a bigger search effort, built and shipped by the people who plan it.

| Searchmaxxed is a good fit if | Look elsewhere if |
| --- | --- |
| You want schema, technical SEO and content from one team | You only want one schema snippet written |
| Your site has many templates, brands or markets to keep consistent | A plugin's default markup covers your small site |
| You want markup tested on every release, not audited once a year | You want a report rather than changes to your site |
| You care about Google and AI answers together | Your developers will build it all and only need a spec |
| You want to own the site, repository, content and data | You're shopping for the cheapest retainer |

Right-hand column? Google's free documentation and the testing tools above will take you a long way.

## Frequently asked questions

### What does JSON-LD stand for?

JSON-LD stands for JavaScript Object Notation for Linked Data, a way of writing structured data as a separate script block in a page's HTML. Google recommends it in most cases, because it's the easiest format to implement and maintain.

### Is schema markup a ranking factor?

No. Google says a structured data manual action removes a page's rich result eligibility but doesn't affect how the page ranks. Schema still helps in two ways. It can lift clicks when it earns a rich result, and it helps engines understand your company and pages.

### How do I check if my website has schema markup?

View the page source and search for `application/ld+json`. Then run the URL through Google's Rich Results Test for Google features, and through the Schema Markup Validator for the full vocabulary. Search Console then reports errors across every page Google has crawled.

### Does AI use schema markup?

Some systems do. Microsoft said in March 2025 that schema helps its LLMs understand content, while Google says its AI features need no special markup. Outside tests suggest ChatGPT and Perplexity read JSON-LD as plain text.

### Should I still use FAQ schema?

Google stopped showing FAQ rich results for every site on 7 May 2026. Existing FAQPage markup that matches a visible FAQ does no harm, and other engines can still read it. New FAQs for rich results won't pay.

### Can I add schema with Google Tag Manager?

You can. Google will usually read it, because it renders JavaScript. Ahrefs reports that GPTBot, ClaudeBot and PerplexityBot don't run JavaScript, so they miss markup added that way. Server-rendered JSON-LD is safer.

### What schema should a B2B website have?

Every B2B site needs Organization, WebSite, BreadcrumbList and Article with a named author. Then add SoftwareApplication, Product, VideoObject, JobPosting or Event, but only on pages that carry those things, because markup must describe what's visible. Give each entity an `@id` and link it to outside profiles with `sameAs`.

## What should you do next?

Open your homepage source and search for `application/ld+json`. Look for three things: one Organization node with an `@id`, a `sameAs` list of the profiles that confirm you, and no markup for unseen facts. If you'd like us to map your entities and build schema into your templates, [let's talk pricing](https://www.searchmaxxed.com/contact).

REFERENCES

1. [Its own advice](https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data)
2. [Google's structured data guidelines](https://developers.google.com/search/docs/appearance/structured-data/sd-policies)
3. [Google says Organization markup](https://developers.google.com/search/docs/appearance/structured-data/organization)
4. [Google requires](https://developers.google.com/search/docs/appearance/structured-data/software-app)
5. [Google's event rules](https://developers.google.com/search/docs/appearance/structured-data/event)
6. [Merchant listing markup](https://developers.google.com/search/docs/appearance/structured-data/product)
7. [From August 2023](https://developers.google.com/search/blog/2023/08/howto-faq-changes)
8. [stopped showing them](https://developers.google.com/search/updates)
9. [according to Ahrefs](https://ahrefs.com/blog/schema-markup/)
10. [said in May 2026](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide)
11. [said schema markup helps](https://www.seroundtable.com/schema-llms-copilot-bing-microsoft-39093.html)
12. [announced it would retire](https://developers.google.com/search/blog/2025/06/simplifying-search-results)
13. [Organization and LocalBusiness reviews](https://developers.google.com/search/docs/appearance/structured-data/review-snippet)

[Get an AI visibility audit](https://www.searchmaxxed.com/contact)

See where buyers find someone else across Google and AI answers.

See where you are invisible →

## Let's make you the answer.

## Related

- [Research](https://www.searchmaxxed.com/research)
- [Searchmaxxed](https://www.searchmaxxed.com/)