Structured data is a machine-readable summary of what a page is about. You write it in a format like JSON-LD and map it to a shared vocabulary such as schema.org. It does not change the visible content of a page. It changes whether a search engine or AI answer engine can confirm what the page is, who published it, and which entity it describes. That confirmation is what most sites actually need.
Here is the surprising part: for many B2B sites, the biggest change from structured data is not a new rich result. It is entity clarity. Entity clarity means the engine can connect your page to a known thing, not just a string of words. If your site sells industrial components, entity clarity is what lets an engine say "this page is about a specific valve model from a specific manufacturer" instead of guessing.
This is a decision guide. I will answer five buyer questions in order, with one comparison table and a checklist you can run yourself. The goal is to help you decide when structured data matters and when it does not.
Does structured data change rankings, or just appearance?
Put simply, structured data does not change the core ranking score of a page. Google's own documentation states that structured data is not a ranking factor and does not change the ranking of your page. What it can change is eligibility for rich results, which are enhanced search listings like review stars, FAQ accordions, or product price and availability.
Rich results can change click-through rate because they make a listing more informative. That is an appearance change, not a ranking change. If your page already ranks at position 8, adding schema will not move it to position 3. It might make the position 8 listing more attractive to click.
For B2B sites, the practical effect is smaller than for e-commerce. B2B buyers often search by part number or standard, and the listing itself is already specific. The bigger gain is in AI answers, where the engine needs to extract facts, not just show a link.
JSON-LD vs Microdata vs RDFa: which format wins?
JSON-LD is a way to write structured data as a block of JavaScript Object Notation, separate from the HTML that users see. Microdata and RDFa are ways to write structured data inside the HTML tags themselves. Google's documentation recommends JSON-LD and lists it as the preferred format for most cases.
The reason is practical. JSON-LD sits in a <script type="application/ld+json"> block, so your content team can edit visible text without breaking the markup. Microdata and RDFa are mixed into the markup, so a template change can silently break them.
There is no ranking difference between the three formats. The engine reads the same triples either way. The difference is maintenance cost. For a B2B site with a CMS and a separate content team, JSON-LD is usually the lower-risk choice.
| Format | Where it lives | Main risk | Good for |
|---|---|---|---|
| JSON-LD | Separate script block | Can drift out of sync with visible text | Most B2B sites with a CMS |
| Microdata | Inside HTML tags | Template edits can break it | Static sites with no CMS |
| RDFa | Inside HTML attributes | Verbose, harder to review | Publishing and library systems |
If you already use Microdata and it validates, there is no urgent reason to switch. The migration cost is real. Switch only when a template change breaks the markup or when your team cannot maintain it.
Which schema types actually matter for a B2B site?
Most B2B sites do not need twenty schema types. They need a small set that matches how buyers search. The table below shows the types that usually earn their place, based on Google's structured data documentation and schema.org.
Article is the baseline for any blog or resource page. Product is relevant if you sell a physical item or a clearly priced service. Organization and LocalBusiness help entity clarity. FAQPage and HowTo no longer produce rich results in Google Search for most sites, but the markup still gives AI engines a clean question-and-answer pair to extract.
BreadcrumbList is cheap to add and helps engines understand site hierarchy. Review and AggregateRating should only be used when you have genuine, verifiable reviews. Marking up reviews you do not have is a policy violation, not a shortcut.
- Article - for news, blog posts, and technical explainers.
- Product - for physical goods and clearly priced services.
- Organization - for the company entity behind the site.
- BreadcrumbList - for site hierarchy and navigation.
- FAQPage - for clean question-and-answer extraction, not rich results.
That is five types. For most small and mid-size B2B sites, that set covers the practical ground.
How do AI answer engines use structured data?
AI answer engines do not publish a full ranking algorithm. What is public is that they retrieve passages from pages and generate answers with citations. OpenAI's documentation for its crawler, GPTBot, describes how it accesses sites, and Anthropic publishes similar guidance for ClaudeBot. Neither promises that structured data will earn a citation.
What structured data does is reduce ambiguity. If a page has Product markup with a name, brand, and SKU, an engine does not have to infer those from prose. If a page has Organization markup with a legal name and URL, the engine can connect the page to a known entity. That connection is entity clarity.
Schema.org defines the vocabulary. The W3C's JSON-LD 1.1 recommendation defines the syntax. Neither ensures an AI answer. They make the page easier to read correctly, which is a precondition for being cited, not a cause of it.
My judgement: for AI answers, structured data matters most when the page contains facts that are hard to parse from prose, such as specifications, prices, dates, and author identity. For a pure opinion essay, the gain is smaller.
When is structured data a waste of time?
There are cases where the work is not worth it. If your site has fewer than twenty pages and the content is plain prose with no products, dates, or specifications, the return is low. If your templates change every month, JSON-LD blocks will drift out of sync and create errors.
It is also a waste when the markup describes something the page does not show. Google's structured data policies require that the markup match the visible content. Marking up a FAQ that is not on the page is a manual action risk, not a win.
Another case: if your buyers find you through direct relationships and email, not search, the search-side gain is small. The AI-side gain may still exist if buyers ask an assistant about your category, but that is a different bet.
Key Facts at a Glance
- Structured data is a machine-readable summary of a page, usually written in JSON-LD and mapped to schema.org.
- Google states structured data is not a ranking factor; it affects rich result eligibility and appearance.
- JSON-LD is Google's recommended format for most sites.
- Five schema types cover most B2B needs: Article, Product, Organization, BreadcrumbList, FAQPage.
- For AI answers, the main benefit is entity clarity, not an assured citation.
- Test with Google's Rich Results Test and the Schema Markup Validator before publishing.
What to check before you ship
Run this checklist on one template first, then scale. Do not mark up every page at once. A single broken template can produce thousands of errors.
- Does the markup describe content that is visible on the page?
- Does the JSON-LD block validate without errors in the Rich Results Test?
- Does the Schema Markup Validator show the expected types and properties?
- Are entity names consistent with the Organization markup across the site?
- Is the author marked up on Article pages, with a real name and URL?
- Are review and rating properties only present when genuine reviews exist?
If you want the exact specification for your own site, ask your developer for the current JSON-LD templates and the validation reports. Those two documents tell you more than any general guide.
Does structured data still improve SEO in 2026?
It can improve appearance and eligibility for rich results, but Google states it is not a ranking factor. Treat it as a clarity and presentation tool, not a ranking lever.
What is a strong format for structured data?
JSON-LD is Google's recommended format for most sites, because it sits outside the visible HTML and is easier to maintain. Microdata and RDFa work, but they are more fragile when templates change.
How do I test my structured data?
Use Google's Rich Results Test for eligibility and the Schema Markup Validator for syntax and vocabulary checks. Run both before publishing a new template.
Can structured data help with AI search?
It can help AI engines read facts correctly and connect your page to a known entity. No major AI engine has published a promise that structured data earns a citation, so treat it as a precondition, not a promise.
What are common structured data mistakes?
Marking up content that is not visible on the page, using review markup without real reviews, letting JSON-LD drift out of sync with the visible text, and marking up every page before testing one template.