Schema validation for SMEs: Cleanly checking JSON-LD

Schema validation means technically and content-wise checking structured data before Google, AI systems, or other services extract outdated or contradictory company information from your website. For SMEs, this isn't just a technical gimmick, but a matter of trust: If the company name, address, phone number, services, or opening hours differ in the JSON-LD file from those on the website, contradictory machine-readable data is generated.

I see this time and again in my work with small businesses in South Tyrol: The website looks clean, the design is appropriate, the texts are well-written – but in the background, structured data describes a different version of the company. An outdated company name, an incorrect phone number, an inaccurate Schema.org type, or LocalBusiness markup without a real location reference are enough to make machine-readable data less than clear.

Structured data does not replace clear positioning. Structured data translates a clear website into a language that search engines and AI systems can better understand.

If you want to first understand why structured data is relevant for company websites, it's worth taking a look at the basics. This article deliberately avoids theory and focuses on a practical testing routine: How do you find validation errors, warnings, and inconsistencies before incorrect data weakens your visibility?

Scheme Validation: The practical testing routine for SMEs

A good schema validation consists of four checks: technical syntax, Google relevance, ongoing monitoring, and manual content matching. Only when all four levels are consistent is your structured data reliable.

• Technical check: Is the JSON-LD syntactically correct and understandable according to Schema.org?
• Check Google: Can rich results be generated in Google search from the markup?
• Watch live: Does Google Search Console detect errors or warnings on real pages?
• Compare content: Does the machine-readable data match the visible content?

This last point is often underestimated. Testing tools can detect many technical errors, but a tool doesn't always understand whether your real company identity is accurately represented. That's precisely where strategic website work begins.

Step 1: Check JSON-LD in the Schema Markup Validator

If you want to validate JSON-LD, start with the Schema Markup Validator. Google describes this tool as a generic schema validator for all Schema.org-based structured data – without Google-specific feature warnings.

In practical terms, this means that the Schema Markup Validator checks whether your markup is fundamentally readable according to Schema.org. The tool is particularly helpful if you want to see whether your Organization, LocalBusiness, Article, Product, or Service markup is formally correct.

Here’s how to proceed:

• Open the Schema Markup Validator.
• Either test the live URL of your page or insert a code snippet.
• Check which schema types are detected.
• Record validation errors and warnings separately.
• For multiple entities, check that the parent company, location, person, and offering are clearly separated.

Important: A flawless Schema Markup Validator does not automatically mean that Google will display a rich result. The tool checks Schema.org in general, not the complete Google search result logic.

Step 2: Use the Google Rich Results Test for search result features

The next step is the Google Rich Results Test. Google describes this test as an official tool that allows you to check which Google Rich Results can be generated by a page's structured data. Google also documents the supported search result features in its own Structured Data Gallery.

The difference is simple:

• Schema Markup Validator: Checks general Schema.org markup.
• Google Rich Results Test: Checks whether Google-relevant rich results are fundamentally possible.

In the Google Rich Results Test, you should pay particular attention to three things:

• Error: Critical problems can prevent structured data from being used for rich results.
• Warnings: Warnings often concern recommended fields that are not mandatory, but improve quality.
• Found elements: Check if Google is actually recognizing the correct entities: Organization, LocalBusiness, Breadcrumb, Product, Article or Event.

For SMEs, the goal is not to force every possible rich result. The goal is to provide the most important business information accurately, consistently, and in a machine-readable format.

Step 3: Monitor Google Search Console for real live errors

Testing tools provide snapshots. Google Search Console shows you what Google recognizes on your actual website during normal operation. According to Google Search Console, reports on structured data or rich result types appear when Google finds valid markup in your Search Console property and that markup belongs to a supported rich result type.

In practice, you check the following in the Google Search Console:

• whether structured data is recognized at all,
• which rich result reports are available,
• which elements are valid or invalid
• which critical and non-critical problems Google reports,
• whether fixed errors really disappear after validation.

I recommend that SMEs don't open the Google Search Console daily. This usually leads to unnecessary anxiety. A more sensible approach is a fixed monthly check – and additionally, a check after every major change to the website, services, team, location, opening hours, or contact information.

Step 4: Match visible content with machine-readable data

The most important check is not automated. You have to compare the visible content of your website with the JSON-LD. Google states in its general guidelines for structured data that content not visible to page readers should not be marked up.

This rule is also strategically important. Machine-readable data must not generate a different version of your company. If the markup includes services, ratings, locations, or awards that people on the site cannot verify, a trust issue arises.

A simple manual reconciliation for SMEs looks like this:

• Company name: Does the visible content contain the same name as the Organization or LocalBusiness markup?
• Address: Do the street, postal code, city and country match the website, Google business profile and JSON-LD?
• Phone Number: Are the NAP data – Name Address Phone – formatted identically everywhere?
• Opening hours: Do the visible opening hours, Google business profile, and structured data match up?
• Service description: Does the markup only describe services that are also clearly explained on the page?
• Social media profiles: Does sameAs' information only refer to official, active profiles of your company?

If you need a clean data foundation, a Single Source of Truth helps SMEs : a binding source for company data from which the website, Google business profile, social media profiles and structured data are fed.

Typical schema errors in SMEs – and how to correct them

The most common problems arise from plugin automation, relaunches, outdated company data, or lack of responsibility.

• Incorrect schema type: A business is listed as an "Organization," although "LocalBusiness" would be more appropriate for a location-specific page. Correction: Check whether the page describes the entire company or a specific local location.
• Contradictory NAP data: Name, address, or phone number differ between your website, JSON-LD, and Google business profile. Risk: Search engines and AI systems may not recognize your business as clearly. Correction: Define a consistent format.
• Missing required fields: A rich result type does not contain all required fields. Risk: The markup is not usable for this search result function. Correction: Add the required fields according to the official documentation.
• Missing recommended fields: The markup is valid, but it incompletely describes the entity. Risk: Google receives less context. Correction: Add relevant recommended fields if the information is real and visible.
• Markup without visible content: JSON-LD tags services, ratings, or people that are not actually on the page. Risk: Loss of trust and potential policy issues. Correction: Remove the markup or make the content visible to users.
• Several organizations mixed together: Operator, brand, location, parent company, and partners are all lumped together in one markup. Risk: Machines cannot clearly distinguish between the entities. Correction: Describe each entity unambiguously and only link them where the relationship is valid.
• LocalBusiness misused: A purely online offering or a supra-regional brand is marked as a local business. Correction: Only use LocalBusiness if there is a genuine local business connection.

These errors demonstrate why our web design and development work is never just about the surface. A modern corporate website must guide people, provide orientation for search engines, and cleanly deliver machine-readable data.

The 15-minute test routine for your website

If you have limited time, check these points once a month. This monthly review routine is sufficient for many SMEs, as long as no major website changes occur.

• Minute 1–3: Open your most important company website and copy the URL.
• Minute 4–6: Check the URL in the Schema Markup Validator and note any technical errors.
• Minute 7–9: Check the same URL in the Google Rich Results Test and distinguish between errors and warnings.
• Minute 10–12: Open the Google Search Console and check reports on structured data or rich results.
• Minute 13–15: Compare the company name, address, phone number, opening hours and service description with the visible content.

You should conduct unscheduled checks after any changes. These include new services, new locations, changed opening hours, new team members, new reviews, new social media profiles, a relaunch, a plugin change, or changes to your Google business profile.

Why schema validation is also relevant for AI search

Search engines are no longer the only systems that extract business information. AI assistants, response systems, and research agents prefer clear, consistent, and verifiable data. Structured data doesn't guarantee mentions, but it reduces room for interpretation.

For a machine-readable website in the context of entity SEO, consistency is paramount: Your company must be recognizable as a distinct entity with its services, location, people, and credentials. Schema.org helps to technically describe this entity. Your positioning explains why this entity is relevant.

That's also why I don't consider schema validation in isolation. If the brand is unclear, the markup will also be unclear. If the website is inconsistent, JSON-LD exacerbates the inconsistency. Technology makes correct information machine-readable – it doesn't replace clarity of content.

When you need professional help

You can perform simple audits yourself. Professional support is advisable when dealing with multiple locations, multiple brands, multilingual sites, complex services, shops, reviews, events, or automated content.

Especially in owner-managed businesses in South Tyrol, I often see organically grown structures: an old website, a Google business profile, social media channels, industry directories, sometimes an online shop, sometimes a booking system. If these sources are not synchronized, contradictory company data results.

In our strategic consulting, we therefore not only examine the code, but first the data logic: Which information is binding? Which page is the reference? Which entities actually exist? Only then is the markup corrected.

Good structured data isn't a trick for better rankings. Good structured data is proof that your company is digitally well-organized.

FAQ on Schema Validation

How often should I check structured data?

For most SMEs, a monthly review routine is sufficient . Additionally, you should revalidate after any changes to services, opening hours, team, locations, ratings, social media profiles, or contact information.

Should I use the Rich Results Test or the Schema Markup Validator?

You should use both tools because they perform different tasks. The Schema Markup Validator checks Schema.org in general, while the Google Rich Results Test shows which Google-relevant rich results are fundamentally possible.

Does every validation error need to be fixed immediately?

Critical errors should be fixed promptly, as they can prevent rich results or trigger incorrect information. Warnings about recommended fields are often less urgent, but should be reviewed if the missing information is real, correct, and readily available.

Why isn't Google showing rich results even though my markup is valid?

Valid markup makes a page generally eligible, but it doesn't guarantee it will be displayed as a rich result. Google decides whether to display an enhanced search result based on the search query, quality, guidelines, content, device, and other signals.

Is it permissible for JSON-LD to contain content that is not visible on the page?

No, for Google-relevant structured data, the markup should reflect the visible content of the page. If JSON-LD mentions services, people, ratings, or locations, users must be able to understand this information on the page.

What data is particularly important for local SMEs?

Consistent NAP data is particularly important for local SMEs : Name, Address, Phone. Company name, address, phone number, opening hours, and service description should match on the website, JSON-LD, and Google business profile.

Conclusion: Schema validation is digital due diligence.

Schema validation is not a task you can complete once and then forget about. Businesses change: services are refined, opening hours change, teams grow, locations are added, and positioning becomes clearer. Structured data must keep pace with this reality.

If you're unsure whether your website is properly readable, have the structured data professionally reviewed. Not because every SME immediately needs a complex schema system – but because incorrect machine-readable data spreads faster than many realize.

Sources

1. Google Search Central / Search Console Help: Rich Results Test – support.google.com
2. Google Search Central: Schema Markup Testing Tool – developers.google.com
3. Google Search Central: General Structured Data Guidelines – developers.google.com
4. Google Search Console Help: Rich result report overview – support.google.com