A Practical Schema Markup Audit Template for 2026
Semantic Summary
Idea: A schema markup audit checks whether the structured data on your pages is present, readable, accurate, and still useful, one page template at a time. A validator checks syntax. An audit also checks truth, conflicts, and what Google still supports.
Challenge: Teams run a validator, see green ticks, and stop. They miss wrong prices, duplicate blocks, markup that scripts remove, and effort spent on features Google retired.
Summary: Inventory templates, sample pages, test raw and rendered output, compare markup with the visible page, check Google status, score findings, and assign owners.
Related reads: GEO Audit Checklist · Content Governance in 2026
A schema markup audit is a structured review of the structured data on a site. It checks whether markup is present, valid, accurate, free of conflicts, and still useful for what Google supports. This template works one page template at a time and fits in a spreadsheet. It covers scope, sampling, checks, scoring, and a fix plan. It does not promise rich results or AI citations. This guide is marketing advice, not a guarantee of results.
What is a schema markup audit, and how is it different from validation?
Validation answers one question: can a tool parse this markup? An audit asks more.
| Question | Validation | Audit |
| Is the syntax valid? | Yes | Yes |
| Is the markup present after scripts and plugins run? | Sometimes | Yes |
| Does every value match the visible page? | No | Yes |
| Do two sources output conflicting blocks? | No | Yes |
| Does Google still support the feature? | No | Yes |
| Are names and identifiers consistent across pages? | No | Yes |
| Who fixes it, and when? | No | Yes |
A page can pass validation and still show a wrong price. Only an audit catches that.
When should you run a schema markup audit?
Treat it as a recurring process, not a one-time project.
- After major releases. Template, theme, or plugin changes often break markup silently.
- Before a redesign or migration. Record what exists, so nothing is lost.
- When Google changes a feature. Several types were retired in 2023, 2025, and 2026.
- On a schedule. A quarterly check suits most sites, and monthly suits stores and event sites.
What should the audit produce?
An audit that ends with a report and no owner changes nothing. Aim for five outputs.
- A template inventory. Page types, schema types, and the source of each.
- A findings log. Each issue, its evidence, and its severity.
- A prioritized backlog. Fixes ordered by risk and reach.
- A “do not fix” list. Markup tied to retired features, left alone on purpose.
- A retest plan. Dates and owners for rechecking.
How do you scope the audit?
Audit templates, not single URLs. A template is a page type that shares the same code, such as product pages, blog posts, or event pages.
- List page templates. Walk your site types: home, category, product, article, event, location, tool, help.
- Note the markup source. Theme, plugin, tag manager, hand-written, or a data platform.
- Sample pages per template. As a practical rule of thumb, check at least five representative pages, plus the newest and the oldest.
- Add edge cases. Out-of-stock items, cancelled events, pages with several categories, and pages with missing fields.
- Record the date and the tool. Results change as pages change.
Our technical SEO checklist covers the crawl and rendering basics that this scope depends on.
What columns should the audit sheet have?
Use one row per template and check. Keep the columns short.
| Column | What to record |
| ID | A short code, such as T-01 |
| Template and example URL | The page type and one live page |
| Schema types found | For example Product, Offer, BreadcrumbList |
| Markup source | Theme, plugin, tag manager, template, or manual |
| Raw vs rendered | Present in the HTML, only after scripts, or missing |
| Syntax result | Pass, warning, or error, with the tool used |
| Matches visible page? | Yes, partly, or no, with the field that differs |
| Conflicts or duplicates | Number of blocks and sources |
| Google status | Supported, optional, beta, or retired |
| Identifier consistency | Same names, URLs, and IDs across pages? |
| Severity and reach | Critical, high, medium, or low, plus the number of affected pages |
| Owner, fix, due date, retest date | One named person and dates |
How do you check each template?
Work through eight checks in order. Each has a clear pass condition.
| Check | How to run it | Pass condition |
| 1. Presence | View the page source and search for “application/ld+json” | Markup exists where expected |
| 2. Rendered output | Compare the raw HTML with the rendered page | Markup survives scripts and plugins |
| 3. Syntax | Run the Schema Markup Validator | No parse errors |
| 4. Google eligibility | Run Google’s Rich Results Test for supported types | Required properties present |
| 5. Accuracy | Compare each value with the visible page | Every fact matches |
| 6. Conflicts | Count blocks and sources per page | One source per type |
| 7. Status | Check the type in Google’s Search Gallery and updates page | Not retired, or marked optional |
| 8. Entities | Compare names, URLs, and identifiers across templates | Same facts everywhere |
The rich results test supports only some types, so use it for those and rely on the validator for the rest. Warnings usually flag missing recommended properties, while errors can block eligibility. Treat errors first.
How NEURONwriter helps: the AI Readiness audit in the AI Visibility module checks whether structured data is present on a page you enter, together with AI crawler access, robots.txt rules, and HTTPS. That covers check 1 quickly for a sample of pages. It reports what it finds. It does not validate the markup or compare it with the page, so the other checks stay manual.
What is the Google status register?
This register keeps the team from fixing what no longer matters. Update it each audit.
| Type or feature | Status in Google | Audit action |
| Product, Offer | Supported for product snippets and merchant listings | Check price, stock, and identifiers |
| Review, AggregateRating | Eligible for some item types; self-serving reviews are not | Check that reviews are real and visible |
| Event | Supported for public events at a physical location | Check dates, status, and venue |
| VideoObject | Supported | Check thumbnail, date, and page |
| BreadcrumbList | Shown on desktop only since 23 January 2025 | Keep, and match the visible trail |
| Dataset | Used for Dataset Search | Check description, licence, and download |
| Speakable | Beta, for eligible news publishers | Optional |
| HowTo | Rich results removed in 2023 | Optional; do not spend time fixing display |
| FAQPage | No rich results since 7 May 2026 | Optional; keep only if accurate |
| Course Info, Claim Review, Estimated Salary, Learning Video, Special Announcement, Vehicle Listing | Retired in 2025, with reporting ended on 9 September 2025 | Do not fix for display |
| Book Actions | Retired in 2025 | Do not fix for display |
Google said that the 2025 change did not affect rankings and that use outside Google Search was not affected. Unused markup does not cause problems for Search, so you do not need to remove it unless it is wrong. Check Google’s Search Central updates page and the Search Gallery before you rely on this table. Our post on HowTo schema explains one retired feature in detail.
How do you score and prioritize findings?
Use a simple scale and combine it with reach. This scale is a practical rule of thumb, not an industry standard.
| Severity | Examples | Action |
| Critical | Wrong price, availability, date, or rating; fake or hidden reviews; markup that misleads | Fix first; consider a temporary removal |
| High | Missing required properties on a key template; conflicting duplicate blocks; markup removed by scripts | Fix in the next sprint |
| Medium | Missing recommended properties; inconsistent names or identifiers | Plan and batch |
| Low | Optional or retired types; cosmetic issues | Leave, or fix when convenient |
Then rank by reach and business risk. A critical error on a template used by thousands of pages comes before a high error on one page. A wrong price matters more than a missing optional field.
What does a filled audit sheet look like?
Here are three sample rows with placeholder data.
| ID | Template | Finding | Severity | Fix |
| T-01 | Product page (example.com/products/sample) | Offer price in markup is 49.00; visible price is 59.00 (reach: all products) | Critical | Generate markup from the live price field |
| T-02 | Blog post (example.com/blog/sample) | Two Article blocks, one from the theme and one from a plugin | High | Disable one source; retest |
| T-03 | Help page (example.com/help/sample) | FAQPage markup present; no Google rich result since May 2026 | Low | Leave as optional; no display fix |
Each row has an owner, a due date, and a retest date in the full sheet.
How do you find duplicate and conflicting markup?
Duplicate or conflicting markup is common when a theme, a plugin, and a manual block all output the same type.
- Count the blocks. View the source, and count each type.
- Identify the source. Disable one source at a time on a test page.
- Compare values. Conflicting names, prices, or dates are the worst case.
- Keep one source. Remove the others.
- Retest. Confirm that the remaining block still renders.
Duplicate blocks with identical values usually do little harm, but conflicting values can confuse any system that reads them. Fix the conflict, not only the count.
Why is schema valid but not showing a rich result?
Valid markup does not guarantee a display. Common reasons include:
- The feature was retired. HowTo and FAQ no longer produce rich results.
- The page is not eligible. The item type, location, or content does not meet the documentation.
- The page is not indexed or was not recrawled. Check Search Console.
- The markup does not match the page. Mismatches harm eligibility.
- Google chose not to show it. Display is never guaranteed.
Which tools can you use to audit schema?
Use a small set of tools with clear roles.
| Tool type | Best for | Limit |
| Schema Markup Validator | General syntax and vocabulary checks | Does not check Google eligibility |
| Google Rich Results Test | Eligibility for supported Google features | Only some types |
| Search Console reports | Errors and warnings across crawled pages | Only types Google reports on |
| View source and developer tools | Raw and rendered output on one page | Manual, so sample only |
| A crawler with custom extraction | Counting blocks and types at scale | Needs setup; does not judge accuracy |
| A spreadsheet | The findings log and the backlog | Needs an owner to stay current |
No tool judges whether the facts are true. That step stays with a person who knows the product, the prices, and the dates.
Does fixing schema improve rankings or AI visibility?
Google does not document structured data as a ranking factor, and there is no published field-level evidence about AI use. Fix schema because accuracy and consistency matter, not for a promised boost.
| Claim | Evidence | How to treat it |
| Accurate markup can keep a page eligible for a supported Google feature | Documented by Google | Audit for it |
| A schema audit improves rankings | Not documented | Do not promise it |
| Fixing schema changes AI answers | Not documented | Test it with a controlled change |
If you want to test an effect, change one thing at a time and log the result. Our guide to AI visibility experiment design shows how.
How NEURONwriter helps: the AI Visibility module monitors the questions you add across Google AI Overviews, Google AI Mode, ChatGPT, and Perplexity. Record a baseline before a fix, and compare after. It tracks answers, not your markup, so keep the audit sheet as your record of what you changed.
What are common schema errors and their likely causes?
| Symptom | Likely cause | Fix |
| Markup missing in the rendered page | Script blocked, timing, or a tag manager rule | Check rendering; move to server-side output |
| Wrong price or date | Hard-coded values or a stale cache | Generate from live data; clear caches |
| Duplicate blocks | Theme and plugin both output markup | Keep one source |
| Missing required properties | Template does not include the field | Add the field or remove the type |
| Rating with no visible reviews | Markup copied from another source | Remove it, or show real reviews |
| Different names across pages | No shared identifier record | Create one source of truth |
Who should own the audit?
Give each template one owner, usually a developer or technical SEO lead, and one reviewer who knows the facts, such as the product or content owner. Agree on a short escalation path for critical findings. Treat markup as part of content governance, so changes to pages also trigger a check of their markup.
How do you perform a schema audit in common situations?
Four situations come up again and again. Each one changes how you test structured data.
- Ecommerce. On product pages, structured data markup must match the live price and stock. Use the workflow above on one template at a time, and treat any mismatch as a fix-first item.
- JavaScript sites. If a framework builds the page in the browser, run the audit on the rendered result, not only the source. Compare the raw HTML with the rendered page to confirm the presence of structured data after scripts run.
- Google Tag Manager. A tag manager can add JSON-LD without a code release. Check that the markup appears in the rendered webpage and that no second source adds a duplicate.
- Website redesign. Record what exists before launch, then audit the same templates afterward. Sitewide changes can remove markup from every page of a type.
Google replaced the older Structured Data Testing Tool with the Rich Results Test for supported features and with the Schema Markup Validator for general structured data testing.
When you are testing your structured data, use both, and remember that the schema.org vocabulary includes many types that Google does not use for SERP features. Use the validator for valid schema.org syntax, microdata, RDFa, and JSON-LD. Use Google’s Rich Results Test and Google Search Console when you want to know about rich snippets, rich result opportunities, and search results features.
Google recommends JSON-LD for most cases, and many teams now follow that advice in new structured data implementation. Keep the audit tied to your SEO strategy, but do not expect it to improve SEO by itself.
A clean audit may protect click-through rate where a feature applies, but no one can promise a CTR change. In the same way, entity connections across pages help only when the facts match everywhere. For AI search and generative search, treat accurate schema as one input among many, together with clear content and crawlable pages.
Think of a schema audit as a recurring check, like any other part of technical SEO. When you fix structured data on a key template, you improve SEO hygiene, and you reduce the risk of wrong facts spreading. You can audit structured data in a day for a small site, but a large site needs a staged plan. The goal is a short list of fixes with owners, not a long report.
Common schema audit mistakes
- Stopping at validation. Syntax is only the first check.
- Auditing URLs, not templates. One template fix covers many pages.
- Fixing retired features. Use the status register.
- Ignoring rendered output. Scripts can remove markup.
- No owner or date. Findings go stale.
- Promising rankings. No evidence supports that.
How does NEURONwriter help with a schema markup audit?
NEURONwriter helps you check structured data presence, draft markup for supported types, and test whether a change affects AI answers. It does not validate markup, compare it with the page, or replace the audit sheet.
- Check presence. AI Readiness checks structured data, AI crawler access, robots.txt rules, and HTTPS on a page you enter.
- Draft markup. Generate Schema Markup creates JSON-LD from page content. Its FAQ names Blog, HowTo, Review, Product, and FAQ as examples, so check the editor’s type list. It generates code from text, so verify every value against real data.
- Test effects. AI Visibility tracks Monitored Discussions across four AI engines, so you can compare answers before and after one fix.
- Plan the pages. SERP analysis and the content editor show whether the page that the markup describes covers what top pages cover.
You can run a keyword analysis in NEURONwriter before you plan your next audit follow-up. Treat every score as a guide, because no tool can guarantee rich results or citations.
FAQ
What is a schema markup audit?
It is a structured review of the structured data on a site. It checks presence, rendering, validity, accuracy, conflicts, and Google status, then produces fixes with owners and dates.
What is the difference between schema validation and a full schema audit?
Validation checks whether a tool can parse the markup. An audit also checks accuracy against the visible page, duplicates, rendering, Google status, and who will fix each issue.
How can I check or audit schema markup on my website?
List your page templates, sample pages, view the source for application/ld+json, run the Schema Markup Validator and the Rich Results Test, compare values with the page, and record findings in a sheet.
What is the first step in conducting a schema audit?
Inventory your page templates and the source of markup for each, so you audit templates and not single URLs.
How often should schema markup be audited?
After major releases and when Google changes a feature, and on a schedule. Quarterly suits most sites, and monthly suits stores and event sites.
Should schema audits be a one-time project or an ongoing process?
Ongoing. Pages, templates, and Google’s features change, so a one-time audit goes stale.
What are common schema markup errors, and how can I fix them?
Missing required properties, wrong values, duplicate blocks, and markup removed by scripts. Fix the template or data source, then retest.
Can duplicate schema hurt SEO?
Identical duplicates usually do little harm, but conflicting values can confuse systems that read them. Keep one source and fix conflicts.
Why is my schema valid but not showing a rich result?
The feature may be retired, the page may be ineligible or not indexed, the markup may not match the page, or Google may choose not to show it.
What is the difference between the Rich Results Test and the Schema Markup Validator?
The Rich Results Test checks eligibility for supported Google features. The Schema Markup Validator checks general schema.org syntax and vocabulary.
What is the difference between JSON-LD and Microdata?
JSON-LD is a separate block of code in the page. Microdata is written into the HTML elements. Google generally recommends JSON-LD.
Does schema markup improve rankings or guarantee rich results?
Google does not document it as a ranking factor, and it cannot guarantee a rich result. Treat markup as a way to stay eligible and accurate.
Does schema markup guarantee AI citations?
No. There is no published evidence that it does. Test the effect on your own prompts with a controlled change.
How should I prioritize schema audit findings?
Rank by severity, reach, and business risk. Wrong prices, dates, and ratings come first, then missing required properties on key templates, then everything else.
Which tools can I use to test schema markup?
The Schema Markup Validator, Google’s Rich Results Test, Search Console reports, view source, a crawler with custom extraction, and a spreadsheet for the findings log.
Is schema markup still relevant?
Yes, for supported features and for clear, consistent data. Several features, including HowTo and FAQ rich results, were retired, so check the status before you invest.
