Our Research Methodology

Six steps

1. Select the page type. Identify the template a question is about: home page, game page, article, help page, review.
2. Map the entities. Name the primary entity and the related ones with Schema.org types.
3. Review the markup. Read the structured data a template would produce and list gaps and conflicts.
4. Validate the structure. Check syntax, then Google's documented requirements, then the visible page.
5. Interpret relationships. Follow @id links and sameAs to see how entities connect.
6. Publish recommendations. Write dated guidance with sources and limits.

Order of sources

1. Google Search Central documentation and its documentation updates log, for anything about Google Search.
2. Schema.org type and property pages, for vocabulary.
3. W3C specifications, for JSON-LD itself.
4. Regulators and laws, for gambling and consumer rules.
5. Reputable trade press or law firm summaries, only when a primary source cannot be read, and named as such.

We do not use vendor marketing, unattributed statistics or forum claims as evidence. The full list is on data sources.

Dates

Every source line shows the date the source itself gives ("last updated", "published") or, where it shows none, "no date shown" with the date we read it. Pages show when they were published and last updated. All sources on the current version were checked on October 26, 2026.

The local checklist

The workbench, the registry statuses and every example on this site use one rule set of 43 checks, written from Google's documentation for the types we cover. It reports three levels: errors (the markup is broken or misses a Google requirement), warnings (likely problems, such as a self-serving rating) and notes (context, such as a retired feature). It is a teaching aid. It is not Google's Rich Results Test or the Schema Markup Validator, and it cannot certify eligibility for any feature.

The same checklist runs over SchemaReel's own structured data when the site is built; a page with an error or warning in its own markup cannot be published.

Illustrative data

Examples and registry records use invented brands and people on the reserved .example top-level domain. They show patterns we see in practice without naming or auditing any real company. They are labeled wherever they appear.

Limits we state

• No search volume, ranking or traffic data is measured or claimed.
• No real site is crawled or audited.
• Legal summaries are for publishers and are not legal advice.
• Google features change; each page records what the documentation said on the date shown.