A data check asks a simple question: does a page tell people and software the same story? If an article says "By Atlas" but its code names someone else, something is wrong. Start with one live page. Then check every kind of page on the site.

It is like checking a book against its library card. A card with the wrong title or writer makes the whole shelf harder to trust.

Four checks: look at the saved article, compare the live page with its code, open the site's content lists, then check the report for all 102 pages.
We followed the same facts from the saved article to the live page, its code, the site lists, and the final report.

Start with one article

Open the page people actually see. Check the title, writer, date, and first answer. Then look at the code behind that page. Do the names and dates match? Google says the data in code should match what the page shows.

If the article is saved in a JSON file, check that file too. Fix a wrong name or date there first. Then reload the live page and make sure both the page and its code changed.

Do not stop at the home page

Check an article, a category page, a glossary entry, and each other kind of page the site has. Look at odd layouts too. Compare the pages you checked with the site's list of published pages. A test that misses half the site is not a full-site test.

Check the site's shelf list

Our public catalog lists eight groups of content, including articles and glossary terms. Each page has links to those lists in its code. Here is part of the code from this page.

A piece of code from this page. The links should lead to real content lists. Slide sideways on a phone to read the long web addresses.
{
  "@id": "https://www.ainowguide.com/article/structured-data-audit/#dataset-page",
  "@type": "WebPage",
  "subjectOf": [
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-content-library"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-pages"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-articles"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-categories"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-glossary-terms"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-glossaries"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-directory-categories"
    },
    {
      "@id": "https://www.ainowguide.com/ai/catalog.json#dataset-ai-now-guide-directory-listings"
    }
  ]
}

The long links lead to the lists in our catalog. Open a few. Are the lists there? Do they describe content the site really has? A working link does not tell us that Google or an AI tool used the list.

What our check found

On September 24, 2026, we checked all 102 pages in the site's published page list. Each page passed our check for eight public sets of content. No page failed. You can read the report and see what we checked.

That is what the report proves. It does not tell us Google put all 102 pages in search results. It does not tell us an AI tool read or quoted them.

Fix wrong facts first

Start with a title, writer, date, or link that does not match. Then fix missing code on other kinds of pages. Add extra details only after the basic facts are right.

The writing guide starts with the answer a person wants to read. The page checklist helps you check the rest. Run the test again when the site changes, and put a date on each new report.