What is extracted
On every check, Search Sentinel reads each <script type="application/ld+json"> block. Blocks inside @graph and arrays are taken apart. Each block is classified by its @type; 42 types are recognized, including Product, Offer, AggregateRating, Review, Organization, LocalBusiness, Article, NewsArticle, BlogPosting, WebSite, WebPage, FAQPage, HowTo, Recipe, Event, VideoObject, BreadcrumbList, JobPosting, Course, SoftwareApplication, Person, ItemList and Service. Other types are grouped as Other; a block that is not valid JSON is marked Invalid.
What is compared
Blocks are matched between checks by type and an identifier, like a Product's sku, gtin, @id or name, or an Article's @id, headline or URL. Three kinds of change are logged:
- Added: a block appears.
- Removed: a block disappears.
- Modified: a block changed. You get a field-level diff: the field path, whether it was added, removed or changed, and the old and new value.
Product "Trail runner 2" 3 fields modified
offers.price 89.95 -> 79.95
offers.availability InStock -> OutOfStock
aggregateRating.reviewCount 212 -> 219
The JSON-LD tab
Open a URL and choose JSON-LD:
- Current JSON-LD blocks: each block with its type, name and the formatted JSON. Blocks that disappeared show
Removedand when they were last seen. - Recent JSON-LD changes: the latest 20 changes with a summary like "3 fields modified, 1 field added", a table of field changes and a side-by-side view of the block.
Alerts for structured data
In the URL's Alert settings, under structured data monitoring:
Enable JSON-LD MonitoringAlert on Added Blocks,Alert on Removed Blocks,Alert on ModificationsSchema Types to Monitor: choose which types send alerts. All are on except WebPage and BreadcrumbList. Unselected types are still tracked in the history but do not send alerts.
Note
Search Sentinel checks that the JSON is valid, but does not validate it against schema.org or Google's rich result requirements.