Schemas are basically a set of types, and each type has its set of properties. As defined by search engines Google, Yahoo, Bing, and Yandex, schema types are arranged in a hierarchal structure and till date, there are 818 Types, 1326 Properties, and 289 Enumeration values in the schema dictionary or vocabulary. Some of the common schema types are:
- Review, Aggregate Rating
- Product, Offer, Aggregate Offer
- Place, Local Business, Restaurant
To give search engine bots context of the content when indexing the web page, the schema is implemented in the code in the form of schema attributes which help the search engine bots identify the entities and their relationships. Schema attributes are basically information in the website code that make up the schema property. While the attribute is simplistic, the combination of attributes is complex and help define the schema property and its relationship. There are generally two types of attributes - Default attribute and Fixed Attributes. As the name suggests, a default attribute is what it is and cannot be changed while a fixed attribute is one which has a fixed value and cannot be changed. Examples:
The schema attributes can be implemented in the website's HTML code using either JSON-LD or RDFA syntaxes. However, Over the last few years, JSON-LD has become the most preferred structured data. Why? JSON-LD is well-implemented. It can be embedded in "script" tags in the head of a document (moving out of HTML body), separated from the data it describes. This makes JSON-LD a format that can be templated, easily implemented and a great option for adding, deleting, or updating purposes.
Other formats like Microdata or RDFa are neither that flexible nor easy to use options for adding, deleting, or updating purposes. RDFa are inline markups applied to HTML documents and define attributes that are added to the HTML.
Above all, search engines love and prefer JSON-LD and thus JSON-LDs are more picked up. Here is what Google has to say - "Google recommends using JSON-LD for structured data whenever possible". Period.
Once the schema code is implemented in the HTML using a Schema Generator, the schema code has to be validated for any errors or warnings using a schema testing tool. If the schema code has errors in case you have included a deprecated schema property, it has an incorrect structuring of the file, or due to other reasons, there will be no point publishing the schema on your website. Instead, rectify the errors and warnings before publishing it on your website.
To see how your content would look as a rich result on SERPs, run the JSON-LD block that contains the schema code on Google's Rich Result Testing tool.
Many businesses often assume that the job is done once the schemas are published on the website - error and warning free. However, this is not the case as the schemas implemented on the website need to monitored closely for errors or warnings that prop up - which can happen if there are any website content changes or Schema.Org changes or updates their vocabulary. Deprecated schema properties will be flagged as an error and for the content to be indexed and displayed on SERPs as a rich result, it should be error-free. Having a system in place to monitor published schema and having impact reporting for your business to assess the difference schemas are making on your website in terms of rich result growth and SERP saturation is as important as taking the first step of publishing schemas on your website.