Even if you’re not a fashionista, it’s likely that you’re aware of Gucci, the Milanese fashion house founded by Guccio Gucci in 1921. You may have picked up a discounted pair of Gucci sunglasses at an airport, and in the coming months, you’re likely to be bombarded with adverts for Ridley Scott’s ‘House of Gucci’ starring Lady Gaga and Adam Driver. That little double-G logomark is omnipresent from Montreal to Mombasa.
The name Gucci is instantly recognizable, but it can actually refer to many things: a company, a family, a style, a product range, a brand. There have been several legal disputes when it comes to defining the name, most notably in 2012 when the Gucci company successfully sued Guccio Gucci’s grandchildren (Guccio and Alessandro Gucci) for using their family’s name for a business.
If it sounds complex to you, spare a thought for Google, Bing, et al., whose business model is dependent on returning the right Gucci every time a user searches.
The issue is multiplied when you introduce voice search. Without a visual interface to scan, voice search typically returns the top result, meaning that the search engine’s algorithm can’t even hedge its bets. The next time someone asks it about Gucci, should it return the price of sunglasses, the location of the nearest outlet, or the screening times of movies?
Solving Search With Structured Data
Gucci isn’t the only term that search engines struggle with. In fact, unless your keywords are exceedingly niche, then there’s a good chance that it’s impossible to correctly pattern-match them on the characters alone.
The critical thing to understand about search engines is that they want the right users to find you. They’re not trying to trip you up or make you work hard to reach number one. If a user searches for a small charter boat named Gucci, then Google doesn’t want to give the user movie times, or sunglasses, or even Lady Gaga, it wants to deliver the charter page for a boat named Gucci.
A decade ago, search engines found that they were so bad at correctly pattern-matching a user’s raw input with abstract data that they banded together to introduce Schema.org, a structured way to add meaning to content.
Structured data allows search engines to correctly interpret the meaning of the data on your website, and perhaps more importantly, understand the site’s relationships to other web content.
With the structured data defined on Schema.org, search engines can understand the meaning behind data more easily. And data with meaning is simpler to match against search terms.
By adding structured data to our websites’ markup, we can help search engines return the right results.
Benefitting From Schema.org Data Types
Schema.org provides hundreds of different types, and they all inherit the properties of their parents. For example, the Festival type inherits from the Event type (a festival is a type of event).
The most important type is Thing because it’s the data type that every other data type is based on; every data type extends the generic Thing, and inherits its properties. The Thing type has basic properties like ‘name,’ and ‘description.’ Because every type inherits from Thing, every other type also has the ‘name’ and ‘description’ properties.
Structured data is essential for being picked up and displayed in rich snippets. For example, when a user searches for the House of Gucci movie, they may want to see information about the production, but they’re much more likely to want to read reviews or see screening times. Both the Reviews type and the Event type — or even better, the more specific ScreeningEvent type — allow search engines to deliver exactly what the user is looking for at the first time of asking.
Sample of rich snippets when searching the term “House of Gucci” on Google
How to Include Structured Data In Your Markup
While each option has suitable use-cases, RDFa is usually the least helpful when building a robust, scalable solution. Microdata can be very useful if you’re adding structured data to asides or snippets. For example, marking up a page about the House of Gucci, the primary structured data would be about the film, but you might want to additionally markup Ridley Scott with the Person schema.
Most often than not, JSON-LD is the best choice for marking up a page with structured data. Although JSON-LD can be injected inline, it tends to be used in the head of the document alongside your other metadata. This separation of data from content saves you from a tangled mess of spaghetti code. If, as the site grows, someone suddenly realizes that Ridley Scott should be marked up not as a Person, but as the more specific schema type Director (which is a subset of Person), it’s simple to make the change without having to hunt through your code for the correct lines to edit.
Does Structured Data Benefit SEO?
There is currently no evidence that structured data helps your site rank better. In which case, what’s the point?
What structured data does, is make your site more visible within search results by making it available to cards, knowledge graph, and other rich snippets that Google may introduce. This will significantly boost your click-through rate.
Additionally, by ensuring that Google understands the meaning and relationships of your data, you increase the chance that your site will be found only by those users to whom it is relevant, which will decrease your bounce rate.
Finally, structured data is heavily relied on by voice search. As voice search continues to grow, sites implementing structured data are more likely to be served to the searcher.
And so, while structured data is not considered a direct ranking factor, in practical terms, it has an enormous impact on how often your site will be seen online and how well it will perform in search engine results.