Evan Sandhaus of the New York Times kicked it off with a very impressive look at how the Times spent the last three years developing and implementing structured data on the site. Called rNews, the data model is Schema.org compatible and intended to help the “machines” figure out page elements. And according to Sandhaus, to make Times web pages “pretty.”
Getting a look at why the New York Times decided to delve into the world of structured data made a lot of sense when Sandhaus began talking about all the elements on a news page. Forget just the main content and images contained in the article, there are also factors like comments, contributed by, provided by, and other elements a typical page doesn’t have.
What really was interesting was that a large newspaper has spent a lot of time (and likely budget) to make their search results stand out and it seems they aren’t done working on this project. Sandhaus noted they have taken to the community, traveling across the country to get feedback and improve the data model.
Aside from the impact on search results, the markup has also resulted in better analytics and insights into how people consume their news.
While Sandhaus talked about the process and the whys of structured data, Adobe’s SEO Manager, Warren Lee, talked about how they are using it in their products and some of the results they are seeing.
According to Lee, by adding Schema markup to their reviews, they saw a bump in CTR and a 20% increase in visits (Chris Silver Smith also noted in his presentation that structured data can result in CTR increases of 15%-50%. Wow!).
Lee also noted that while Schema has some really great value for search, it also has some value “beyond the click” with authorship. With search and social integration becoming ever more frequent and AuthorRank happening somewhere down the road, the rel=”author” tag has a lot of value.
My favorite takeaway from Lee’s presentation was his recommendation to make sure you understand regional implications when using structured markup. For example, in Japan reviews are typically used as complaint platforms. If you have a product you are marketing in Japan, reviews are probably not something you want to stand out in search.
The great thing about Schema is you can choose what you want to highlight on your site. While Lee mentioned that it was best for things like reviews, events, products and recipes, Chris Silver Smith showed some other ways to use it, including on your logo.
He also noted that when crawlers come to a web page, the first things they look for are Schema markup, open graph tags, meta tags, and then content. Don’t be afraid to use both Schema markup and Open Graph tags. Giving the crawlers more information can only help you.
Aside from the improvements in CTR and the appearance in search results, Silver Smith also made the excellent point that Schema markup can result in increased real estate in the search results, as well as in Twitter. For those unfamiliar with Twitter cards, you can now add a “card” to your tweet with things like article summaries, photos, app profiles, and more by adding meta tags to your page.
Each speaker made it very clear that if you are going to add structured data to your site, make sure you test, test, and test it again. Google’s structured data testing tool is a great feature in Webmaster Tools and will show you what your site will look like in the search results. Be sure to use it.
Interested in adding structured data to your site? Each of the panelists noted that ideally, you should develop it yourself but if you can’t, try out Google’s structured data markup helper. Another product within Google Webmaster Tools, the panelists described it as “brittle” but noted that it will allow you to easily mark up your content.