Everything you need to know about SEO, delivered every Thursday.
The Next Generation Of Structured Data: Taking Markup To The Next Level
Contributor Janet Driscoll Miller recaps an SMX West 2015 session covering the latest in structured data.
As SEOs, we’ve all heard for several years that we need to integrate structured data into our websites.
At SMX 2015, the speakers at “The Next Generation of Structured Data: Taking Markup To The Next Level” provided compelling new ways to use structured data beyond just your website and offered innovative ways to track the effectiveness of structured markup.
Justin Briggs: Innovative Applications For Structured Data
The first speaker in this session was Justin Briggs, CEO and Founder of Briggsby, who focused on some creative applications of structured data.
Search is changing — it’s becoming more mobile and more conversational. Up until now, optimizing for search was really focused on optimizing articles for information retrieval. But today with wearables, voice search and the Internet of Things, search is refocusing into becoming more of an interface that allows you to interact with data and take advantage of various functionality.
Ultimately, search will be able to predict your needs and do things on your behalf.
The knowledge that search engines must build to achieve to reach that goal, however, starts with the marriage of structured data and mobile applications. The steps toward that marriage have already begun, and Justin covered some of the methods that SEOs can use today to integrate information into mobile — beyond the website.
Deep App Linking
Deep app indexation is based on deep URIs, similar to how websites have URLs. By using schemas, you can define an action that can be performed on that deep-linked app content.
Deep linking to mobile app content provides you with a way to drive app engagement with the information inside of the app. Further, on mobile devices, users are asked if they wish to use that app by default for that content in the future, which may mean that they don’t have to visit your website again since they can access the information through the app.
And when the new mobile algorithm update drops on April 21, 2015, if your app is installed on a searcher’s device, it will rank higher in search results for that person.
Knowledge Graph = Action Graph
The Knowledge Graph in Google has been changing as well. It’s now not just a “knowledge” graph — it’s an action graph. Using structured markup, you can identify for Google what types of content are on your site, including content like music, and then direct Google where to take the searcher when they take action on that search.
You can also mark up events to show in the Knowledge Graph. Be sure to mark up the site with the detailed event information on it.
You can also use structured data to personalize Google Now cards. For example, Google can read your Gmail and, from that information, understand that you have a hotel reservation for a particular date and time.
Google can then surface relevant results when you’re performing searches (such as “nearby restaurants”) at that time — it doesn’t have to rely on your phone’s location services but rather the information from your hotel reservation.
Consider adding structured data to your emails for this purpose. Information like location and time/date are great for generating Google Now cards. Justin recommended referencing Actions In the Inbox. You can read more about structured data for emails on Google’s developers page on the subject. Events can also be pushed to Google Now cards.
Mike Arnesen: Measuring The Effectiveness Of Structured Data
Mike Arnesen, Director of Analytics & Optimization for SwellPath, challenged the audience to be testing the effectiveness of their structured markup. You’ve been told to do it, but how do you know it’s making a difference? Mike offered a truly innovative way to track effectiveness using Google Tag Manager and Google Analytics.
Google Tag Manager is composed of three basic parts: tags, rules and macros. Tags are the same as the tags you may already hard-code on your site, like Google Analytics tracking code or Google AdWords conversion pixels.
Mike referred to his approach and methodology as “semantic analytics” — applying the principles of the semantic web to your analytics data and collection methodology. So how can you implement semantic analytics using Google Tag Manager?
First, start with a macro to detect if semantic markup exists in the source code of a web page. Then, create a second macro that provides the semantic values for each type of markup you’re tracking. This macro should dynamically build the labels in Google Analytics.
Next, you want the rule to fire the tags. The first rule should look to see if semantic markup exists on a page. The rule will trigger the tag if the semantic markup exists on the page.
Finally is the tag, which triggers (based on the rule) if semantic markup exists on the page. This tag uses custom dimensions in Google Analytics with the semantic label/value.
Did it work? You can test your setup by applying semantic markup as a secondary dimension to see how these pages with semantic markup have traffic, how much and more.
A More Elegant Solution
The application layer is the CMS — where you house your data to create an experience (what the user sees). Between the application and experience later is the data layer, which surfaces data housed in your CMS databases in a usable way. The data layer is where Google Tag Manager and Google Analytics reside, for example.
Google Tag Manager is automatically set up to use macros to grab info from the data layer. Then create a macro to build the Google Analytics values.
Instead of having hard-coded values, create a custom HTML tag using macros as variables. Then you can ensure that when you add new pages to the website that fit a particular rule, the tag will be triggered automatically, saving you from having to code everything manually when you add new pages.
Google has a newly updated testing tool for semantic markup that also recognizes JSON LD. So don’t just implement structured data — track its performance!
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.