Last week at the State of Search Conference, there was no shortage of conversation around Schema Markup. From beginning to end, everyone was talking about the shift in search to include all kinds of machines (voice search, Google Home, Alexa, Echo, etc), and that with the emergence of these new platforms, the game is changing.
“Position Zero doesn’t matter, in voice search there is position only” Dr. Pete Meyer
Increasingly, people are getting their answers in search, meaning that they never get to our websites. Who is controlling these answers in search? Google. In order for Google to provide the most relevant answers, they need to be able to understand the content. The common language that Google and other machines speak, is schema markup.
In this article, I’ll share how Dr. Pete Meyer from Moz, Jon Earnshaw from PI Datametrics, Alan Bleiweiss and Duane Forrester of Yext, and Martha van Berkel (me!) from Schema App, told a story about the emerging need for Digital Marketers (not developers) to take control of their brand’s data. By building a brand’s knowledge graph, a brand can be understood, stay relevant, and perform in this new machine world.
Losing Control of Brands In Search
Today, we give Google content and we hope that we will get the traffic in return. But this may not happen in the new search, where voice search and platforms like Google Home provide answers, with no website traffic. Page one results don’t exist, or position zero, just the answer.
Similarly, Google no longer sends the user to the image on a webpage. Instead, Google shows the user the results and details in a cached result. While this is good for user experience, it means the image provider never gets the click.
The same experience is happening for Job Postings and Local Businesses. Google is curating the content directly into search results. This makes sense for Google, as a result of their curated content, users spend more time in their search engine, providing them more time to show ads. However, this does mean that we have lost control of our brand, since the customer experience is being managed by Google with our data, off of our websites.
Voice Search requires a Different Approach
If there is one place that the loss of control of our brand is VERY clear, it is in Voice Search. Jon Earnshaw, from Pi Datametrics, in his presentation Voice Search Tipping Point, shared that by 2020, 50% of all searches will be in voice. Just think, for 50% of all searches, only one person’s website will be highlighted. If you thought search was competitive now, with voice search it is even more critical to be the #1 result.
How is voice different? In addition to our behaviours evolving to adopt this new way of finding information, voice search moves away from searching for keywords such as “ribs dallas” and instead has us searching like human asking, “Where can I get the best ribs in Dallas?” This evolution means that searches will require elements of our brand to be understood in context so that they can respond to these human requests with answers, not just search results. In the example below, Jon has a conversation with the voice search engine to get answers.
So whether you are trying to perform in voice search and be the only answer, we must start answering real questions, in human language. We must go beyond keyword optimization and start optimizing with context (explaining what something is and how it relates to other things). If providing context means that we are defining things and their relationships to one another, we are actually saying we need to build and manage a knowledge graph.
Your New Job: Managing your Brand’s Knowledge Graph
Duane Forrester closed the conference with his presentation, stating that we all have a new job to be Digital Knowledge Managers. While I used the term “Managers of your brand’s knowledge graph”, we were saying the same thing.
How do you become a digital knowledge manager and what will be included in this new job description? In a world where there will be more phones than people, and everything in our lives will be connected (car, fridge, etc), the key will be to be able to manage how something relates to something else and define it explicitly. The practice of doing this is called building a knowledge graph.
A knowledge graph maps concepts and the semantic relations between concepts.
Managing Your Brand’s Knowledge Graph
As new Digital Knowledge Managers trying to stay relevant in the new machine world, we must think about how we manage our brand’s data and how it is being understood by machines. In my (Martha van Berkel) presentation, I shared how to Manage your Brand for the Machine Channel.
In order for machines to understand data, the data needs to be in their language, Schema Markup. To inform what schema markup to write, the first step is to define your brand’s knowledge graph.
Here’s how to build a knowledge graph, using my 1965 Austin Healey as an example.
Step 1: Define the Unique and Important things about your brand
Start by defining the important things about your brand. What do customers look for about your brand? What makes it unique?
Note: that this is not what the important webpages are on your brand’s website.
For my car, that meant that it was the star of a movie, that it is an Austin Healey, it is similar to an MG Midget and that it is the subject of the Austin Healey Club of America.
Step 2: Define Relationships
The next step is to define the relationship using Schema Markup between the important things about your brand. This includes linking to Wikipedia to define things explicitly. For example, my car was in the movie Losing Chase, as defined in Wikipedia here.
Tip: Figuring out how things are related can be difficult. Schema App CoFounder, Mark van Berkel, created the free tool, Schema Paths to help figure out how to relate two things in Schema Markup.
Step 3: Map Google Features
Next, you want to look at the Google Features for Schema Markup and see what applies to your brand. Most people start here and then don’t ask the important question, of what is important or unique about the brand. By making sure you include the Google features, you can enhance your search engine result pages, while also defining your brand for the machine channel. There are over 40 features to choose from to enhance your brand in search and email.
Usually, when you get your knowledge graph beautifully designed, you have to hand it over blindly to your development team to implement. This yields delays, and other challenges. Earlier this year, we interviewed people around the world who do schema markup by hand and asked them what challenges this introduced.
Challenges of Doing Schema Markup…the Old Fashioned Way
Challenge #1: Maintaining Schema Markup
Since January 2017, there have been 44 updates to the Google recommendations and Schema.org vocabulary. Imagine if you had to make manual updates to 100’s or 1000’s of pages with each of these changes. Wouldn’t it be great to have a dashboard where you can see the health of your markup and update it in minutes vs hours? Schema App gives you this visibility.
New Approach: This is the first driver to changing the way you manage your schema markup, moving to a centralized platform where you can make updates quickly, instead of trying to maintain hundreds of lines code, and be dependent on a developer.
Challenge #2 Showing Value
Showing the value of schema markup is not very different than how you show value from your SEO efforts, with a bigger focus on Rich Results. Over the last quarter, we have seen great improvements in reporting rich results, in Google Search Console Structured Data Report, in SEMRush SERP features report and in Moz’s SERP Features report.
New Approach: Add schema markup into analytics to slice your data by any property, an approach called Semantic Analytics.
Here we have added the blog author from our schema markup and can now see what Author on our blog drives the most traffic. This can be used to inform content strategy. Similarly, you could create a report on features of a product, and use the data to help drive product or sales strategies.
Challenge #3 Schema at Scale
If you are a developer, doing schema markup at scale, isn’t hard, you just manually add code to your site. However, this approach results in challenges in maintenance (remember the 44 updates you would have to do since January to stay up to date!).
New Approach: Define your strategy using your knowledge graph, and automate the rest. Today, Google has the Data Highlighter that allows you to centrally manage your schema markup, without writing code and automating the deployment. However, it only does basic schema markup and doesn’t allow you to connect elements of your knowledge graph across your brand and with Wikipedia. Schema App’s Highlighter solves this problem.
Another approach to doing schema markup at scale is to use a Data feed, such as Google Adwords data feed and converting it into Schema Markup, and using a Tag Manager (Google Tag Manager, Tealium, Adobe Tag Manager) to deploy it to your site. This approach re-uses existing data feed investments and automates the deployment in a way that the digital marketer can control. Schema App can convert any data feed into Schema Markup.
Read about other approaches for doing Schema Markup at Scale in this detailed article.
Watch the State of Search presentation, How to manage your Brand for the Machine Channel.
Google Uses Schema Markup to Understand
There was some skepticism about schema markup during the show. A few local marketers, said they weren’t convinced (based on data and results they have seen) that Google uses Structured data to improve local results.
In the interview with Gary Illyes from Google, Jennifer Slegg asked, “What tips do you have for people on the topic of Featured Snippets?” Gary answered that they are changing a lot, since they are still in development and they are trying to improve it. As a result, he didn’t have any tips to share.
After the session, I asked if Google uses the schema markup on the page for search. Yes was the answer, they use it all. This makes sense since the machines that drive search are trying to understand the content and schema markup brings context. Gary also said that he hopes in future they won’t need schema markup, that the machines will be smart enough to figure it out on their own. This won’t happen overnight, so in the meantime, we can use Schema Markup to explain to the machines what our brand is about.
Alan Bleiweiss, re-iterated in his presentation, Owning Answer Boxes, the Knowledge Graph, & Featured Snippets, the importance of using the right schema markup. If search engines are using it to understand the context and to inform their search and knowledge graph, then be sure to get it right.
The World is Changing
The world of search is changing. No longer is optimizing for Google enough, we must think about voice search, Echo, Chatbots, and even how day to day things like our cars, will understand brands. In order to do this, digital marketers have a new job. They must manage how their brand is understood by machines, which means they need to manage their brands knowledge graph, and learn how to do schema markup.
In order to do this in a way that allows you to stay in control, Digital Marketers should think about creating and managing their brand’s knowledge graphs, and finding smart and innovative ways to automate the rest.
The future is bright, and full of machines. Your job is to make sure your brands are understood in this new world, where websites are less relevant.