Anyone trying to implement schema markup to a large, multi-location business website is eventually going to find themselves with maintenance and scalability issues. Here, Schema App’s Martha Van Berkel explains the various approaches you can take to make this hugely worthwhile process much, much easier.
Why is Schema Markup Important to your Business?
Schema Markup is the language of machines. It was defined by Google, Yahoo, Bing, and Yandex back in 2011 to help them understand exactly what a web page is about. The result of adopting schema markup across your site is that the search engines and other machines truly understand the content, and match you with better searchers… or leads! But don’t take my word for it, Google has shown evidence of this over the last year.
At Google IO 2017, Google highlighted several case studies that showed the following outcomes:
- Pages displayed as rich results get 20%-82% higher click through rate than those that do not.
- Brands have experienced 1.5x more time spent on rich results pages and a 3.6x higher interaction on those pages.
Just this month, Google added three case studies into their search documentation. These all focus on results achieved by adopting schema markup.
- ZipRecruiter grows conversion rate 4.5x with the job search experience
- Eventbrite boosts traffic 100% with the event search experience
- Rakuten increases time on site 1.5x with the recipe search experience
While these metrics are great, the value of doing schema markup is much greater than just search metrics. Schema.org has become the de facto standard for machines trying to understand content. While search marketers are today adopting it to improve their brands’ organic search results, schema markup implementation also helps assistants understand your brand’s content, can augment your analytics, and provides understanding to your new chatbot.
While we all used to write our web content for machines (think keyword and site structure best practices for ranking), today we can write for humans, and use schema.org to ensure that the machines of today and tomorrow truly understand your brand and its content.
Code vs Knowledge Graphs
Now that you are (hopefully) convinced you should translate your website into schema markup, let’s talk about the difference of adding schema markup code, and creating a knowledge graph.
Schema Markup is code that sits on your web pages and is not seen by humans, but read by machines that translate the content of the page into a standard vocabulary. This code can be in the form of JSON-LD, RDFa, or Microdata.
A knowledge graph is a connected set of ‘things’, where the ‘things’ are defined explicitly, as well as how they relate to other ‘things’ (e.g. how a location is connected to a brand or a service offering is offered by a company).
The knowledge graph is a critical aspect of Google’s search technology and is the underlying fabric of the semantic web. It is also the vision of the future of the web, from Tim Berners-Lee, founder of the world wide web.
Berners-Lee describes the semantic web in the following way:
“The Semantic Web is about putting data files on the Web. It’s not just a Web of documents but also of data. The Semantic Web of data would have many applications to connect together. For the first time there is a common data format for all applications, for databases and Web pages.”
Why do I mention this? Well, that’s because the difference in creating schema markup code, creating contributions to the semantic web, and building a knowledge graph is in the practice of connecting your data to itself and other definitions on the web.
In a recent interview I did with Steve Macbeth (Sr. Manager, Microsoft, Executive Sponsor of Schema.org), he said:
“Creating connection points is very important because semantics is better than no semantics. Semantics without the ability to connect to other data is almost as valueless as no semantics. I mean, semantic data is only valuable, in my opinion, when it can be bridged to other data.”
How does this apply to you? As you create schema markup, for a single page or a large group of pages, think about how that ‘thing’ (e.g. Local Business), relates to other ‘things’ in your business, or other ‘things’ on the web. Connect these ‘things’, making your schema markup “semantic”. Take the time to create the connection points. It is here that your code becomes a knowledge graph.
Challenges doing Schema Markup: Scale and Maintenance
In March 2017, Schema App did a survey to gauge the “State of Schema Markup”. We asked schema markup practitioners around the world about the key challenges they face doing schema markup. Amongst the 75 respondents, scale and maintenance were two key themes.
It’s not surprising that these two challenges were mentioned, as they are highly related. If you are trying to do something at scale, and are unable to automate, maintenance at scale is the result.
In order to prepare you to tackle these two challenges as you attempt schema markup at scale (for a multi-location business, for example), I’ll detail the options for scalable schema markup and suggest tools to aid with maintenance.
Multi-Location Examples of Schema Markup at Scale
Before we jump into how you can do schema markup at scale, let’s discuss some real-life situations and how these Schema App customers overcame the challenge of scale and maintenance.
A national retailer was looking to add schema markup to their locations as well as their online products. They were challenged with a large backlog in their IT team and were looking for a way to do schema markup at scale, with the goal of using best practices and creating a knowledge graph for their brand, instead of just code.
They used Schema App’s data feed translation services to translate their location and services data into JSON-LD. Schema App converted these data sets into robust JSON-LD and deployed the data through Google Tag Manager. This took a total of three weeks.
Fortunately, because they used a tag to deploy the schema markup, they can make updates quickly, even though the first deployment took half a year.
Multi-location Fast Food Company: WordPress Plugin
A national fast food company with a WordPress website was looking to create location pages and optimize them with detailed schema markup. Since they had to create a WordPress plugin to integrate with their location data feed, they built the schema markup generation directly into the WordPress plugin. This way, when there are updates to their data feed, the WordPress plugin simply creates a new location page and automatically optimizes it with robust schema markup. Since it’s a custom solution, they will need to continue to work with their vendor to maintain and update their plugin.
For their blogs and additional pages, they use Schema App Structured Data Plugin.
Multi-location National Retailer with a Custom Platform: Highlighter and Tag Manager
A national retailer was struggling to get schema markup done on their custom platform. This was a result of limited IT resources and the agency of record not having access to make changes in the platform directly. Because they wanted to optimize the site for voice search in addition to Google, they opted to use the Schema App Highlighter.
The agency was then able to use the Highlighter to optimize the locations, products, blog, and category pages in a day, and deploy the schema markup through Google Tag Manager. The Tag Manager had been set up for analytics and so the agency had the ability to update and deploy tags without waiting for the development cycle. The site was optimized in less than two weeks, and now updates can be done in minutes.
Schema at Scale
When you are optimizing 100’s or 1000’s of pages with schema markup, there are a few different approaches you can take. The solution that works best for you will depend on the resources you have on your team, how you want to do maintenance, your web platform, IT procedures, and more. Below I’ve detailed the key approaches (and the pros and cons of each) to help you figure out what will work for your multi-location business.
When I speak about scale, some people dismiss the challenge because they have development resources at their disposal to build templates. This approach to schema markup at scale is quite common. In order to get schema markup across all their pages, they include the schema markup in the web page templates using microdata or they program a mapping of page data into JSON-LD.
Pros: Schema Markup on page, read by all search engines.
Cons: Requires a developer. Maintenance is manual. Markup can be broken with page changes. Updates can be slow depending on developers backlog.
Since the majority of people don’t have unlimited development resources, using an off-page highlighter to generate the schema markup is a great option.
Google Data Highlighter is a free tool from Google that allows you to map elements of the page into schema markup. Just submit and you’re done! However, Google’s tool is limited to a small part of the schema.org vocabulary, and once it’s submitted, the schema markup is not available to other search engines or voice assistants.
Pros: No development resources are required. Free. Easy to use.
Cons: Limited to basic schema markup use cases. No schema markup generated for other consumers (Alexa, Bing). No control of what pages are optimized.
Pros: No development resource required. All schema markup vocabulary. Ability to do complex schema markup (semantic). Easy to use. Creates Schema Markup for all consumers (Alexa, Bing, etc). Full control of what pages are deployed.
Cons: Paid solution.
Data Feeds: APIs, RSS, Merchant Feeds
A great way to do schema markup at scale is to use a Schema Markup translator that can take your data, map it into JSON-LD and allow you to deploy it. This is something you could build, or you could find a service, like Schema App, that does it for you.
What’s really fun about this approach is that if you use a common data feed, such as Google Merchant Feed, the translator service may already have the feed setup done, so you just have to pay for the translations, not the setup.
This approach of using a feed translator works great on large sites with 10’s of 1’000s of pages. We’ve used this translation service for HubSpot blogs, Ecommerce feeds, RSS deal feeds, company directories, and more.
A feed translation with a Tag Manager deployment contributes significantly to speed of deployment.
Pros: Ability to create detailed schema markup.
Cons: Requires a data feed, and therefore some IT/development involvement.
If you can’t do more complete automation or templates, then the best thing is to use a Tag Manager to deploy your schema markup. This can be Google Tag Manager, Tealium, Adobe DTM, etc. Using a tag manager puts the digital marketing team in control of the schema markup deployment and provides the ability to do maintenance and updates outside of a development release.
Pros: Marketers are in control of deployment. Deploy across templates without development release.
If you are on a content management system or platform with an ecosystem of plugins, you may find some availability to automate some or all of your schema markup, Great! Do it, but be sure to pick a plugin that adds unique schema markup to each page (e.g. don’t add ‘Organization’ schema markup to all pages).
In February, 2017, John Mueller shared the importance of adding unique schema markup to each page:
“So, in general, the structured data on a website, or on a page, should be specific to that particular page.” John Mueller, Google, Webmaster Hangouts
This makes total sense, since you want the search engines to understand what that specific page is about. If the same schema markup is on all pages, how does the search engine know which page is about (for example, the organization)?
Schema markup plugins for e-commerce sites
For e-commerce sites, look for a plugin or add-on that creates schema markup where the product versions, reviews, offer variances, etc. are nested within the schema markup. This means that the schema markup for the product shows up as one item in the structured data testing tool, rather than five. For reviews, it is important it is connected with the ‘thing’ that the reviews are about.
Schema markup plugins for blogs
For blogs, look for a plugin that uses all the fields from the blog to provide rich information. Also, if it is indeed a blog and not a news site, then make sure that the plugin optimizes the blog with ‘BlogPosting’, rather than the generic ‘Article’. ‘BlogPosting’ is a more specific type of ‘Article’ for blogs.
In a perfect world, the plugins would give you the opportunity to connect the organization/Local Business with the products and blogs so that you are creating a knowledge graph.
Pros and Cons vary depending on the plugin. Very likely the Pros are: easy to use, inexpensive or free, ‘out of the box’. Cons can include: hard to customize, rigid, or basic schema markup. It all depends on the plugin you choose.
Do you maintain your schema markup today? Most don’t, and that’s because it can be very difficult, especially at scale. But we all should be maintaining our schema markup, especially as it’s constantly evolving and changing.
Did you know?: In 2017, there were over 50 updates to schema.org and Google recommendations.
Do you know of all the changes, and did you make updates to reflect those changes?
Here are some tools to help you do maintenance of schema markup at scale, as well as some strategies to automate parts of the process.
Tools for Understanding Health of Schema Markup
Google Search Console
Google Search Console is the source of truth for what Google is seeing from your Schema Markup and the errors it is reporting. Within Google Search Console, go to the Structured Data Report (see below). This report shows you what schema markup is on the page, and what errors it is reporting. When you first add schema markup to your site, you will see the number of items increase.
Key Structured Data Events List
Semantic Search Marketing expert, Aaron Bradley, maintains a Google Sheet that tracks the events in schema markup. This spreadsheet shows a list of the changes in documentation across Google, Bing, and schema.org. While the spreadsheet is a great way to stay on top of changes, it is a very manual process to then update your schema markup.
Schema Markup Analyzers at Scale
When attempting to do schema markup maintenance at scale, it can be helpful to have the ability to analyze your schema markup and see the health of the markup (alongside actionable information). There are a few different analyzers available that provide this level of detail.
Check out the following tools to help you gain insights on your and your competitors’ sites’ schema markup:
Both of these tools are designed at scale. Schema App supports the entire schema.org vocabulary and extensions, and Botify has a segment of the schema.org vocabulary.
In an ideal situation you would have proactive monitoring setup for your schema markup, so that when something needs to be updated, you are told to do it, or it is done automatically. Today, Schema App provides error reporting on all your schema markup in one central location (see below). In time, Schema App will also provide notifications and ‘one-click’ to ‘auto-click’ fixes.
There are many different options for doing schema markup at scale. As I’ve detailed above, each approach has its pros and cons, and requires different skills to be successful.
No matter which approach you choose, I encourage you to think first about building knowledge graphs for your brand, rather than just code. This means thinking about how each entity (‘thing’) on your site is related and connected to others, and to other ‘things’ on the web.
Today, schema markup is foundational to your organic search strategy, in addition to being a fundamental part of the future of search, informing voice assistants and even virtual reality.
- How to Manage your Brand for the machine channel
- How to Automate Schema Markup at Scale with Speed
- Interview with Steve Macbeth, Schema.org Executive Sponsor from Microsoft
- Interview with Aaron Bradley, Semantic Search Marketer at Electronic Arts (EA)
- Schema Paths Tools
- How to create a Schema Markup Strategy
- Structured Data Testing Tool Error Guide
- 60+ Schema Markup Tools