Content Models: Getting Started with Structured Content

IC_ContentModels-01by Natalya Minkovsky (@hejhejnatalya)

In Intelligent Content: The Elephant and Its Parts, Marcia Riefer Johnston breaks down the definition of intelligent content into its components, one of which is structure. Marcia writes, “Technology can do its magic only after structure has been put in place. Technology doesn’t create the structure; people do. You could restructure your content today without buying a single tool.” In this post, I will share tips and ideas for introducing your organization to structured content using a technology-agnostic content model.

If you understand the basics of intelligent content but aren’t sure where to start, creating a content model – a framework that documents the structure of the content – is the first step to take to ensure that you’re setting your content up for success. And because content modeling can be tackled in small, manageable increments, it can be an excellent first step for an organization overwhelmed by a large volume of unstructured, uncategorized content.

Enter the content model

What is a content model? “A content model documents all the different types of content you will have for a given project. It contains detailed definitions of each content type’s elements and their relationships to each other,” Rachel Lovinger writes in Content Modelling: A Master Skill. The content model is the place where you capture the structure that will inform the way your content will be assembled, reconfigured, and reused.

For example, once you’ve determined that the “research report” content type is part of the overall model, you could propose that research reports include both a short title and a full title to make reports better adapt to social media sharing. You might include a mock-up of the way a properly structured research report would appear in various channels, like Facebook, Twitter, Google+, etc.

In your content model, you will want to include a list of the content types that will make up your digital presence together with the content and metadata elements that make up each content type. You should also document the relationships between the content types and the way the elements will be used, whether together or apart from each other (the “assembly model”). For example, a company’s overall content model might include “biography” and “speech” content types. The “speech” content type may comprise elements like speech title, event date, and speaker name. If the company president often gives speeches, her biography on the website may pull in the title and date of her three most recent speeches automatically.

That little word automatically is where the intelligent content magic comes in. Content that’s structured in the sense we’re talking about, and then coupled with tools that can take advantage of that structure, lends itself to automatic, strategic – intelligent – uses. IC_ContentProgramChart-01

Content model

This example of a high-level content model shows the relationships between various content types on a website. The boxes (biography, speech, blog post, etc.) represent content types. The arrows represent opportunities to automatically – dynamically – connect or reuse content. A program page might connect to speeches, blog posts, research reports, and FAQs related to that program. Speeches and blog posts might also connect to employee biographies. For example, recent speeches and blog posts by the company president might appear on her page. Research reports about a certain program might be aggregated into a filterable, sortable list.

The level of detail in the content model is up to you; you can use a diagram, spreadsheet, or another format that works for you and your team. The important thing to remember is that your content model should support your priorities (more on that in a moment) and should be tailored for your audience. The content model you hand off to a web developer will probably be quite different from the content model you show at meetings with senior leadership. A diagram allows you to focus on the big picture; a spreadsheet can be overwhelming by comparison. While your boss may want to know that you will be using microdata to mark up your content for Google and other search engines, your web developer needs to know which item types the content model includes and a mapping of content elements to itemprops.

Some good news: Chances are, you’re not starting from scratch when it comes to your content model. Even if the content hasn’t been broken down into its distinct attributes yet, and even if the metadata leaves a lot to be desired, every website has content that’s begging to be set free through structure. Yes, set free. Structured content implementation could come with some new rules and restrictions for authors and managers, like shorter character limits and more required fields. But rather than constrain the content, structure enables – frees – that content to be used automatically across multiple channels and devices and in a variety of previously impossible contexts. A piece of content can be created once and used in many places (aka single sourcing). That’s intelligence.

FAQs are one of those content types that naturally lend themselves to structure. While currently the FAQs on a given web page may be “blobs” of text from the computer’s point of view (text that has embedded formatting but no meaningful fields or structure), you already know that every FAQ is made up of a question and an answer, at a minimum. Each FAQ probably also has a category that describes what the FAQ is about, and may have other distinct elements like a product image or tutorial video. Similarly, if your organization puts out news releases, publishes research reports, or has a management team that gives speeches, add those to your list of possible content types to be included in the overall content model.

Think big, but start small

Remember my earlier point about the content model supporting your priorities? To make your initial foray into intelligent content more manageable, evaluate your content model and pick a workable place to start. Think big, but start small. While some changes may require extensive markup or even a new content management system, other improvements will be easy and inexpensive to accomplish. Why not start there? To help get buy-in for intelligent content at your organization, you may want to prioritize an area of content you know is dear to a decision-maker’s heart. You could also start with the content that you know you can start reusing right away so that you can demonstrate early on how intelligent content is saving your organization time and money.

Managers love a success story.

Once you know which content types you’ll be starting with, break that content into the distinct content and metadata elements that support your objectives. For example, if you want to increase engagement on Twitter, and your blog posts are shared on Twitter more often than other kinds of content, you might want to implement Twitter Cards for your blog posts. To implement Twitter Summary Cards, you will need to add metatags to your site and make sure that the post titles and descriptions adhere to character limits set by Twitter. On a recent blog project, this meant adding a new “short title” field to the blog content type and providing more clear instructions to authors about how the description field would be used – relatively minor changes. Keep breaking off pieces of the overall content model, working on executing one or two content types at a time until your organization is ready for a complete content overhaul, a new content management system, or another major initiative.

The classics are classics for a reason

Adaptive, reusable, and discoverable, oh my! There are a lot of intelligent content concepts to keep track of. Classic examples help bring these concepts to life and make them memorable.

Like the FAQ content type I mentioned earlier, the recipe is a classic content type. It lends itself to structure: ingredients, instructions, cooking time, and nutrition facts are all distinct attributes of a recipe. Recipes have become my go-to classic example of intelligent content at work because recipes are everywhere: blogs and websites, mobile apps like Epicurious, social networks like Pinterest, and in search-engine results. (I’m not alone in this. Many content strategists, including Ann Rockley in her classic book Managing Enterprise Content, use recipes as a way of clarifying the basics of structured content and its potential for intelligent uses.) It’s easy to separate the “intelligent recipes” from the not-so-intelligent ones by looking at how effectively a given recipe has adapted to a variety of devices or media channels. search_lasagna

Search-engine-results page for “lasagna”

A well-structured recipe traditionally includes elements like ingredients, instructions, photo, cooking time, calories, and nutrition facts. Online recipes might also include social elements, such as average rating and number of reviews.


Show them what you’ve got

As you’re getting started with intelligent content, even more important than telling your management about the benefits is showing your content in action. Although it’s tempting to use the classic examples of recipes and music genres to demonstrate structure and metadata, some stakeholders respond best when you can show them how the concepts would apply to their content. Take a real piece of content – a blog post, a product FAQ, a research report – and demonstrate how you would break it into smaller pieces and use metadata to support your company’s priorities and strategies.

This is where content strategy comes in. Decisions about metadata, structure, etc., must support the company’s business goals. Metadata for metadata’s sake helps no one.

Keeping the previously established priorities in mind, involve your key stakeholders in demonstrations. You might ask your stakeholders to share a blog post on Twitter and interact with the resulting Summary Card on their mobile phones. Giving people visual proof of the immediate impact of structured content will take you far in getting them on board. Twitter_Card

Twitter Summary Card

This Twitter Summary Card previews a Pew Research report’s image, headline, and description – elements defined in the report’s structure.

Conclusion and words of encouragement

Whenever you get an opportunity to present and discuss your content model, include reminders of the benefits and ROI of intelligent content:

  • Reuse: Content reuse may constitute a business case in itself, considering the efficiencies and cost savings that can be achieved as more and more content components are reused.
  • Multi-channel publishing: Multi-channel publishing is another area for improvement. Separating content from presentation helps get us closer to the desired “create once, publish everywhere” method.
  • Translation: For organizations that provide content in multiple languages, translation is a huge opportunity for intelligent content ROI.
  • Accuracy and integrity: Organizations that struggle to maintain the accuracy and integrity of content that is often copied and pasted, resulting in endless variations that are impossible to keep up to date and consistent, can look to intelligent content to address that pain point.

Don’t be afraid to roll up your sleeves and get started with intelligent content. To quote Cleve Gibbon: “No model survives first contact with real content … So go ahead and break your models, test them by running various scenarios using real content. Inspect, then adapt your models because that’s design. This kind of model design is best done early and often.” Once your organization sees some practical examples of the structured content model in action, ideas and enthusiasm for doing more with intelligent content will grow.

We would love to hear from you: How is your organization getting started with intelligent content? Please share your tips, travails, and triumphs in the comments.

About the author

Natalya MinkovskyNatalya Minkovsky is a writer and content strategist who lives and works in Washington, DC. She’s an organizer of DC’s Content Strategy Meetup and a department editor for the Content Marketing Institute’s Chief Content Officer magazine. Natalya spends a lot of time thinking about grammar, plain language, taxonomy, open source technology, and user experience design.

Learn more

Ready to learn more about content models and structured content? Check out these articles:

Intelligent Content Conference 2015

Want to learn more about structured content, content models, and intelligent content in general? Here are three things you can do right now:

Conference sessions related to this post’s topic:

Remember, when you REGISTER for the Intelligent Content Conference, use discount code ICC100 to save $100.

Images courtesy of Joseph Kalinowski/Content Marketing Institute

  • Scott Abel

    This a nice, brief look at some of the value of content models. However, content models play a much more critical role in the automatic exchange of both narrative and transactional content.

    In fact, it’s the incompatibility between content models that introduce challenges. This is a huge and often overlooked fact that needs some attention. Unless you are a specialist paid to translate and transform the flow of information between disparate content models, incompatible models introduce innecessary hurdles and expense. And, they prevent the true magic of intelligent content from being realized.

    If you create a content model for a business that is incompatible with other departments in the same organization — as well as incompatible with other organizations — you now have a new problem: content cannot easily be aggregated together, compared, and acted upon by machines in meaningful ways.

    Content models encoded in XML can drive applications, power transactions, enable commerce, update inventories, create invoices, process and collect payments, arrange for and instigate delivery of electronic and physical goods and much more.

    I’m excited to see this topic get the attention it deserves. The overarching value of content modeling is really not about web sites or recipes (although these common examples help us begin to understand their importance), instead content modeling is the beginning of the creation of hyper-efficient, data-powered business that leverages the power of a common model to describe our content — what is is (and what it isn’t), what it means (context), how it will be represented, and what other types of content are related to it.

    I’d love to hear what you think. Please consider adding your views to the comments section.

  • Marcia Riefer Johnston

    Scott, I like your phrase “the power of a common model.” Reminds me of the power of standards. Lots of conversation (tug of war?) has to happen to arrive at the level of agreement required, but when done well, all that conversation pays off. You end up with something consistent that’s bigger than any one participant could have created—and everyone benefits.

  • Berry Braster

    A great blog about intelligent content! In my view it’s a bit too much focused on content used for websites, where if you take content used by industries like OEM, which is what we mainly deal with at Etteplan, there’s a very specific need for intelligent content, but on the level of how it should be used in the (after market) service industry. Providing end users like field technicians (of which we have hundreds of thousands running around worldwide, most of which being non-native English speakers with various levels of technical and cultural backgrounds) with intelligent content demands a content strategy that requires a different approach than what is described in this blog, although one thing certainly applies: a common model. The use of controlled authoring standards like plain language (or Simplified Technical English) in combination with a solid structure (e.g. DITA), and single sourcing is the basis for any content strategy, and by being able to connect them with other parts of the business (e.g. accounting, inventory, etc.) it should support the full life cycle of a product, and connect with other parts of the business.

  • Marcia Riefer Johnston

    Berry, thanks for the thoughtful reply. You’re right; web-based examples illustrate general principles that apply to non-web content, too. I’ll keep that in mind as we develop other examples.

  • Pingback: Content Models: Getting Started with Structured...()

  • Natalya

    Scott, I agree that interoperability is a major consideration for content modeling. I was glad to see your comment about it since I didn’t directly mention content compatibility in the post. In my current position, I work with government agencies, and recently become involved with a working group that’s tackling the compatibility issue. For government agencies, like the businesses you mention, being able to aggregate, reuse, etc. content from different departments and divisions would help a lot. If different government agencies could adopt the same model, the benefit would be even greater, given the need to collaborate on different topics and programs. I’ll also add that the bulk of my experience is with narrative, informational websites and with social media content, so I hope others will weigh in with comments about how they are approaching content models for other kinds of content and data.

  • Marcia Riefer Johnston

    Natalya, Let me know when you’re ready to tell your story about “tackling the compatibility issue.”

  • Pingback: The Next Generation of Content Marketing: 6 Insights About Intelligent Content()