The theme of this year's IA Summit was designing cross-channel user experiences. A cross-channel experience is one in which an individual will use more than one device to complete a task and will experience a seamless transition between devices to enable completion of the task. Google has recently referred to this as the multi-screen world in their mobile apps article.

The Kobo eReader is one of my favorite examples of a cross-channel experience. The Kobo device has Internet access that enables a user to save bookmarks and download books (amongst other things). The wonderful thing about the Kobo is that the eReader is just one device in a family of devices and applications that all speak the same language.

Kobo Family of Devices and Applications

I can download an eBook on my eReader and start reading at home. While waiting for the bus I can continue reading on my iPhone as the Kobo app knows where I stopped reading at home. Once at my destination, I can pull out my iPad and continue reading where I left off on my iPhone. 

The context of my cross-channel experience of reading a book is never lost during the process and I complete my task of reading the book using multiple devices or channels along the way.

This is all accomplished in part due to metadata. The book I was reading was an information object with extra data associated with it (such as the page I was last on) that was automatically updated as I read the book to preserve my context throughout my experience.

Adam Ungstad gave a fantastic presentation on the basics of metadata at the IA Summit. I felt he presented a very clear and concise way of introducing metadata to everyone and included all of the rules and concepts you have to think about when planning your metadata in order to enable cross-channel experiences.


There are three main types of metadata (as described by Adam):

  1. Descriptive
    • Ways to discover and identify information objects (color, shape, version, etc?)
    • This type of metadata enables context
  2. Administrative
    • How to manage an information object (expiry, creation date, modification date, content owner, etc?)
    • This type allows delivery of the right content to the right user, in the right channel, at the right time
  3. Structural
    • Information that's used to combine information objects (name can be composed of salutation, first name, last name, middle initial, etc?)
    • This type allows for interoperability in information exchanges

The metadata for your experience will fall into one or more of these three categories based on how the system intends to use its information objects. But the main enabler of cross-channel experiences is the implementation of metadata standards to ensure that each application or device is speaking the same language. Wikipedia gives a sample of some of the different metadata standards.

Adam specified three main components for defining a metadata standard:

  1. Semantics
    • One word could have many meanings. For example, a bow could be referring to the front of a ship, a person bending at the hips, a weapon which shoots arrows, or a kind of tied ribbon.
  2. Syntax
    • One meaning, many words. For example, a pepper is also known as a capsicum in some parts of the world
  3. Lexical rules
    • Defining what is and what isn't allowed. For example, it is not allowed to name your child Talula Does The Hula From Hawaii in New Zealand

Once metadata has been defined and a set of standards have been put in place to govern them, different systems are able to speak to each other using the same language, thereby enabling cross-channel experiences.

Karen McGrane's talk at this year's IA Summit was one of the highlights for me of the conference. Her presentation focused on two particular companies and how they handled consumption of content on multiple types of devices. One was a publishing company and the other a national syndicator of a network of radio stations.

The companies handled the situation in very different ways (you can view Karen's presentation for more details). Ultimately the most successful method was to focus on strategically structuring the content and creating an API that could deliver content via the use of focused metadata to the different channels that could potentially consume their data.

The organization truly separated their content from presentation by spending the time to structure it in multiple sizes, written for reuse, and with meaningful metadata associated with each piece. Through the API, their product owners were responsible for the presentation of the content instead of the content developers.

Much like how responsive web design appeared when traditional fixed-width designs did not display well on the plethora of mobile devices and screen resolutions that could potentially view a website, the idea of structuring your content for multiple device consumption has appeared as we have to struggle with how to best present content in so many different mediums.

This new landscape has forced us to prepare for a bigger world where support for every device (current and future) is no longer the most important concern. This is a pursuit that is neither realistic and nor sustainable.

Karen has a great series of slides in her presentation showing the evolution that print is no longer the center of the digital world, neither is the web, and neither is mobile. It all begins with content. She blames traditional CMS systems as part of the problem. She goes on to quote Dan Willis from his "Make it Semantic from the Start" article.

"Traditional publishing and content management systems bind content to display and delivery mechanisms, which forces a recycling approach for multi-platform publishing."

This is a very important statement. When you really look at it, the traditional CMS is one that promotes the creation of a full piece of content for one particular purpose. There isn't anything built-in that attempts to change it for other mediums. In a lot of cases, it encourages the author to combine the presentation with the content, which limits the uses of that particular piece of content.

Dan Willis continues on in his article to say that if we can design a CMS that encourages the creation of presentation independent, semantic chunks of content, then we can truly separate presentation from the content itself. The delivery mechanisms can then combine the content in a way that's appropriate for them.

In order to get there, Karen highlighted three main things for businesses to focus on:

  1. Write for the chunk, not the page
    • Focus not on how it will look on the page, focus instead on chunks of content that can be recombined and reused, in many different ways
  2. Syntax
    • Metadata can programmatically create pages, prioritize content on a page, and can personalize content
  3. Lexical rules
    • Currently a CMS workflow usually consists of writing content, sending it for approval, and publishing. Perhaps writing for the chunk can be incorporated into a workflow when creating the content?

The fortunate thing for us is that the rapid growth of mobile platforms has exposed a lot of opportunities for our industry to support this. Mobile should be used as a catalyst to prepare our content to be accessed from anywhere, from our fridges, to our phones, to our watches.

I'll leave you with the same quote that Karen did at the end of her presentation.

"I've never seen anyone regret having flexibility in how they deploy their content" – Jeff Eaton, @eaton