User Centric Content Models

Content doesn’t organize itself.  That’s why we have content models.  

A lot of advice about creating content models misses an important dimension: how the user fits in. Many content models are good at describing content.  But not many are very user centric.  I want to suggest some simple steps to help make content models more centered on user needs.

Two popular ways of thinking about content models are (1) that the content model is like a database for content (the technical approach), or (2) that the content model is a structural representation of a massive document (the structured authoring approach).  When combined, these approaches transform a content model into a picture of documents-as-a-database.   

Content models generally focus on showing what information is relevant to various topics.  Some models can be very sophisticated at representing the publisher’s perspective, and all the details it might want to manage.  But even in sophistical models, the needs and motivations of audiences are hard to see.  

Content models show numerous fields and values.  Each topic could become a screen that could be configured in various ways.  One CMS vendor says of content modeling: “it’s very similar to database modeling.” 

But actually, designing content to support user goals very different from designing  database to store records.  Databases are a bad analogy for how to model content.

Audiences don’t want to read a database. Even if they are interested in the topic.  A database is fine for scanning for short bits of information to get quick answers. It’s less good for integrating different fragments of information together into a meaningful whole. People need support bridging different fragments of information. 

A content model should aim to do more than show a picture of how topics can be broken into chunks.  

Neither are content models about navigation paths, as if they were a site map.   True, different chunks, when linked together, can allow users click between them.  It’s nice when users can jump between topics.  But it’s not clear why users are looking at this content to begin with. Many models may look like a collection of linked Wikipedia articles about baseball teams, baseball players, and pennant races.  It’s a model of what we could call brochure-ware.  It’s a database of different articles that reference one another.  The connections between chunks are just hyperlinks. There’s no obvious task associated with the content.  

What Users need to Know

Most explanations of content models advise publishers to model stuff that people might want to know about. I call this the stuff-to-know-about perspective. It’s a good starting point.  But it should not be the end point of the content model, as it often is.

When we look at stuff people might want to know about we start with topics. We identify topics of interest and then look at how these topics are connected to each other.

Suppose you and I are going to take a trip to a place we have never visited. Let’s imagine we are going to Yerevan in Armenia. We’d want to consult a website that presents content about the city. What might the content model look like?  As a thought experiment, we are going to simultaneously think about this situation both as content modeler and as a prospective tourist.  We’ll see if we can blend both these perspectives together.  (This technique is known as wearing two hats: switching roles, just like we do all the time in real life.)   

As content modelers, we will start with stuff we as tourists will need to know about.  We’ve never been to Yerevan and so we need to know some very basic information.

If we are going to travel there, we will want to:

• Find a place to stay, probably a hotel 
• Take transport within the city
• Find restaurants to eat at
• Visit tourist sites
• Check out local entertainment

These user needs provide the basis for the content model. We can see five different topics that need to be covered. There needs to be profiles of:

• Hotels
• Transport options 
• Tourist sites
• Restaurants
• Entertainment venues  

Each profile will break out specific aspects of the topic that are of most interest to readers.  Someone will need to figure out if each hotel profile will mention whether or not a pillow menu is available.  But for the moment, we will assume each profile for each topic covers  important information users are looking for, such as opening times.  

As modelers, we have identified a bunch of stuff that tourists want to know about.  But it’s not obvious how these topics are connected to one another.  It’s like we have several piles of tourism brochures: a pile on hotels, a pile on tourist sites, and so on, each stacked side by side, but separate from each other.  If you’ve ever walked into the tourist information center in a city you are visiting, and walked out with a pile of brochures, you know that this experience is not completely ideal.  There’s loads of material to sort through, and decisions to coordinate.  

Modeling to Help Users Make Decisions

If we only adopt a topic perspective, we don’t always see how topics relate to one another from the users’ perspective. It’s important for the model not only to represent stuff people need to know about. We also need the model to account for how audiences will use the content. To do this we need to dig a little deeper. 

As modelers, we need to look at the choices that users will be making when consulting the content. What decisions will users make? On what basis will users make these decisions?  We need to account for our decision criteria in the content model.

As a prospective tourist, I’ve decided that three factors influence my choices. I want to do things that are the best value, the best experience, and the most convenient.  This translates into three criteria: price, ratings, and location. 

It turns out that these factors are dimensions of most all of the topics.  As a result, information about these dimensions can connect the topics together. 

I want to go to places that are convenient to where I am staying or spending time.  All the different venues have a location.  Different venues are related to one another through their location. But we don’t have any content that talks about locations in general. This suggests to new content type: one on neighborhoods.  This content type can help to integrate content about different topics, revealing what they have in common. People both want to know what’s nearby and get a sense of what a neighborhood feels like based on what’s there.

Many venues also so have ratings and prices. This information also presents an opportunity to connect different types of content. We can create a new content type profile for the “best of” highlights in the city.   It can show the top rated restaurants according to price category. And they can show the top rated attractions. This could be a list that links to the more detailed profiles.   We now have a way to decide how to prioritize things to do.  This content type helps users compare information about different items. 

Modeling to Help Users Act

As tourists, we now know what we want to do. But are we able to do it? 

Remember, we’ll be in Armenia. We don’t know if the familiar apps on our phones will work there. Neither of us speak Armenian, so making phone calls seems intimidating. We need a way to make sure we can actually do the things we’ve decided we want to do.

For the content to really support our visit, we want the content to give us peace of mind about the risks and disappointments we worry about. We don’t want to waste our time unnecessarily — or worse, find that we can’t do things that we had planned on doing.  We want to avoid a long queue at the museum. We want to make sure that we can get a table at a well-known restaurant.  We went to go to a show, without having to visit the venue before hand to buy the ticket.

These needs suggest additional features that can be added to the content model. We want the ability to buy tickets after we decide to visit a museum or club. We want to be able to make reservations for a restaurant.  We want a booking widget.  A tourist website can create a widget that connects to outside services that enable these actions.  In some cases, the website can pull content from other sources to give readers the ability to see whether or not and option is available at a particular point in time.

Helping people act sometimes entails thinking about content beyond the content you’ve created yourself.   It can involve integrating with partners.

The Three Steps of Content Modeling

 This post is necessarily a very high-level and incomplete overview of content modeling.  There are many more possibilities that could be added, such as including a calendar of events and special offers.  But my goal here has been to provide some simple guidance about how to model content. 

The three steps to creating a user centric content model are:

1. Identify the topics  that users need to know about, and what specifically about those topics matter to users
2. Identify the criteria that users have when making decisions while consulting this content
3. Identify what actions users want to take after consulting the content, and what additional information or features can be added to help them

This process can surface connections between different chunks of information, and help to ensure that the content model supports the customer’s journey.

— Michael Andrews