Structured Content: An Introduction
Structured content allows us to create meaningful relationships between pieces of information. It enables users to navigate freely, using formal and informal information-seeking behavior, improving content findability and satisfaction.
This article is an introduction to the basic process of structuring content, and is based on the workshop I ran at UX Ireland 2016. I hope it acts as a starting point that prompts you to find out more, because I strongly believe structured content has the potential to improve user experience so significantly, it deserves more time and attention.
We will cover why it is worth structuring content, and why it plays a role in user experience. We will also look at some specific examples of how to structure content, at quite a high level.
A bit of background:
To structure content, you must define what content is and what relationships it has, and how it can be described. It must be structured in a way that maps to the audience’s mental model, the content provider’s constraints and your publishing system’s flexibility.
Investing time structuring content makes good business sense because it extends the lifespan of content, as it makes it easier to reuse across multiple devices in the present, as well as in the future. It sits well alongside service-oriented architectures, as APIs can be developed to work well with clearly organised and structured data.
User Experience & Structured Content
User experience is an amorphous term. It acts as an umbrella for various specific practices, which all have the intention of improving the experience people have when they use digital products.
Fields that have fallen under this umbrella include:
Usability: the practice of designing/engineering efficient, effective and satisfying functionality and experiences
Information Architecture: the practice of organising information in relevant, meaningful and findable systems and structures
Content Strategy: the practice of strategically planning, structuring, and maintaining content to meet user’s information needs
Service Design: the practice of designing cross-platform services that deliver pleasing experiences both on and offline
Interaction Design: the practice of designing interactive experiences that help people use systems and get appropriate sensory feedback to guide them
Visual Design: the practice of designing visually pleasing interfaces that positively convey brand values and encourage user engagement
The problem with this umbrella is that some fields simply don’t get the time and attention they need to deliver the right experience ‘components’. Great experiences are created from all these fields - not just visual design, which often gets the lion’s share of attention because it’s so immediately visible.
Structured, scalable systems are reliant on IA and content strategy. We can’t rely on developers and content management systems to magically make everything work while we put the UX lacquer on the top.
Also, let’s not forget that words and written text are the most important part of the Internet. From a content perspective, words are both information and the tool to retrieve information. They are enhanced and supplemented by other media types, but without words our sites and apps couldn’t work.
“It’s pointless to design a user experience without a deep understanding of the content that will fuel that experience”
- KRISTINA HALVORSON
A note about content auditing and workflows:
I’m not really digging into the complete workflow in this article, but in brief, a typical flow would involve a good amount of auditing and analysis of any existing content early on. This provides a clear view of where there are gaps in content, missing competencies, and structural or organisational problems.
Most problems involving content are those created by humans during the design and creation of a site/system, those made during content creation itself, and those made by a lack of maintenance. Auditing content shines a bright light on these issues and gives valuable information that will help you avoid recreating mistakes.
Structured Content Basics
Structuring content involves removing our reliance on the idea of ‘pages’. We need to look at the broader sense of what a piece of content is beyond the idea that it is a page. Content is made up of clearly visible qualities, like a title, description or body text, but it also has a set of less visible qualities that are just as important. These include details like IDs, relationships, content types etc.
A piece of content can be used in many different ways, if it has been stored and labeled in the right way. It may sit on a web page in one context, an app in another, in printed documents, or on product packaging materials to name a few.
There are three main levels to think about with structured content:
High level: the overall site-wide structure of all the content such as the main navigation, supplementary navigation etc
Groups: how similar and related content is grouped together into structures and labelled using taxonomies and content types
Specific: how individual pieces of content are collected and structured and their data labelled with metadata and microdata
Each of these areas help to improve the findability of content in different ways.
Content Findability
The speed, accuracy and ease a user can locate a specific piece of information is very important. Making a piece of content findable involves optimising the visible content and the hidden data associated with it. This applies for both on site and off site user journeys.
Findability breaks down into four specific areas:
Onsite:
Site navigation:
- Menus
- Filters
- On site search
Other content:
- Relationships (taxonomies, categorisation etc)
- Link strategy
- Content types
Offsite:
Search engines:
- Keywords
- Content quality
- Data visibility
Social media:
- Titles
- Assets (e.g: images, videos)
- Sharability (is it worth sharing?)
Creating groups and relationships between content
Let’s use cheese as an example. If we think of a list of cheeses, this could include Cheddar, Stilton, Mozzarella, Feta, Red Leicester, Parmesan, Brie, Emmental, Gorgonzola and Halloumi. They are all cheeses, but they’re not all created equally, are they? They have specific qualities and information associated with them that as humans we instinctively understand, but computers or content management systems wouldn’t.
For instance, Cheddar, Stilton and Red Leicester are all from the UK, whereas Mozzarella, Gorgonzola and Parmesan are Italian. Stilton and Gorgonzola are blue cheeses, and Feta and Halloumi are both made using sheep or goat’s milk. Unless we specifically map these qualities of ‘country’, ‘type’ and ‘milk’ to these cheese types, we won’t be able to associate them together in a structured way.
Labels and hierarchies need to fit with the domain model of the content you’re dealing with as well as the audience’s mental model. Be wary of subjective qualities, they need extra care and testing to ensure they match user’s mental models. For instance, in our cheese example flavour strengths could mean very different things across audience groups.
A common problem with tagging content is that tags are often not used consistently, or they get made up on the fly. This leads to sprawling, unruly tags that don’t help with findability. For instance, imagine one content creator tagging a cheese as ‘strong’, another tagging one as ‘mature’ and a further one saying ‘full flavoured’. Having a controlled vocabulary or term list to describe flavour strengths would avoid this, and allow cheeses with similar strengths to be related to each other.
Taxonomies
A taxonomy is an agreed, controlled list of terms used to describe content.
Taxonomies play a crucial role in structuring content. They help content management systems relate content and pages together in consistent patterns.
Depending on the way you can use taxonomies with your content creation and maintenance team, and your content management system, you may use the following types of taxonomy too:
Term list: A standardised, controlled list of terms or tags that are used consistently
Hierarchy: A structured taxonomy of terms that have a structural relationship with each other (e.g: parent — child)
Thesaurus: Maps relationships between items to help systems understand how they relate
A note of caution: if you are creating a hierarchical taxonomy, be wary of trying to create the ‘parent’ items first and then fitting the children underneath. It is very tempting but difficult to do this unless you are very, very used to the domain you are working in. It is much more advisable to group the items first and then see what natural hierarchy comes from them.
If you are going to use the parent items as navigation labels I strongly suggest testing them. Labelled hierarchies can be checked using tree testing or card sorting. These are both relatively quick and easy ways to determine if your labels are working and match to your audience’s mental model. They will also help you weed out any corporate jargon that might have crept in!
Grouping content by type
If we think back to our cheeses, we could have various content types associated with them. For instance:
Product: “Cheddar Cheese”
Review: “Mature Cheddar Cheese Review”
Recipe: “Cheese Soufflé Recipe”
News: “Cheese health benefits announced!”
Press releases: “New soft cheese launched”
Travel guides: “What to do in Wensleydale”
Diets: “The Four Cheese Diet”
Case studies: “How I lost weight eating cheese”
If we look at the example content types above, we can start to map relationships between them. If we consider ourselves to be a cheese retailer, then ‘Product’ would sit at the heart of our activities. We need to associate products with the other content types. Lets say we have the product ‘Cheddar Cheese’, and a recipe that uses Cheddar (‘Cheese on Toast’). ‘Cheddar Cheese’ is linked to ‘Cheese on Toast’, so we need to map that relationship.
Content Modelling
The process of content modelling can map relationships between domain level concepts and more detailed content. If we take the example of a Product’s relationship with a Recipe further, we can also look at how both a Product and a Recipe could have Reviews about them, and they could both feature in a Travel Guide.
These relationships can be mapped out onto a simple diagram. I’d recommend working on post it notes, or a whiteboard to do this before going anywhere near any diagram software. You need freedom to explore how items interconnect and software has a knack of shutting that down very quickly. This is not an activity for one person to do, either: you must involve your domain experts as they have valuable insight.
When it comes to building the system you’ll use to publish your content, having a clear content model diagram will help you explain to your development team how these content types and relationships interconnect.
You can take the content model further and make it into a spreadsheet. Give each content type a worksheet, and list out all the things that are associated with it. You can really dig into the detail with these worksheets as you can begin to plot in the dependencies on data from third parties as well as within the system you are using. Again, I don’t believe this is a solo activity. Having other people work with you on this will help challenge your assumptions.
Site maps
So, we’ve looked at how groups of content can be linked together using taxonomies and how content types can have relationships with each other with content models, but what about the high level navigation we need to give users to access content?
Site maps are an important communication tool. They help you communicate the overall structure of how your content sits together, and how it can be found in a hierarchical sense.
Site maps can be diagrams showing pages and how they are structured into primary and secondary navigation, or they can be very detailed spreadsheets that itemise the details about every page.
I like to use diagrams to start the process of organising the main pages as they are easy to change. They come with a warning: if you are using ‘page stacks’ (multiple pages stacked on top of each other to indicate there are a bunch of pages in a section) then you’re not really making a site map as you’re not specifying what is going on there… and who’s going to do that if you won’t?
After creating an initial diagram, I move into a spreadsheet to list pages and their titles as much more detail can be recorded against them.
Faceted navigation
Faceting navigation can help users find content through a refinement process. This can be very useful to people when they are not certain what they are looking for and can improve user’s perceptions of the effectiveness and efficiency of a site or app.
There are two main types of faceted navigation to consider:
Fixed: such as geographic (French Cheeses) or alphabetical facets (Cheeses beginning with ‘A’)
and
Ambiguous: such as task based (Cheeses under £5) or audience segments (Beginners’ Recipes)
Other facets you can consider using are:
By Audience: e.g. Skill levels such as Beginner, Intermediate, Advanced
By Subject: e.g. How to guide
By Product: e.g. by Brand, Price, Features
By Format: e.g. Video
Choose the facets you provide based on evidence of your audience’s needs.
Specific content structures
So far we’ve looked at how content can be grouped and labeled, and how it can sit together in navigation patterns, but now we need to look at the specific content itself, and how that can be structured.
Although we may ultimately produce ‘pages’ from our content, we don’t necessarily make pages themselves. There could be multiple pieces of content that get pulled together to create a page.
If we only think of pages, we can end up just making blobs of content with a title and body copy and that’s not structured. We need to consider the data about the content so that it can be findable by off site searches and our content management system.
Helping machines understand content
Search engines, content management systems, APIs and feeds don’t automatically understand what content is about. We need to markup our content to make it make sense to these systems as they power the findability of most content.
Metadata
When you last moved house, you probably packed your stuff into cardboard boxes to transport it. All the boxes looked the same, so you wrote labels on the boxes to remind you and the people moving your stuff what was in each box.
These labels are metadata, as its data about content. ‘Plates’ or ‘Kitchen’ are perfect box labels to describe both what is in the box, and where it needs to go. Many thanks to my friend @morpchic for help with this example!
The metadata you assign to a piece of content allows it to be used just like the boxes. It means a content management system understands that this piece of content is a ‘Recipe’ and it involves ‘Cheddar Cheese’ and so it is related to all the other ‘Recipes’ and to the product ‘Cheddar Cheese’.
Microdata
Microdata goes into even more specific detail than metadata, as it stores specific information in context. If we use the box example: if we opened the box labelled ‘Kitchen’ and found a plate inside, it would have a label on it saying ‘Cupboard next to dishwasher’ and ‘Porcelain’ and ‘Ikea’.
There are many different standardised schemas that can be used to store microdata in context with content. Popular ones include RDF, Schema.org and Dublin Core. The type of microdata you use will depend on the domain you’re working in, the content management system you have and how much customisation you can do to add this additional data to your content.
Marking up content with microdata such as Schema.org can help Google to understand content more meaningfully. You can test it using their structured data testing tool.
Go beyond pages
I hope this introduction to structured content has sparked a desire to explore how content can be made more reusable, findable and useful to users in the present, while being flexible to fit to future uses too.
The processes and techniques I’ve described could all have a lot written about them in their own right. If you are keen to learn more there are some excellent books that will give you a good overview: Content Everywhere by Sara Wachter-Boettcher, and Content Strategy for the Web by Kristina Halvorson and Melissa Rach.
The infamous Polar Bear Book is also very useful, always.
