Designing connected content
How we’ve used content modelling and structured content for digital products
When you’ve spent time crafting content that meets user needs and has been signed off by stakeholders, you’ll probably want to use it in as many places as you can. You may have a definitive event listing or product description and there’s no need to reinvent the wheel for every other channel or page.
Except sometimes as content practitioners, we can find ourselves doing just that, especially when we’re working across large digital ecosystems.
The main benefits of structured content and content modelling are:
Reusability: to be able to create once, publish everywhere (the ‘COPE’ model) where the master version of the content is created with structure applied and then it can feed other locations.
Consistency and accuracy: the structure acts as a template, so your content will have less variation and be more complete.
Findability and accessibility: structure can make it easier for users and machines alike to find your content and understand what it’s about. Helping search engines and AI-powered machines to gain a deep, semantic understanding of your website content offers a host of SEO-related benefits. It also improves accessibility by making the content readable and reliable for assistive technologies.
Efficiency and cost savings: having structure makes it easier to create, update, and manage content, which saves time and money. For the people responsible for content on a project, it frees up time to focus on crafting the content that will be tailored to meet specific user needs.
Lauren Pope has an excellent how-to guide that expands on these benefits.
In recent projects for Paul Hamlyn Foundation and BookTrust, we have been structuring content to provide a better experience for users and editors. In this blog we’ll explore how we’ve used these methods on recent client projects and what we’ve learned from this approach.
The goal: structured content
Moving from well-organised to structured content involves putting the structure before the content itself.
Content that is organised logically, in a way that makes sense to users, makes it easier for them to find information quickly and take action. Well-organised content is largely a result of user-centred information architecture, user journeys and the hierarchy of information.
Structured content is planned, developed and connected outside a specific product, channel or interface so that it can be used in any. It’s not tied to its original form or purpose. Like a LEGO brick, which can be used to make a house, a ship or a tree, the content pieces become free to be part of lots of different wholes. The structure comes from breaking content down into its constituent parts and arranging them following a template. You can put the same content together in different ways over and over again. Just like LEGO.
The Government Digital Service has been exploring a system that lets publishers manage reusable snippets of information to store facts in one place and share them across multiple pages. Charities could use something like this to update beneficiary advice, impact data or call-to-actions to the latest campaigns.
Standardising funds content for PHF
As part of being an equitable funder, Paul Hamlyn Foundation wants the information about their funds to be easy to understand for everyone, including first time applicants who may be less familiar with the funding world. People who are looking for support from foundations are usually busy, wearing many hats and looking to find out quickly if there’s a fit between a funder and their work.
We structured their funds content based on what we heard from prospective applicants during user research conversations. The most important information for applicants included:
the amount of money available
how long the funding lasts and
when people can apply
These insights led to how we structured the fields in the content management system (CMS). We found that each fund had its own nuances. For some funds the deadline is continually ‘rolling’ but we knew that this was better than not mentioning a deadline at all. The one-sentence summaries also took some careful crafting to effectively summarise each of the funds in clear language.
By finding a base structure that they could all start from, it allows the fund content to be easily pulled into different places such as their fund finder tool, as well as the application guidance and listing pages, ensuring consistency across all these locations.
The method: content modelling
Content modelling is the process of defining:
the types of content you need
the attributes of each one
the relationships between them
Carrie Hane and Mike Atherton, ‘Designing Connected Content’
At William Joseph, we often talk about ‘chunks’ of content. The types of products and projects that content modelling works well for are when:
there are repeatable chunks of content, such as event listings and product descriptions
the information is or could be used elsewhere online, such as apps, email, social media or third-party websites
you have a decentralised publishing model and multiple stakeholders are adding or editing content
For Paul Hamlyn Foundation we applied a model to their funds and funded projects. Whereas, for BookTrust we applied a model more broadly to their book reviews, people, programmes, resources, booklists and articles.
Domain modelling rooted in user insights
‘Domain’ is quite a grand word but in the context of content modelling, the domain is the subject matter that you’re creating the content model for.
This is typically done during the discovery stage of a project and involves immersing ourselves in the subject domain(s) of our client. Some methods that are helpful to explore domains are:
Desk research: including schema.org and large language models (LLMs) to explore how the concepts exist elsewhere online.
Stakeholder interviews: these can be particularly helpful for understanding why things are organised the way they currently are.
Usability testing: We tend to start each project with a round of generative user interviews to look at the current product and some comparators to bring in the voice of the user from the outset.
Shareable resources for BookTrust
During the BookTrust discovery phase, we interviewed professionals who are supporting families and children, such as health visitors and early years educators. One of the things we learned through these sessions is that book recommendations were most helpful when supported by practical resources they could use in their roles.
BookTrust had plenty of resource content on their old site, but a lot of it was hidden away as news articles or blog posts. As a result, one aim for the product and the content model was to better connect books and programme information to resources that partners could use and share with families.
One use case that we heard from early years educators was that they might go to the BookTrust website if they were looking for books and related activities for an awareness week or month.
This example booklist for Deaf Awareness Week shows how content can be brought together. Books that have reviews on the site are selected by the editor as complete chunks of content rather than entering them individually for each booklist page.
Further down the page, there are related resources for tips and programme information. Surfacing relevant related content isn’t unusual but we’re able to do it at scale. BookTrust have over 250 booklists, because of the way that the content has been modelled in the CMS.
Building your content model
Once you have your domain map, you need to pick the concepts from it that will become content types, which will have attributes.
Content type: a template you use to create multiple pieces of content that have the same purpose and structure.
Content attributes: the different elements that come together to make up the content type ‘template’.
Content modelling involves identifying the concepts where you’d actually want to have a reusable template. One thing to look out for is content that is unique or specific – you don’t want to include them. For example, it wouldn’t be worth modelling a content type for your ‘About us’ page, as this is a single entity, not a content type that you’ll repeat.
Implementing a content model into a CMS
The aim of a content model is to integrate it with a CMS to fully realise the structured approach that has been planned. We’re fortunate that Craft – the CMS we use most often and that the BookTrust site is built with – is very flexible and suitable for content models.
This step involves defining the field type and parameters of each attribute. This means content migration or population won’t involve adding all the copy into a text editor in a CMS, but rather having individual fields for the different attributes.
An example of how we adapted the BookTrust content model was for people entries, with ‘name’ data split into three separate fields:
first name
last name
display name
In the majority of cases for BookTrust’s content, people have a first and last name and that is what should be displayed on the front end, but some edge cases include a title or an initial, like Michael J. Rosen, that needed to be displayed. So we included a ‘display name’ field in cases where editors need to override first and last name. This way we could be more inclusive and comprehensive about how names were entered into the CMS.
Name fields are something that comes up in other areas of designing products for charities, such as form design and donation journeys. We recognise that the most equitable solution would just be a single name field that covers all naming conventions, but there’s often a tech reason for needing them to be separate. In this case, a single name field would mean we would lose the future possibility to sort books content by last name, which is a typical convention in the books world.
Final thoughts
Designing connected content using structured content and content modelling isn’t a new approach, but it has increased importance in the so-called age of ‘AI-first’ development. As Patrick Stafford has written recently, when considering what ‘AI-first’ means for content designers:
‘Without structured content systems in place, products risk becoming fragmented, inconsistent, and confusing for users. Clear content architecture, scalable patterns, and strategic messaging frameworks are no longer “nice to haves” – they will be survival mechanisms.’ - UX Content Collective
There is also a human aspect to having more structured and connected content; it will necessitate alignment on content strategy across your organisation (rather than siloed working or conflicting priorities) and will make content operations smoother.
This blog was originally presented as a talk at DiNG: Design and Innovation Nottingham in May 2025
Resources
- In this talk ‘Content is Smart Now’ for Content Folks, Sandie Markle talks about content engineering and why understanding the basics of structure, metadata, and systems thinking is becoming a core skill for content professionals. Sandie also mentions the LEGO analogy.
- Lauren Pope’s guide to Content Modelling and Structured Content: The Basics
- Carrie Hane’s Content Modeling: What It Is and How to Get Started
- What does “AI-first” mean for content designers? - Patrick Stafford, UX Content Collective