Content modelling is on everyone’s lips these days, yet it’s a term that seems not to have existed just a few years ago. Is it some entirely new concept? As usual, a quick look on the Web reveals several definitions, some of which concur, and most of which differ in emphasis.
So, for Cleve Gibbon, a content model is a representation of the types of content and their inter-relationships. For example, a car dealership may have content types for vehicle, dealer and manufacturer. This, of course, is where you start when modelling a relational database.
For Rachel Lovinger, a content model documents all the different types of content you will have for a given project, which I would call a content audit, although she actually starts with a very clear and simple diagram of the content relationship. She then goes on to reveal a full content audit, listing the content types, quantities, and so on.
Wikipedia has no separate entry for content model. It does have an entry called “indecs Content Model”, although I trust I am not alone in finding that entry very unenlightening. Among other things, it never gets round to defining what a content model might be. The implication, however, is that the content model is a statement of resource relations and categories, which I can agree with.
None of this is at all contentious, so you could say, what’s the problem? Perhaps a more useful answer then is to learn what content modelling is not. For Cleve Gibbon, it’s not a website sitemap, or a diagram of the web pages. For Mike Atherton, it’s not (necessarily) a hierarchy: “knowledge is wild and untamed. Hierarchy tries to impose rigid structure, sometimes arbitrarily.” Hence for Atherton the content model is simply a visual representation of types of content with no attempt to place hierarchy on it. You begin to see how the content model is now drifting towards facets of content, that enable the content to be retrieved on a website. For Atherton, the content model is explicitly not the content inventory. Instead, Atherton emphasises the process of content modelling, which involves talking to experts and users to articulate their subject (hang on, wasn’t that the user journey interview?). The resulting diagram looks to him like this:
The Atherton content model example
Shown like this, the content model can only be a very first pass. Yes, Disney Parks has restaurants, and has attractions and characters, but that doesn’t get us very far in terms of understanding the process, and hence of delivering content (or solutions) to the user. The real benefit might be to understand that the restaurant becomes important either before or after visiting an attraction; or that visitors might prefer snacks to meals. Considerably more work is required to create a content model that corresponds to the user journey, and a user journey is a key component that seems to be missing from many content model descriptions.
Content modelling, in other words, is close to a consensus – it’s just that the consensus is a mix of activities that have been taking place for years but under slighly different labels. And content modelling is in itself incomplete. Unless we are careful, the content model leaves us in just as much of a muddle as when we started. It's only when the raw data that has been gathered is put into some kind of coherent relationship and/or sequence that the information flow starts to become truly structured. To be effective, it needs user journeys and some thinking (usually by the architect) to make some kind of sense of it.