How to achieve reuse?

Some time ago, the Rockley Blog talked about identical vs derivative reuse.

“When content is reused identically, it is reused without change. Derivative content is reused with change, content stays related to the original component.”

Often content cannot be used for multiple purposes directly as such. In her article, Rockley focused on two issues, the channel and the audience, and how addressing these issues does not mean content could not be reused.

To reuse content effectively, as much as possible of your content should be global and generic. Global content can be published in different contexts as-is, without changes (either identically or through filtered reuse like conditional text). Differentiate only where you must. These must-haves depend a lot on the item you are documenting, the environment for which the content is created, and also on the legacy practices in your organisation.

There are certain methods you can use to achieve reuse, and I’ll discuss two of those briefly here: information design (info architecture of your content) and conditional text.

Information design
In the information design of modular content, you define both the granularity, so size, of your modules and their content. Design as many modules as possible as global content.

  • To address variation in product features, for example, try to confine the varying information into separate modules. If there is a feature or component that is not available in all products, instruct it in separate files.
  • To address variation in the needs of separate audiences, design the modularity so that different modules speak to different audiences. For example, describe the advanced tasks in separate modules, which can be added to the content only when the audience benefits from this information.
  • Like Rockley already describes, to address the needs of publishing formats, you may select to use different modules of the content for different channels. In some industries, the amount of printed documentation is actually decreasing due to various drivers. In cases like these, web channels may actually include more content (more modules) than the print one.

Conditional text
Especially if you need to address several sources of variation in your content, some kind of filtering such as conditional text is just about a must. By tagging complete modules and especially smaller elements inside them, you can reuse the same files directly and still cater for the various needs of software, hardware, products, audiences, countries, market areas, publishing channels…

When the differences are tagged with conditional text, you can maintain the content in a single source and reuse the file identically. The content is just filtered differently; you configure separately for your publishing which tags apply.

Make reuse happen

Rockley describes the caveat of content reuse:

“One thing is important to understand, not all content can be reused identically. You never want to compromise the quality of the content in order to reuse it. Derivative or unique content is OK, just ensure that it doesn’t dominate your content set, rather it is an appropriate part of your content set.”

Especially if your organisation does not have a long history in content management, it is beneficial to always start with the premise of reuse. When considering how to design and author the content, let your first question be “how can this be written as global, reusable content”. This way you are always forced to analyse the situation, and derivative or unique content is created only for solid, valid reasons, not out of tradition.

One Response

  1. [...] When you are writing for reuse, you must suddenly know a lot more. You must know not only one product very well; you must know all the products that may reuse your content, to some extent. The minimum that you must learn is what features (hardware, software, whatever) typically vary between the products. This way, you can create global, generic content that suits all of them, and know where content must be tagged to allow for variation (like discussed in a previous post). [...]

Leave a Reply