The Scriptorium Content Ops Manifesto

Content Operations (content ops or ContentOps) is the engine that drives your content lifecycle.

The Scriptorium Content Ops Manifesto describes the four basic principles of content ops:

1. Semantic content is the foundation.
2. Friction is expensive.
3. Emphasize availability.
4. Plan for change.

1. Semantic content is the foundation.

Single-channel publishing processes, where content is written, edited, and then sent to a specific output type, are obsolete. Instead, content is pushed, pulled, assembled, disassembled, and rearranged for many different targets.

To accommodate these diverse requirements, the content needs to carry contextual information, such as:

• Tags. Tags are identifying information, such as title, abstract, link, or emphasis.
• Metadata. Metadata provides additional information. For example, you might have a booktitle tag with isbn metadata. The booktitle tag gives you the common name of the book, but the isbn metadata identifies the exact edition and could be used to create a link to an online bookseller or database. At a higher level, metadata lets you classify information, such as by subject matter, author, or product family. A classification system lets you sort and filter information.
• Connecting related content. To assemble small chunks of content into a larger document, you need sequencing and hierarchy information. For example, a magazine is made up of a collection of articles. To generate a print version of the magazine, you need to specify the order of the articles. For AI consumption, you need to capture relationships among content components, typically with knowledge graphs.

2. Friction is expensive.

In a content lifecycle, friction refers to productivity impediments—processes that require human intervention:

• Copying and pasting information from one location to another
• Manual formatting (or reformatting) of content to accommodate different delivery channels or languages
• Manual collaboration, review, and approval workflows
• Process workarounds to address special one-off requirements
• Creating redundant copies of information instead of using existing content

A sufficient degree of friction in the content lifecycle makes it impossible to scale up content.

AI can either reduce or increase friction. For example: 

• For a content consumer, an AI chatbot might reduce friction by supplying only specific information.
• For authors, AI assistants can eliminate manual tasks, such as tagging content and writing short descriptions.
• AI-powered content creation tools can create friction by generating incorrect information.
• AI-driven workflows could also result in content silos as information is copied from one location to another.

Content scalability refers to your ability to expand your content lifecycle to process more content, more content types, more channels, more languages, and more variants. The greater your scalability needs, the more critical it becomes to eliminate friction.

Workflow and governance are also common areas of friction. In some industries, complex approval workflows and multiple layers of quality control are necessary. But many organizations have time-consuming, multifaceted governance processes that don’t match the risk profile of the content. If your content is regulated, can have health and safety risks, or is otherwise high risk, heavy governance may be needed.

Friction slows down your content lifecycle and introduces waste into the process. Eliminate it so that you can operate more efficiently and more accurately. Software tools, especially content management systems, translation management systems, and automated rendering engines, are essential components. 

3. Emphasize availability.

Customers expect content on demand, in their preferred format and language. For content ops, that means focusing on content access:

• Ensure that information is current and accurate. Push updates early and often. 
• Ensure that information is accessible. Do not assume that all of your readers have perfect vision, hearing, and fine motor control. Instead, provide at least two ways to access information—at a minimum, a podcast should have a transcript, a graphic should have a descriptive caption, and videos should have captions. Provide keyboard navigation options in addition to clickable regions. Choose colors carefully, so that a color-blind person can use your material. 
• Provide a variety of delivery options to accommodate your customer’s needs. Consider how you might improve the customer experience with personalized information delivery.
• Give customers entry points into the information: for example, search, faceted search, filtering, category-based navigation, linking related information, and curated pages for specific topics.
• Localize and translate your content. Remember that in many countries, you will need more than one language to reach your target audience. The US, for example, has roughly 40 million native Spanish speakers.
• In addition to traditional content publishing (the organization pushes out content for users), look at a pull model, in which a user can request relevant information from a repository.
• Recognize that end users will remix your information using conversational (AI) interfaces, which means that clarity and consistent structure in the original text are of paramount importance.
• Consider the user’s context in time and space to improve the relevance of information.

4. Plan for change.

Content ops will evolve with new forms of content, new tools, and new business requirements. With this in mind, prioritize flexibility:

• When you choose a platform, have an exit strategy in mind from the beginning.
• Audit your systems regularly to ensure that they are still meeting your needs.
• Conduct forward-looking needs analysis to identify emerging trends and requirements.
• Build out metrics and measurements that help you understand your content ops systems’ overall performance.

Thanks to Rahel Anne Bailie, Patrick Bosek, Carlos Evia, Jeffrey MacIntyre, and Kevin Nichols for helping to clarify my content ops thinking.