Structured vs Unstructured Authoring: What a CCMS Actually Changes

Most regulated and technical documents in Australia are still written in Word. A rule determination, a Product Disclosure Statement, a maintenance manual: someone opens last year’s file, saves a copy and starts editing. That works until the same clause lives in forty documents, or three people need to change one document at the same time.

Structured authoring and component content management systems (CCMS) are the usual answer. Here is what each one means, what actually changes, and when Word is still the right tool.

Structured authoring versus formatting in Word

In Word, structure is a matter of appearance. A heading is text that looks like a heading. Styles help, but a style describes how text looks, and nothing stops an author from bolding a line instead of applying one. The file can’t tell a warning from an ordinary paragraph, or a defined term from any other phrase.

Structured authoring means writing against rules about what content is. A schema sets which elements exist (a procedure, a step, a warning, a clause, a definition), what each one may contain and in what order. The content is usually stored as XML, and the look is applied later, at publishing time.

Two things follow. Software can check a document against the rules before anyone signs it off. And because the content says what it is rather than how it looks, the same source can be published to PDF, Word, HTML or a data feed without anyone reformatting it.

Unstructured authoring isn’t wrong. It is formatting, and formatting can’t be validated, reused or queried.

What a component content management system does

A CCMS stores content as components instead of whole documents. RWS defines a CCMS as a platform that “manages content at a granular level”, where each component “can be a paragraph, image, table or any discrete piece of information that can be created, tracked and reused independently.”

In practice that gives you five things:

• Reuse. A clause, warning or product fact is written once and referenced wherever it is needed. Change it once and every document that uses it changes with it.
• Versioning. Each component has its own history, so you can see who changed what, when, and which version went into which release.
• Workflow. Review and approval are built in and assigned by role, instead of running through email.
• Assembly. Documents are built from components, sometimes automatically from metadata and rules.
• Multi-format output. One source publishes to print, PDF, HTML and other channels.

A CCMS is not the same as a document management system. A document management system versions files. A CCMS versions what is inside them.

The standards underneath

Many CCMS products are built on open XML standards. Four come up most often.

DITA (Darwin Information Typing Architecture) is an OASIS standard. Its technical committee describes it as “a document creation and management specification that builds content reuse into the authoring process.” It is topic-based and can be specialised for particular domains.

S1000D is the “International specification for technical publications using a common source database”. It is maintained by ASD, AIA and A4A, is free to download, and is used for defence systems, civil aviation, construction and ship industry products.

DocBook is maintained by an OASIS technical committee that describes its work as “defining a schema for describing XML documents for general and technical publishing”.

JATS (Journal Article Tag Suite, ANSI/NISO Z39.96) “provides a common XML format in which publishers and archives can exchange journal content”, in NISO’s words. Version 1.4 was published in October 2024.

Legislative and regulatory content often uses its own schema, and that is fine. The schema should fit your documents, not the other way round.

Component content management system examples

Here are five products, described in their makers’ own terms.

• Quark Publishing Platform (QPP). Quark calls QPP NextGen “content automation software for modular, metadata-driven and compliance-controlled omnichannel publishing.” It says you can “use any XML schema”, including its own Quark Smart Content schema, DITA, DocBook and S1000D, and publish “to print, PDF, HTML5, Web, XML, tablet and mobile apps”. We are a Quark and QPP partner for Australia and the USA.
• Adobe Experience Manager Guides. Adobe describes it as “an enterprise-grade component content management solution (CCMS)” that adds native DITA support to Experience Manager.
• RWS Tridion Docs. RWS positions it as an enterprise DITA CCMS, and says it connects natively with its Trados and Language Weaver translation products.
• Paligo. A cloud CCMS that says it is “based on DocBook”, with versioning, branching, release workflows, publishing and translation management built in.
• Heretto. A CCMS built around DITA for technical documentation teams, with publishing from the CCMS to web, PDF and customer portals.

They are all capable products. What separates them for a given organisation is fit: which standard you need, who will be authoring, and which systems the content has to connect to.

What about open source?

The best-known open-source tool in this space is the DITA Open Toolkit, released under the Apache License 2.0. It describes itself as “the open-source publishing engine” for DITA content, with outputs such as HTML5 and PDF.

It is not a CCMS. It doesn’t store components, track versions or run approvals. Commercial products can use it underneath: Adobe, for example, lists custom output “through DITA-OT” for AEM Guides.

An open-source route usually means pairing a publishing engine with an XML editor and version control, then building the rest. That can suit a team of technical writers who are comfortable in XML. In our experience it rarely suits lawyers and policy staff.

When it pays off, and when Word is fine

Structured authoring costs effort up front: a content model, templates, migration and training. It pays off when at least one of these is true:

• The same content appears in many documents. Product facts shared by a PDS and a target market determination, a definition used across a set of rules, a warning repeated in every manual.
• You maintain many versions. Product variants, jurisdictions, or point-in-time versions of an instrument that must stay retrievable.
• Sign-off is regulated. You need to show who approved what, and that what was published is what was approved.
• You translate. Translating a component once and reusing it costs less than retranslating whole documents.
• You publish in several formats. PDF, Word and HTML from one source, instead of three versions drifting apart.

Word is fine when documents are one-offs or short-lived, written by one or two people, with little shared content and no audit requirement. If none of the five points applies to you, a CCMS adds process you won’t get value from. Better templates may be enough.

Authors don’t have to see the XML

The old objection to structured authoring is that authors have to learn an XML editor. That is no longer a given.

Quark says that in QPP, subject-matter experts and technical writers can work in Microsoft Word, PowerPoint and Excel, or directly in a web browser. They “don’t need to learn XML schemas or content modeling”, because XML metadata and tagging are “automatically generated behind the scenes”. Authors use Quark Author, its web-based authoring tool, or Quark XML Author, its Word plug-in.

That is how it works at the Australian Energy Market Commission, where we implemented QPP. AEMC’s legal and policy teams write regulatory documents in the browser and publish PDF, DOCX and HTML automatically from one structured source, with no XML specialists needed.

Moving from Word and PDF

Most migrations start from a shared drive of Word files and PDFs. Quark says QPP uses AI “to transform unstructured PDF and Word documents into structured, reusable components using smart auto-tagging.” Tools like this speed up the mechanical part.

The harder part is deciding what the components should be. Convert before you design the content model and you get the same documents in smaller pieces, with the same duplication. So we work in this order:

1. Audit what you have. Not just how many documents, but how consistently they were produced.
2. Design the content model. Which components exist, which are shared, and what metadata they carry.
3. Pilot on documents you nominate. Real documents, not a demo set.
4. Migrate the rest. Carry identifiers and version history across where they exist.

Our post on AI in structured content publishing covers where AI helps with conversion and where human review still matters.

Parallel drafting for legislative bodies

Legislative drafting has a problem most document workflows handle badly: several people amending the same instrument at once. In Word, each drafter takes a copy, and someone spends days reconciling them.

We built a custom extension to QPP NextGen for legislative agencies and legislative bodies that brings Git-style branching to document management. Drafters branch a bill, rule or amendment from a point in time, work independently, and merge back with automatic conflict detection at document and component level. A large restructure can run alongside routine amendments without either one overwriting the other.

Branching as an idea isn’t unique to us. Paligo, for one, lists it among its built-in features. Ours is built on QPP, specifically for legislative agencies and legislative bodies. There is more on it under regulatory and legislative publishing.

Where to start

Pick the document family that hurts most: the one with the most shared content, versions or sign-offs. Model it, pilot it, and compare the time it takes to produce and approve against Word.

If those documents also need controlled delivery once published, our Content Portal delivers them securely to the right readers, with SSO, role-based access and watermarking. For the wider picture, see our technical publications practice and our Quark Publishing Platform page.

This guide is written for Australian organisations. In the USA, start with Shakewell LLC at shakewellusa.com.

---

Sources: Quark Publishing Platform (Quark), Experience Manager Guides overview (Adobe), Tridion Docs, What is Tridion Docs and Component Content Management System (CCMS) (RWS), Paligo DocBook CCMS (Paligo), Heretto, DITA Open Toolkit, OASIS DITA TC and OASIS DocBook TC (OASIS), S1000D (ASD, AIA, A4A), and JATS: Journal Article Tag Suite (NISO).