Explainer
What Is Structured Authoring?
The Short Answer
Content Constrained by a Schema
In structured authoring, the author does not draw a document. The author fills in typed components, such as a procedure, a warning, or a parts list, and a schema says what each type may contain and in what order. The W3C's XML specification describes the mechanism: a document type declaration can define constraints on the logical structure, a DTD is a grammar for a class of documents, and a document is valid when it complies with those constraints.
The second half of the definition is separation. The source says what the content is; a stylesheet or publishing pipeline decides how it looks in a PDF, a web page, or an interactive manual. That separation is what makes single-source publishing possible, and the schema is what makes validation possible.
The Contrast
Structured vs Unstructured Authoring
Unstructured authoring covers word processors and desktop publishing: the author sees the page and shapes it. Both are fine for one-off documents. They strain when the same content must stay consistent across many deliverables.
- What the author controls
- Unstructured: everything, including fonts, spacing, and section order. Structured: the words and the choice among element types the schema allows at that point.
- What a mistake looks like
- Unstructured: a procedure with no warning, or a heading styled by hand, passes unnoticed. Structured: the file fails validation, and the tool can say why.
- What a unit is
- Unstructured: a document, or a file of pages. Structured: a topic, data module, or similar component small enough to reuse and review on its own.
- How output is made
- Unstructured: the file is the deliverable, so a second format means a second copy. Structured: stylesheets and publishing pipelines produce each output from the same source.
- How change spreads
- Unstructured: someone searches for every copy. Structured: a shared component is edited once and every publication that references it picks up the change.
Structure does not require authors to type angle brackets. Editors can present a form-like or word-processor style view while the structure sits underneath; see Word with guardrails for the middle path and why Word breaks at scale for the limits of the other.
The Standards
Where the Schema Comes From
You can invent your own model, but most programs adopt a standard, because the schema is the hard part and because customers and regulators already know it. Four are common in technical publications.
- DITA
- Topic-oriented document types that you can extend and constrain for your own content. DITA standard.
- S1000D
- Content stored as data modules in a common source database, governed by an issue of the specification plus your program's business rules. S1000D standard.
- MIL-STD-40051
- The single standard for preparing U.S. Army and related technical manuals, with published DTDs that define the structure. MIL-STD-40051 standard.
- DocBook
- An XML vocabulary built for hardware and software documentation. Its standards committee closed in 2024, so it is stable rather than evolving. DocBook standard.
They differ in who they serve and how tightly they bind. DITA is flexible by design, since it lets you extend and constrain its types. S1000D and MIL-STD-40051 are program standards: a contract cites them, and the issue or revision you cite matters. Our DocBook and DITA comparison and DITA and Markdown comparison help when the choice is open.
The Trade
Benefits and Costs
The benefits follow from the two properties. Because content is constrained, it can be checked automatically for completeness and consistency. Because it is separate from format, one source produces several outputs, and a change to a shared component reaches every deliverable that uses it. Translation and review also work on components rather than whole documents, so effort follows what changed.
The costs are front-loaded. Someone has to model the content: decide what a component is, which types exist, and what metadata they carry. Authors need training and an editor that enforces the rules. Legacy documents have to be converted or kept alongside the new source (our guide to convert or ingest covers that choice). And storage, branching, and workflow for many components usually lead to a component content management system.
The Decision
When Is It Worth It?
It pays when you can point to the problem. The same content is maintained in several places. Variants of a product or a customer multiply. The same material goes out in more than one format. A contract or regulator requires a specific structured standard. Quality problems trace to inconsistency that a rule could have caught.
It does not pay for a handful of stable documents in one format. Before committing, work through the structured content readiness checklist, and read why authoring teams need a CCMS for what happens once the volume arrives. If the data may also feed software, the XML and JSON comparison is a useful next read.
FAQ
Questions We Hear
What is structured authoring in technical writing?
Writing content as typed, validated components, such as topics or data modules, against a schema that says which elements are allowed where. Formatting is applied later by a stylesheet or pipeline, not by the author.
What is the difference between structured and unstructured authoring?
In unstructured authoring the author decides the layout and the file is the deliverable. In structured authoring a schema decides what can appear, the tool checks it, and the output is generated. The practical difference is that structured content can be validated, reused by reference, and published to several formats from one source.
Is structured authoring the same as XML?
Not exactly. XML is the most common carrier, and DITA, S1000D, and MIL-STD-40051 all use it, but the idea is the rule set, not the syntax. Some structured systems use other storage; what makes them structured is that content is constrained to a defined model and separated from formatting.
Do I need a CCMS for structured authoring?
No. You can author structured files in an XML editor and keep them in version control. A component content management system becomes worth it when reuse, variants, translation, and multiple authors make files hard to track. Our guide to what a CCMS is covers that threshold.
When is structured authoring not worth it?
For a small set of documents that rarely change, are not reused, and go out in one format. The cost of modeling, tooling, and training is real, so it should be paid against a problem you can name; the readiness checklist helps you name it.
Is topic-based authoring the same as modular authoring?
The two terms overlap and are often used interchangeably. Topic-based authoring stresses that each piece is a self-contained, typed unit that answers one question, such as a DITA topic or an S1000D data module. Modular authoring stresses that the pieces can be reused and assembled into different deliverables. Structured authoring adds the schema on top: the unit is validated against a model, and the formatting is applied separately.
Get In Touch
Moving From Documents to Structure?
We build, convert, and maintain structured content in DITA, S1000D, and MIL-STD-40051. Tell us what you have today and we will say what structure would change.