Glossary · 2 · Topics and information design
Progressive disclosure
Also known as: Progressive information disclosure, Layered information, Progressive information layering
In technical communication and interaction design, progressive disclosure is a design principle that presents the information or controls a reader needs for the current step and keeps the rest one deliberate step away, available on demand. Jakob Nielsen's 2006 description of the pattern — deferring advanced or rarely used features to a secondary screen so that applications are easier to learn and less error-prone — states its purpose: lower the entry cost for occasional users without removing capability for experts. In documentation it appears as short overviews that link to detail, expandable sections, layered help, tooltips and reference material kept outside the procedure. Progressive disclosure is a presentation and sequencing principle, not a standardized requirement, and deferring content does not change which information a product must be supplied with.
- information design
- technical communication
- chunking
- Reading behavior
- Findability
In one sentence
Progressive disclosure shows what the current step needs and defers the rest to a second level — in interfaces, help and documentation.
Example
The commissioning topic for a servo drive lists the six parameters needed for a standard start-up and links to a reference topic with the forty optional tuning parameters, so a field technician is not scrolling a parameter table while the machine waits.
Progressive disclosure comes from interaction design rather than from a standards committee. Its empirical roots are usually traced to work at IBM in the early 1980s, where John M. Carroll and colleagues studied "training wheels" interfaces and minimal instruction: novices made fewer errors and learned faster when advanced functions and advanced explanation were kept out of the way until they became relevant. The term became common vocabulary in user experience work through Jakob Nielsen's Nielsen Norman Group article of December 3, 2006, which summarizes the pattern as deferring advanced or rarely used features to a secondary screen, so that most users can finish their task without meeting that complexity at all while experts pay only one additional click. In technical communication the same idea has circulated under labels such as "progressive information disclosure" and "layered information," and it is one of the principles that connects minimalism with the question of how a topic is shaped and delivered.
The principle is about sequencing, not about hiding. Two conditions make it work: the first level must be enough for the common case, and the path to the second level must be labeled in the reader's language, so that a reader can tell what lies behind it before opening it. Where those conditions fail, disclosure turns into concealment — readers do not know that the information exists, and search and retrieval systems may not reach it either.
How it applies
- Put the common case on the first level. In a procedure, state the prerequisites, the normal path and the expected result; move variants, troubleshooting branches, optional parameters and background theory to linked topics or clearly labeled expandable blocks. This is a judgment about typical use, so it should rest on task analysis and support data, not on guesswork.
- Label the door, not just the room. "Show advanced print options," "Calibration for multi-axis setups" and "Why the interval differs for food-grade oil" tell readers what they will get; "More," "Details" and "Read on" do not. Weak labels destroy the information scent that readers use to decide — see information foraging.
- Do not defer safety-relevant content. Warnings, residual-risk statements and prerequisites belong with the step they apply to and visible by default, not behind a collapsed section, a tooltip or a tab. Which information must be supplied with a product, and how prominently, follows from the applicable regulations and standards for information for use and instructions for use — applying a design principle never establishes compliance by itself.
- Software documentation: disclose in the interface and in the help. Default views with an "advanced settings" area, wizards that ask one thing at a time, inline hints that link to a full reference topic, and context-sensitive help that opens the one topic matching the current dialog are all implementations of the same principle.
- Machinery documentation: disclose along the task and the role. An operating topic can carry the daily sequence, with maintenance intervals, lubricant specifications and spare-part numbers in referenced reference topics; service personnel reach the deeper level through a part number, an error code or a QR code rather than through the operator's page.
- Keep the layers addressable. If the deferred content is a separate DITA topic or iiRDS topic with its own metadata, it can be linked, filtered, searched and reused. If it exists only as script-driven markup inside another page, in-page search, print output, offline delivery and retrieval pipelines for AI assistants may all miss it.
- Check accessibility and output channels. Disclosure controls must be operable by keyboard and announced to assistive technology; collapsed content should remain findable by in-page search, and the print or PDF rendition should expand everything, since paper has no "show more."
- Limit the number of levels. Two levels usually suffice. Each additional level adds a decision and a chance for the reader to guess wrong, which is the common failure mode reported for deep drill-down structures.
Progressive disclosure vs. staged disclosure
Progressive disclosure is hierarchical: the first level is complete for most readers, and the second level is optional, so many readers never open it. Staged disclosure is sequential: the content is split into steps that every reader passes through, as in a multi-step form or an installation wizard. The practical difference is the exit condition — progressive disclosure assumes the reader can stop after level one, staged disclosure assumes the reader continues to the end. Mixing them produces the familiar defect of a wizard whose first screen is unusable without information that appears only on screen three.
Progressive disclosure vs. Every Page is Page One
Both principles answer the same observation, that readers arrive mid-document and read only what their task requires, but they act at different scales. Every page is page one argues that each topic should stay on one level and be self-sufficient, rather than being a node in a drill-down hierarchy that only makes sense after its parent pages. Progressive disclosure, applied in that setting, therefore works inside a topic (first the normal path, then optional detail) and outward through links to peer topics, not by stacking a chain of overview pages that readers never see. Used carelessly, a layered table of contents becomes exactly the hierarchy that web-delivered documentation is poorly served by.
External references
- Jakob Nielsen: Progressive Disclosure (Nielsen Norman Group, December 3, 2006)
- Nielsen Norman Group: Progressive Disclosure (video summary)
- Jakob Nielsen: Progressive Disclosure — From Training Wheels to Week-Long AI Agents (UX Tigers)
- Interaction Design Foundation: Progressive Disclosure (The Glossary of Human-Computer Interaction)
- Andrea L. Ames: Applying Progressive Information Disclosure to User Interface and Content Design (presentation slides)
- IEC/IEEE 82079-1:2019, Preparation of information for use (instructions for use) of products — Part 1: Principles and general requirements — the current edition; a committee draft for a third edition is listed at ISO and remains a draft.