Skip to main content

Information Architecture for Complex Software

Information Architecture in complex professional software has almost nothing in common with IA on the web. Web IA organises content so people can find it — navigation hierarchies, taxonomies, search. Complex software IA organises the working environment so people can do their work: maintaining orientation across multi-session investigations, supporting non-linear exploration, and managing multiple simultaneous representations of the same data.

The shift is from content organisation to work organisation.1 Your users aren't browsing — they're investigating, creating, analysing, and iterating. Their IA needs to support the way professional work actually unfolds: with backtracking, parallel exploration, provisional judgments, and goals that emerge only as the work progresses.

Three Architectures Working Together​

Complex software demands three types of architecture functioning as an integrated whole:2

Data Architecture​

How data is structured, connected, and accessed across the application. Professional tools pull data from multiple sources, in multiple formats, with complex interdependencies. Your data architecture determines whether users can seamlessly cross-reference information or must manually export, transform, and reimport between disconnected views (a depressingly common outcome).

How users move through the application's capabilities and their own work — entry points, wayfinding cues, movement patterns. In non-linear software, navigation can't be a fixed path; it must support configurable traversal of a problem space. Users need to go where the work takes them, not where your sitemap predicted they'd go.

Representational Architecture​

How the same underlying data appears through multiple views. Professional software typically maintains 4–6 simultaneous representations: a spatial view, a properties panel, a timeline, a node graph, a data table, an output preview. The representational architecture determines how these stay synchronised, how changes propagate, and how users toggle between them without losing context.

The ICV Model​

Barbara Mirel's ICV model offers a framework for understanding the conditions that shape IA decisions in complex software:1

  • Interactivity: Conditions across multiple contexts that dynamically open or close possible actions. The interface must adapt to what is currently possible, not just what is theoretically available.
  • Contexts: Four overlapping contexts shape every user's work — the problem context, the technology/data context, the work domain context, and the subjective context (expertise, preferences, time pressure). IA decisions must account for all four.
  • Variability: Every professional workflow has stable elements (conventions, business rules, recurring patterns) and variable elements (situational factors, emerging insights, shifting priorities). The IA must support both — providing structure without rigidity.

What This Section Covers​

The following pages explore the key concepts for designing IA in complex professional software:

  • Task Landscapes — The unit of design for non-linear workflows, replacing traditional task hierarchies
  • Wayfinding — How users maintain orientation and navigate complex problem spaces
  • Multiple Representations — Managing simultaneous views of the same data

Footnotes​

  1. Mirel, B. (2004). Interaction Design for Complex Problem Solving: Developing Useful and Usable Software. Morgan Kaufmann. ↩ ↩2

  2. Mirel, B. (2004). Interaction Design for Complex Problem Solving: Developing Useful and Usable Software. Morgan Kaufmann. The three-architecture split (data, navigational, representational) is the author's synthesis, built on Mirel's account of complex problem solving rather than terminology drawn directly from her text. ↩