Skip to main content

Linear vs Non-Linear UX

Linear UX​

At the heart of traditional user experience design lies the linear workflow — a predictable journey where the user’s destination is anticipated and every step can be meticulously planned and journey mapped to within an inch of its life. They are straight forward, and to borrow a term from literature, paratactic (workflows are one after the other).

The characteristics of linear UX:

  • A well defined goal
  • They are easy to test.
    • With limited steps and places to go wrong, observing users provides simple solutions when things do go wrong. Users may also have done similar tasks in other environments in the past allowing for self correction.
  • Known entry and exit points.

Non-Linear UX​

There are various characteristics of non-linear apps:

  • The end goal can be exploratory, unknown, ever changing, or dependent on ability.
  • They can have iterative workflows
    • A user takes several steps towards a goal
    • They decide to change something
    • They go back and alter/ edit some states
    • And continue in this vein
  • No set entry/ exit point.
  • The first step of a workflow may not be known.
    • What do you do when you start a new spreadsheet document?
    • Set up styling options.
    • Input Data.
    • Name columns.
    • Import from another source.
    • Make macros.
    • Any of these and more, in any order.
  • At what point are you done?
    • Are you finished when you meet your end goal?
    • Did you run out of time?
    • Was there a missing feature?
    • Was the task beyond yout current skill level?
    • Will you come back to continue later?
    • Did you need to export data to work with it somewhere else?
    • And so on.
  • They can involve using data from external sources.
  • It's possible that more than one person is required to operate the app at the same time.

Again, to borrow a term from literature, they are, hypotactic (one workflow inside another, and another inside that, and so on).

The Seven Traits of Complex Problem Solving​

Barbara Mirel, in her seminal Interaction Design for Complex Problem Solving, identifies seven defining traits that distinguish complex non-linear work from well-structured tasks.1 These traits can serve as a checklist — the more of these your software's domain exhibits, the more it demands non-linear design thinking:

  1. Ill-defined goals that clarify only through the process of inquiry itself
  2. Stakeholder diversity with competing priorities and perspectives
  3. Complex, heterogeneous information from multiple sources and formats
  4. Dynamic emergence — the problem shifts and evolves as the user investigates
  5. No preset entry or stopping points — users begin and end based on judgment, not procedure
  6. Iterative, opportunistic patterns — regularity mixed with improvisation and serendipity
  7. "Good enough" solutions — there is no single correct answer, only provisional resolutions

Mirel calls the design target for this kind of work "structured openness"1 — support for the recognisable professional patterns that users follow, without constraining the improvisation, backtracking, and emergence that make complex work productive. The software should scaffold professional conventions without enforcing them.

Footnotes​

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