Shannon Entropy Audit

Description

Quantifies the uncertainty a user faces at each decision point or in an information space using Shannon entropy, giving a principled complexity measure for IA and navigation.

Shannon entropy, H = -Σ p log2 p, measures the average uncertainty in a probability distribution in bits. Applied to interfaces, it quantifies how unpredictable a user's next choice is at a decision point, how evenly traffic spreads across navigation items, or how much information a notification stream carries. An entropy audit computes these values from usage logs or from structural models, then compares them with what users need. High entropy at a point where most users want the same thing indicates poor prioritisation; very low entropy across a whole menu indicates items nobody uses. Rigour comes from defining the random variable precisely, using sufficient log data for stable probability estimates, and pairing entropy with outcome measures such as success rate so that lower entropy is not mistaken for better design in itself.

Attributes

Loss type
Latent
Effort
Days

Steps

  1. Define the decision points or distributions to measure (e.g. first click on the home page, menu selections, notification types).
  2. Extract choice frequencies from logs over a representative period, ideally thousands of events per point.
  3. Compute entropy and normalised entropy (H / log2 n) for each point.
  4. Compare entropy with task success and time to identify points where uncertainty is high and outcomes poor.
  5. Model redesigns by estimating new distributions and the resulting entropy reduction.
  6. Validate changes with A/B tests or tree tests and re-measure entropy post-launch.

Sources

  • Information Theory Inference and Learning Algorithms
  • A Mathematical Theory of Communication (Claude Shannon)
  • MacKay, D. J. C., Information Theory, Inference, and Learning Algorithms (2003)
  • Kolmogorov complexity literature (MIT Press)
  • Shannon, C. E., A Mathematical Theory of Communication (1948)
  • MacKay (Information Theory)
  • Kolmogorov via MIT Press
  • Shannon's works
  • The Inquiry Series, Ch 1 — Entropy, Complexity & the Information Architecture of Interaction

Also known as

  • Shannon Entropy in UX
  • Information entropy analysis
  • Entropy audit
  • Shannon entropy audit
  • Choice-uncertainty audit
  • Information-theoretic complexity audit
  • Interface entropy audit

Categories

  • Information Theory, Physics & Mathematics