The average duration of a user's continuous interaction period with a catalogue or discovery interface, measured from session start (first action or page load) until session termination by explicit end or inactivity threshold. Definitions must state the inactivity timeout, treatment of background page activity, cross‑device continuation rules, and aggregation method (mean versus median).

Principle

Session length is an ambivalent usability indicator: longer average sessions may reflect deeper engagement or greater search difficulty; interpretation requires task context, user intent segmentation and complementary success metrics (clicks, item views, conversions).

Demonstration

Illustrative scenario → Using a 30‑minute inactivity timeout, a library logs session durations for 1,000 sessions and computes mean and median. Mean = 18 minutes, median = 7 minutes; analysis segments sessions by task type (known‑item lookup, research) to interpret whether length indicates search difficulty or legitimate engagement.

Misapplication

Interpreting a single session‑length statistic as a universal measure of usability without segmenting by task, user type, device, or considering outliers—leading to erroneous conclusions such as 'long = bad' or 'short = good'.

Consequence

Useful for interface tuning, timeout policy, personalization and support prioritization; misinterpreted session length can drive inappropriate UI changes, penalize legitimate research behaviour, or misallocate support resources.

Reversal

For research or exploratory tasks, long sessions are expected and desirable; for transactional lookups, short sessions with immediate success are preferable. Therefore median or task‑stratified metrics may reverse the signal given by the mean.

Boundary

Clearly within: continuous interaction windows recorded in the discovery interface with stated inactivity rules. Boundary case: sessions that span devices where continuity is tracked versus those counted as separate sessions. Clearly outside: cumulative account usage over days or long‑term engagement metrics.

Semantic Tension

Engagement (longer sessions indicating depth) ↔ Efficiency (shorter sessions indicating quick success): the same session length can be read as positive or negative depending on task and outcome metrics.

Synthesis

Search Session Length is a contextual performance indicator: its value arises only when paired with task classification and success measures, and when session delimitation rules are explicit.