── ── Mental model
User Insight Engine
User research produces data. The Engine produces insight — the causal link between an observable behavior and the deep-layer driver that causes it. Three layers: Surface (what users say), Behavioral (what users do), Deep (why — cognitive and social drivers). Four Deep Layer drivers: Loss Aversion, Social Proof, Cognitive Load, Trust Cost.
How it works
Step 1 — Behavior Gap: "[User segment] is [doing/not doing] [specific action] at [workflow point], despite [Surface Layer evidence they intend to / are capable of doing it]."
Step 2 — Three-Layer Synthesis: Collect Surface (surveys, interviews, tickets), Behavioral (clickstream, funnels, session recordings, cohort retention), Deep (ethnographic observation, JTBD interviews, diary studies, driver-isolating A/B). Stop-rule: no Deep Layer evidence = analysis incomplete; do not proceed to interventions.
Step 3 — Map to Driver: High trial-to-abandonment → Trust Cost / Cognitive Load. Low feature adoption despite awareness → Cognitive Load / Social Proof. Sudden drop-off after initial engagement → Social Proof / Trust Cost. Feature used in unintended order → Cognitive Load.
When to use it
- user says 'why aren't users doing X despite saying they would', or 'our survey scores are high but churn is high', or 'we shipped a feature that tested well but nobody uses it', or 'we have conflicting signals from research', or 'I can see the drop-off point but don't know why'
When not to use it
When the decision is routine and reversible, applying a formal method costs more than it returns.
Worked example
Taylor's Behavioral Observation at Bethlehem Steel (1898–1901)
Source: Frederick Winslow Taylor, The Principles of Scientific Management (1911, Harper & Brothers). Available at: https://archive.org/details/principlesofscie00taylrich
Install this skill (free, MIT)
npx skills add deciqAI/knowledge-skills