UX & Design

Optimal UX design patterns for shareable, socially-engaging AI research report pages — combining the

Dark-mode-only is the single highest-friction barrier to mass adoption and shareability. All four providers independently confirmed that light mode produces superior reading comprehension for long-form text, that professional content platforms universally default to light mode…

AI Research Query UX Patterns

The "Research Plan Preview" pattern is the single highest-priority implementation: All four providers independently converged on Gemini Deep Research's editable plan as the strongest model for a high-stakes ($1–$60) multi-provider platform. Showing users an AI-generated…

Optimal Research Report Ordering

Move Detailed Synthesis to Position 2, immediately after Executive Summary. All eight providers converge on this recommendation, grounded in the Minto Pyramid Principle, intelligence community BLUF tradecraft, and Nielsen Norman Group eye-tracking evidence showing readers…

Citation UX and Evidence Navigation

Progressive disclosure is the universal master pattern: Every provider independently confirmed that hiding citation metadata behind hover/tap interactions — rather than displaying it inline — is the single most effective technique for balancing verifiability with readability…

AI Research Platform Onboarding Strategy

Time-to-value is the single most critical variable: All 8 providers independently confirmed that users make a "stay or go" decision within 60–90 seconds, and that 60–80% of new users abandon a product after a single session if they don't experience immediate value. For…

UX Patterns for Dense Research Reports

Progressive disclosure is the single most universally validated pattern across all five providers: structure reports in at least three tiers (executive summary → narrative with inline evidence → full claim/source explorer), ensuring casual readers get a clean 500–800 word…

AI Query Refinement UX Patterns

Progressive disclosure is the single most universally validated pattern across all four providers: layer complexity into 3 tiers (inline summary → expanded refinements → deep-dive reasoning), auto-expand only the highest-confidence section on first load, and limit visible…

Secure Guest Document Access Tokens

High-entropy, cryptographically random tokens are non-negotiable: All four providers independently confirm that tokens must use CSPRNGs with a minimum of 128 bits of entropy (256 bits recommended for long-lived tokens), URL-safe base64 encoding, and must never be derived from…

Document Sharing UX Patterns

The "Share Dialog" is the canonical UX pattern across all major platforms (Google Drive, Notion, Figma, Dropbox), functioning as a single modal hub that uses progressive disclosure to balance simplicity for casual users with granular control for power users. All four providers…

Inline Annotation UX Patterns

The select-to-annotate popover with margin placement is the universal foundational pattern: All four providers independently confirmed that text selection triggering a contextual action (highlight/comment) combined with margin-anchored comment cards is the dominant interaction…

Conversational UX for Research Results

Progressive disclosure + content-grounded empty states are the single highest-leverage pattern for new user onboarding: all four providers independently confirmed that replacing blank chat boxes with report-derived starter prompts dramatically reduces blank-page anxiety and…

Prompt Refinement UX Patterns

The "asymmetric control" principle is the single most important design constraint: the system proposes, educates, and suggests, but the user always retains override authority and can submit at any moment. Every provider independently converged on this principle, making it the…

Research how premium intelligence and research platforms present their final deliverables to maximiz

The Pyramid Principle is non-negotiable for premium perception: All four providers independently confirmed that leading with the conclusion (Situation-Complication-Resolution + MECE structure) is the single highest-impact, lowest-effort change that separates a $15-60 deliverable…

Deep Engagement Mechanics for Research Platforms

Separating agreement from quality assessment is the single highest-impact design decision for intellectual platforms. LessWrong's two-axis voting system (karma vs. agreement) is independently validated by five providers as the most important structural choice, preventing echo…