| name | behavioral-design |
| description | Apply behavioral psychology, UX principles, and decision science to product design. Use when designing user flows, evaluating UX quality, or diagnosing adoption problems. |
Behavioral Design Skill
Purpose
Provide the behavioral psychology and UX design toolkit for building products that work with human nature, not against it. Covers decision science, cognitive psychology, emotional design, and habit formation.
Decision Science
Hick's Law
More choices = longer decision time = more abandonment.
- Threshold: >4 choices creates measurable decision fatigue
- Fix: Reduce choices, smart defaults, progressive disclosure
Decision Fatigue
Users have a finite daily budget of decisions. Each decision depletes it.
- Late-in-session decisions get less attention
- Users default to "do nothing" when fatigued
- Fix: Front-load important decisions, batch related choices
Default Effect
75-90% of users accept defaults.
- Defaults should serve the USER, not just the business
- Opt-out > Opt-in for desirable behaviors
- The default IS the experience for most users
Framing Effect
Identical information presented differently leads to different decisions.
- "90% success rate" vs "10% failure rate" feel different
- "Save $50" vs "Don't lose $50" trigger different responses
- Review every piece of copy for framing opportunities
Cognitive Psychology
Cold Start Problem
Zero-data states are where most users churn.
- Users need value within 60 seconds
- Pre-populate with examples or sample data
- Show what's possible before asking for input
- "Empty state = churn state"
Progressive Disclosure
Show only what's needed for the next decision.
- Hide complexity behind progressive layers
- Let users discover depth, don't force it
- Match information density to user sophistication level
Reactive Onboarding
Let users DO first, then teach.
- Do not block action with education
- After the user acts, explain what happened and why
- Learning by doing > learning by reading
- Tutorial overlays have <5% completion rates. Stop using them.
Doherty Threshold
Responses >400ms break the feeling of direct manipulation.
- Users perceive delays >1s as a different interaction mode
- Skeleton screens > spinners > blank screens
- Optimistic updates for user-initiated actions
- If delay is unavoidable: show useful information during the wait
Context Craving
Sometimes users want clarity MORE than efficiency.
- Not every flow should be minimized
- Key moments (payment, data deletion, account changes) benefit from extra context
- Match speed to stakes
Behavioral Psychology
Social Proof
People follow others, especially similar others.
- Types: User counts, testimonials, peer activity, expert endorsement
- Most effective: Showing what similar users did (not generic "1M users")
- Backfires when: Numbers are small, audience is irrelevant, feels manipulative
Loss Aversion
Pain of losing is 2x the pleasure of gaining.
- Users overvalue what they already have (endowment effect)
- Frame feature adoption as "don't lose X" rather than "gain X"
- Show users what they'll miss if they don't act
- Use ethically — this is powerful and easily abused
Commitment Escalation (Foot-in-the-Door)
Small compliance → larger compliance.
- Start with a tiny ask (one-click action)
- Gradually increase commitment
- Each completed action makes the next more likely
- Personalization before payment leverages endowment effect
Curiosity Gap
Incomplete information creates a craving to fill it.
- Tease value before delivering it
- "See how your metrics compare" > "Analytics dashboard"
- Progress indicators create curiosity about completion
- Don't be manipulative — deliver on the tease
Mimetic Desire
People want what they see others wanting.
- Showing demand signals increases desire
- "Trending" and "popular" labels drive behavior
- In B2B: case studies showing peer success
- Warning: drives imitation, not necessarily good decisions
Reciprocity
People feel compelled to return favors.
- Give value before asking for commitment
- Free tools, templates, insights → builds obligation
- Unexpected generosity is most effective
Emotional Design
Cognitive Dissonance
When beliefs conflict with actions, users feel discomfort.
- Do not remind users of behaviors that conflict with their self-image
- Do align messaging with the identity they want
- Users cope by: lying to themselves, changing behavior, or adopting offsetting behavior
- Products that surface contradictions lose users (Netflix "hours watched" backlash)
Aha Moment
The moment a user understands the true value.
- Everything before the Aha moment is at-risk retention
- Identify your Aha moment precisely
- Make the path to Aha as short as possible
- Measure what % of users reach it and optimize ruthlessly
Peak-End Rule
People remember the PEAK experience and the ENDING.
- Invest in making the peak moment amazing
- End every flow on a positive note
- A mediocre middle is forgiven if peak and end are strong
Accomplishment Signaling
The dopamine hit from completing things drives repeat behavior.
- Celebrate completions (progress bars, congratulations, milestones)
- Break large tasks into small wins
- Show cumulative progress ("You've saved 14 hours this month")
- Make effort visible — people value results more when they see the work
Anxiety Reduction
Anxious users don't convert.
- Identify anxiety points (payment, data sharing, irreversible actions)
- Add safety signals: "You can change this later," "Money-back guarantee," "Preview before submit"
- Undo > "Are you sure?" confirmation dialogs
Habit Formation
The Hook Model
Trigger → Action → Variable Reward → Investment
↑ │
└──────── investment creates ──────────┘
- Trigger: Internal (emotion, need) or external (notification, context)
- Action: Simplest behavior in anticipation of reward
- Variable Reward: Unpredictable positive outcomes (social, material, self-achievement)
- Investment: User puts something in (data, content, social capital) that improves next cycle
Habit Metrics
- Time between uses (decreasing = habit forming)
- % of sessions initiated by internal vs external trigger
- Usage without prompting/notification
