# Peak End Rule

> Predict and shape memory formation by focusing on peak moments and endings rather than duration or average experience

- Skill: `lev-os/peak-end-rule` (Agent Skill)
- Install (CLI): `npx skillmds@latest add lev-os/peak-end-rule`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lev-os/peak-end-rule/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: lev-os (https://skillmd.com/u/lev-os)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/lev-os/peak-end-rule

---


# Peak-End Rule

**What**: A cognitive heuristic where people judge experiences based almost entirely on how they felt at the peak (most intense moment) and at the end, rather than the total sum or average of the experience.

**When to use**: When designing user experiences, services, medical procedures, or any interaction where memory formation and future behavior matter more than moment-to-moment experience.

**Introduced by**: Daniel Kahneman and colleagues (1993-1999) through colonoscopy and cold-water experiments

## Core Mechanism

Memory is not a faithful recorder—it's a storyteller that samples two key moments:
1. **Peak**: The most intense point (positive or negative)
2. **End**: The final moments

**Duration neglect**: Time spent in an experience has surprisingly little impact on memory. A 20-minute colonoscopy with a less painful end is remembered as better than a 10-minute one with worse ending.

## When to Apply

**Use when:**
- Designing customer experiences (retail, service, digital products)
- Planning medical procedures or unpleasant tasks
- Structuring presentations, performances, or events
- Creating learning experiences or training programs
- Building habit-forming products or rituals

**Skip if:**
- Real-time optimization matters more than memory (safety-critical systems)
- Experiences are too short to have distinct peaks/ends (< 1 minute)
- Users won't encounter repeated instances (one-time interactions)
- Ethical concerns about manipulating perception (medical informed consent)

## Execution Steps

### 1. Map the Current Experience Journey
Document all phases of the experience from start to finish. Identify where peaks (positive and negative) currently occur and how it ends.

### 2. Identify or Create a Positive Peak
Design one exceptional moment that stands out—a surprise delight, moment of achievement, or emotional high. This becomes the memory anchor.

### 3. Engineer a Strong Ending
End on an upbeat note even if the middle was challenging. The last 5-10% of the experience disproportionately shapes memory.

### 4. Minimize Negative Peak Intensity (If Unavoidable)
If painful/difficult moments are necessary, reduce their intensity even if it extends duration. Kahneman's colonoscopy study: Less painful for longer = better memory.

### 5. Create Multi-Peak Experiences
Don't blow the entire budget on one peak. Multiple moderate peaks can create a "highlight reel" effect while maintaining strong ending.

### 6. Test Memory, Not Just Satisfaction
Ask "How do you remember the experience?" rather than "How was it?" Measure recall and willingness to repeat.

### 7. Sequence Carefully (Improving vs. Declining Trends)
Experiences that improve toward the end are remembered better than those that decline, even with identical averages.

## Real-World Applications

**Healthcare (Kahneman's Research)**: Adding a less painful period at the end of colonoscopies made patients rate the experience better and return for follow-ups more readily—even though they endured more total pain.

**Disney Parks**: End rides with a climactic moment, ensure visitors leave through gift shops with positive associations, design fireworks shows as park-closing peaks.

**Software Onboarding**: Create "aha moment" as early peak (user achieves quick win), end each session with progress visualization or positive reinforcement.

**Restaurants**: Strong opening (greeting, bread) + memorable dish (peak) + excellent dessert/farewell (ending) matters more than every intermediate course being perfect.

## Key Indicators

**Signs you're applying it well:**
- Users recall specific moments rather than duration
- Positive word-of-mouth focuses on peaks and endings
- Willingness to repeat exceeds real-time satisfaction scores
- Memory improves over time (rosy retrospection)

**Red flags:**
- Duration complaints despite good peaks/endings (you may need baseline comfort)
- Negative peaks overshadow positive ending
- Ending feels tacked-on rather than natural culmination
- Ethical manipulation concerns from users or team

## Common Mistakes

**Ignoring negative peaks**: You need to manage both. A terrible moment in the middle can dominate memory despite great ending.

**Weak endings**: Letting experiences fizzle out wastes the most impactful memory moment.

**Confusing satisfaction with memory**: Real-time feedback ("How are you feeling now?") doesn't predict memory-based decisions.

**Over-optimizing for memory at the expense of ethics**: Manipulating medical recall without improving actual outcomes crosses ethical lines.

## Related Frameworks

**Complementary**: Duration Neglect (same research lineage), Availability Heuristic (memorable moments shape judgment), Recency Effect (why endings matter)

**Contrasting**: Total Utility (classical economics assumes all moments matter equally), Experience Sampling (measures real-time vs. remembered utility)

**Sequential**: First map experience → Identify/create peaks → Engineer ending → Measure memory vs. real-time experience

## Practical Examples

**Uber/Lyft**: End ride with driver name, rating prompt, and "You saved X minutes" messaging—positive ending after potentially stressful trip.

**Apple Unboxing**: Peak moment when device revealed, positive ending with "Designed by Apple in California" closing statement, setup assistant that ends with "Welcome to iPhone."

**Sprint Retrospectives**: Even if sprint was difficult, end retro with wins/appreciations so team remembers period positively and stays motivated.

## Measurement

**Quantitative signals:**
- Net Promoter Score (measures memory-based recommendation)
- Retention/repeat usage rates
- Time to next engagement (sooner = better memory)
- Recall accuracy tests (what do users remember?)

**Qualitative indicators:**
- Story themes in testimonials (do they mention peaks/endings?)
- Emotional language intensity in reviews
- Comparison to real-time sentiment data
- Word-of-mouth content analysis

## Scoring Criteria

**Practitioner Weight**: 10/10 — Kahneman is a Nobel laureate who empirically validated this through controlled experiments with real medical procedures

**Clarity & Executability**: 9/10 — Clear two-point focus (peak + end), actionable for designers, though requires thought about what constitutes a peak

**Proven ROI**: 9/10 — Demonstrated in healthcare compliance, customer experience design, and has been widely adopted by top product companies

**Novelty**: 8/10 — Counterintuitive that duration doesn't matter; challenges rational economic models of utility

**Cross-Domain Applicability**: 9/10 — Applies to UX, healthcare, education, relationships, service design, entertainment, presentations

**Total Score**: 45/50 (Tier 1: Canonical)