Cognitive Psychology
Peak-End Rule
Memory of an experience tracks the peak moment and the end, not the average
Daniel Kahneman, Barbara Fredrickson, and colleagues (1993, 1996, 2000) showed that retrospective evaluations of an experience are dominated by two moments — the peak (best or worst) and the end. Duration is largely ignored ("duration neglect"). In a striking colonoscopy study (Redelmeier and Kahneman 1996), patients who underwent a longer procedure with milder pain at the end remembered it as less unpleasant than patients whose shorter procedures ended at peak intensity. Cold-pressor experiments (Kahneman et al. 1993) showed people prefer a longer painful experience to a shorter one if the longer ends with reduced pain. The findings inform a "two selves" framework — the experiencing self versus the remembering self — which often disagree. Practical applications: hospital procedure design, hospitality, customer experience, end-of-life care, and product design all benefit from optimizing peaks and endings rather than reducing total negative time.
- ResearchersKahneman, Fredrickson, Redelmeier (1990s)
- Core findingMemory tracks peak + end, not average or duration
- Famous studyColonoscopy (Redelmeier & Kahneman 1996)
- Cold-pressor studyKahneman et al. (1993)
- Related conceptDuration neglect
- Two selvesExperiencing self vs. remembering self
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why peak-end matters
- Healthcare. Procedure and end-of-life design optimize peaks and endings, not averages.
- Customer experience. Onboarding peaks and goodbye moments shape Net Promoter Score.
- Hospitality. Theme parks, hotels, and cruises engineer memorable bookends.
- Education. Course endings (final lecture, summative project) outweigh midterms in remembered satisfaction.
- Product design. First-run and exit experiences punch above their weight in reviews.
- Therapy. Session structure benefits from emotional peaks and gentle endings.
- Personal life. Plan endings deliberately — vacations, relationships, projects.
Common misconceptions
- Average experience drives memory. It largely doesn't; peaks and ends dominate.
- Length matters. Duration is largely neglected in retrospective evaluation.
- Optimizing for the experiencing self optimizes memory. The remembering self's verdict can disagree.
- The rule licenses ignoring average quality. It doesn't — averages still affect the experiencing self's actual life.
- Endings can be tacked on. Forced or inauthentic endings can damage trust and feel manipulative.
- The peak must be positive. The peak is the most extreme moment in any direction; pain peaks dominate similarly.
Frequently asked questions
What was the colonoscopy study?
Redelmeier and Kahneman (1996) studied 154 patients undergoing colonoscopy without sedation (then standard). Patients reported pain every 60 seconds. Procedures lasted 4-69 minutes; pain varied. Crucially, half were randomized to a control procedure that ended at peak intensity; the other half had an extended ending — three additional minutes with the scope still inserted but motionless, providing milder discomfort. Despite enduring more total pain, the extended-end patients rated the experience less aversive and were more likely to return for follow-up colonoscopy.
What was the cold-pressor study?
Kahneman, Fredrickson, Schreiber, Redelmeier (1993). Participants submerged a hand in 14°C water for 60 seconds (short trial), then later in 14°C water for 60 seconds plus 30 additional seconds at slightly warmer 15°C (long trial). When asked which to repeat, two-thirds chose the long trial — preferring more total pain because it ended better. The result is counterintuitive but robust; participants chose worse experiences to live through in exchange for better memories.
What is duration neglect?
When evaluating a remembered experience, total time matters surprisingly little. A 30-minute meeting with a great peak and ending is remembered fondly; a 5-minute meeting with the same peak and ending is remembered similarly. This implies that brief unpleasant moments at the end of a long pleasant experience can dominate the retrospective verdict — the famous "ruined the whole trip" pattern when a vacation ends badly.
What's the two-selves framework?
Kahneman's distinction in Thinking, Fast and Slow (2011). The experiencing self lives moment to moment; its well-being is the average pleasure across time. The remembering self constructs narratives shaped by peaks and ends; its well-being is the rated quality of summarized experience. Choices depend on the remembering self (we plan for memories) but well-being depends on the experiencing self. The two often disagree — explaining why people pursue intense, photogenic experiences over more steadily pleasant alternatives.
How is it applied in healthcare?
Procedural and end-of-life care increasingly designs for the rule. Slightly extending a procedure with a benign closing reduces remembered suffering and improves follow-through with future care. Hospice and palliative care prioritize meaningful peaks and gentle endings; hospital discharge experiences can outweigh weeks of better-quality care if poorly handled. Patient satisfaction measures, increasingly tied to reimbursement, are systematically biased toward peak-end memories.
How is it used in hospitality?
Hotels, restaurants, and theme parks deliberately design memorable peaks and endings. Disney parks engineer "hello and goodbye" moments. Cruise lines bookend trips with welcome and farewell rituals. Restaurants offer complimentary mignardises after dessert. Recovery from poor service is most leveraged in the final interaction; a gracious resolution after a problem can produce higher loyalty than a problem-free experience.
What are limitations?
Several. (1) The rule emerged from short, controlled experiences; long-running events (years-long relationships, careers) follow more complex memory dynamics. (2) Individual differences exist — some people weight average more heavily. (3) Peaks must be salient and clearly demarcated. (4) Highly emotional or self-defining moments can override the rule (a single trauma may dominate regardless of position). (5) Repeated experiences blur distinct peaks and endings.