Judging an experience mainly by how it felt at its most intense moment and at the very end, not the overall average.
Imagine you go to a birthday party. You played boring games for two hours, but then there was an amazing magic show, and at the very end, you got the biggest slice of cake. When someone asks you about the party, you say it was awesome — because your brain only really saved the magic show and the cake, and forgot the boring parts in between.
The Peak-End Rule describes how retrospective evaluations of experiences are disproportionately shaped by two specific moments: the point of greatest emotional intensity (whether positive or negative) and the final moments before the experience concludes. The overall duration of the experience has remarkably little influence on how it is remembered, a related phenomenon known as duration neglect. This means that a long, moderately pleasant experience can be remembered less favorably than a shorter one that contained a vivid high point and a satisfying ending. The bias operates across both painful and pleasurable experiences, distorting memories of medical procedures, vacations, customer interactions, and everyday events in ways that often diverge significantly from what moment-by-moment reports would predict.
The same glitch looks different depending on the terrain. Finance, medicine, a relationship, a team — same mechanism, different costume.
Investors tend to evaluate the quality of a trading period based on the most extreme gain or loss and the portfolio's final position, rather than on cumulative returns. This leads to preferring volatile strategies that ended well over steadier approaches with higher total yield, and it distorts reference pricing — consumers anchor on the highest price seen and the most recent price rather than the average.
Patients' willingness to return for preventive screenings such as colonoscopies is strongly influenced by how painful the worst moment was and how the procedure ended, rather than total duration of discomfort. This has led to clinical strategies where practitioners deliberately taper pain at the end of procedures to improve patient recall and compliance with future screenings.
Students' retrospective evaluations of a course are disproportionately shaped by the most emotionally charged moment (a particularly exciting or frustrating class) and the final session. A course with mostly mediocre lectures but a powerful final project presentation may be rated higher than a consistently good course with a flat ending, distorting teaching evaluations.
People's overall assessment of a relationship, a date, or a shared experience is heavily colored by the most emotionally intense interaction and how things ended. A long relationship with years of steady contentment but a bitter breakup is often remembered primarily as painful, while a brief but passionate fling that ended on a sweet note is recalled fondly.
Product designers exploit this bias by investing disproportionately in 'wow moments' and the offboarding or checkout experience. A smooth, delightful final confirmation screen (like Mailchimp's celebratory animation after sending a campaign) can dramatically improve user satisfaction scores even if earlier workflow steps were frustrating.
Employees' annual satisfaction is often driven by the most emotionally intense event of the year (a major win or a painful conflict) and the tone of their year-end review, rather than the day-to-day quality of their work life. Managers who deliver negative feedback early and end reviews on a constructive note tend to leave employees with more positive retrospective evaluations.
Voters' memories of a political term or campaign are dominated by the most dramatic event (a scandal, a crisis response) and the candidate's final impression before the election. A presidency with steady governance but a chaotic final month may be remembered far more negatively than performance data would suggest.
Barbara Fredrickson and Daniel Kahneman, 1993. Formalized in the paper 'Duration Neglect in Retrospective Evaluations of Affective Episodes' (Journal of Personality and Social Psychology, 1993) and the landmark experiment 'When More Pain Is Preferred to Less: Adding a Better End' (Psychological Science, 1993) by Kahneman, Fredrickson, Schreiber, and Redelmeier.
In ancestral environments, the most emotionally intense moment of an encounter — such as the peak danger during a predator attack or the peak reward during a successful hunt — carried the most survival-relevant information about whether to repeat or avoid that situation. The ending of an event similarly signaled the final outcome (escape or capture, fed or still hungry). Encoding these two data points was far more computationally efficient than tracking every moment, allowing rapid future decision-making about threats and opportunities.
Recommendation systems and sentiment analysis models that weigh all moments of user interaction equally may fail to predict user satisfaction, because humans evaluate experiences via peak and end moments. AI systems trained on retrospective user ratings inherit the bias — they learn to optimize for peak intensity and endings rather than consistent quality, potentially leading to manipulative design patterns that game user memory rather than improving overall experience.
50 cards are free to preview. Buyers unlock the rest of the deck plus the interactive training — Spot-the-Bias Quiz unlimited, Swipe Deck with spaced repetition, My Blindspots, Decision Pre-Flight, the Printable Deck + Cheat Sheets, and the Field Guide e-book. $29.50$59.
