Peak-End Rule
Retrospective evaluations of some experiences disproportionately reflect their most intense and final moments rather than a simple average of every moment, which can make the remembered version of an experience diverge sharply from the experience as it was actually lived.
What Is It?
The peak-end rule describes a documented pattern in how memory constructs a retrospective evaluation of certain experiences: rather than simply averaging the quality of every moment, memory can weight the most intense moment, the peak, and the final moments, the end, disproportionately heavily, with duration sometimes receiving surprisingly little weight in the process. The finding grew out of research by Daniel Kahneman and colleagues in the early 1990s, including a well-known experiment in which participants held a hand in painfully cold water for a fixed period, then repeated the experience with the same painful period followed by a slightly less painful (but still uncomfortable) extension; despite the second trial involving strictly more total pain, participants consistently preferred repeating the second trial, because it ended on a comparatively better note. A 1996 clinical study by Donald Redelmeier and Daniel Kahneman tracked moment-to-moment pain during naturally occurring colonoscopy and lithotripsy procedures and found that patients’ retrospective evaluations were shaped disproportionately by especially painful and final moments rather than by total duration. A later 2003 randomized trial by Redelmeier, Katz, and Kahneman went further, deliberately extending some colonoscopies with an added, less painful final interval; those patients subsequently remembered the procedure more favorably despite it having lasted longer.
The rule reveals a genuine and important distinction between two different things that are easy to conflate: the moment-to-moment experience of something as it happens, and the remembered evaluation of that experience afterward. These aren’t the same measurement, and organizations, like individuals, tend to act on the remembered evaluation rather than the moment-to-moment reality, since the moment-to-moment reality typically isn’t the thing anyone actually has access to after the fact.
Why Does It Matter?
The pattern is directly relevant to both employee and customer experience design. A project that ran smoothly and reasonably for months but concluded in a chaotic, rushed handoff may disproportionately color how the whole project is remembered and discussed afterward, even if much of the preceding work went well. Conversely, a genuinely difficult project that happens to close with a strong final week may be remembered more favorably than its average day-to-day experience actually justifies. Retrospectives, exit interviews, and end-of-project surveys can be especially vulnerable to letting recent or extreme moments stand in for the whole episode, though the strength of this effect varies by the kind of experience and how it’s being evaluated, and the leap from controlled pain studies to multi-month organizational projects is a plausible extrapolation rather than something separately demonstrated.
What Changes Once You See It?
You start treating a strong ending as something worth deliberately protecting, not because a good ending fixes what came before it, but because it’s genuinely what most people will remember and repeat afterward. You get more careful, when gathering feedback in a retrospective or exit interview, to separate “how do you remember this experience” from “how did this actually go on an average day,” since those two questions can produce very different, equally honest answers from the same person. You also become more skeptical of a uniformly glowing or uniformly damning account of a long project, since it may be reporting the peak and the end accurately while genuinely misrepresenting the middle.
Common Misunderstandings
- It isn’t a license to engineer an artificially good ending onto a genuinely bad process and call the underlying problem solved. Manufacturing a better-remembered experience is a manipulation of memory, not an improvement of the thing actually being remembered, and that distinction matters, especially when the same people will experience the process again.
- It isn’t a claim that the middle of an experience doesn’t matter at all. It matters for the actual, moment-to-moment wellbeing of the people living through it; the rule is specifically about how the experience gets remembered and evaluated afterward, which is a separate question from how good it was to actually live through.
- It doesn’t mean every long project should be redesigned purely around its ending. Some experiences, especially ones repeated often or ones where the moment-to-moment reality has its own real stakes, deserve to be evaluated and improved on their actual average, not just their retrospective impression.
- It isn’t unique to negative experiences like the original cold-water and colonoscopy studies. Peak and end effects have also been observed in some positive and mixed experiences, a great product launch remembered mainly for its strong final moment, a demo remembered mainly for its most impressive point, although the pattern isn’t equally strong or universal across every kind of episode.
Diagnostic Question
Are we evaluating the whole experience, or letting its most intense and final moments stand in for everything that happened between them?
Explore Further
Field Notes
- None yet.
Related Field Guide
Origin
Identified by Daniel Kahneman, Barbara Fredrickson, Charles Schreiber, and Donald Redelmeier in “When More Pain Is Preferred to Less: Adding a Better End,” Psychological Science (1993), with further clinical evidence from Donald Redelmeier and Daniel Kahneman, “Patients’ Memories of Painful Medical Treatments,” Pain (1996), and a randomized extension by Donald Redelmeier, Joel Katz, and Daniel Kahneman, “Memories of Colonoscopy: A Randomized Trial,” Pain (2003).