The Peak-End Rule
Learn it → Apply it → Save as Skill
People remember an experience by its most intense moment and how it ended, not by its average or its length. Two experiences that were equally long and equally difficult overall can be remembered very differently depending on where they peaked and how they closed.
Learn it. Apply it. Save as Skill.
Dr. Wallace's Favorite Mental Models is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
When people recall an experience later, their rating tracks the single most intense moment (the peak, good or bad) and the final moments (the end) far more than it tracks how long the experience lasted or how it averaged out.
The memorable shorthand
Important: the middle bars barely register in memory. Only the tallest moment and the last moment carry disproportionate weight.
Use it when someone will remember this later.
The model matters most whenever a person's later memory, rating, or retelling of an experience will shape a decision: a repeat purchase, a referral, a hiring judgment, a review.
Map → Find the peak → Engineer the end → Check duration.
The rule tells you where to spend your limited attention: on the single most intense moment and the closing moments, not evenly across the whole experience.
Chart the experience over time. Mark where intensity spikes, up or down, and how it currently ends.
Identify the single most intense moment, the best or the worst. That moment will dominate what gets remembered.
Deliberately design the final moments. A resolving or positive close carries weight far out of proportion to its length.
Resist over-investing in shortening or smoothing the middle. Duration is remembered far less than people assume.
Recognize the situation.
Open a use case to see what the model changes.
Customer Service
A support interaction has painful wait time
Hold time alone rarely determines the remembered rating. Find the worst moment in the call (being transferred, repeating a story) and fix that first, then make sure the close is a genuine resolution rather than an abrupt hang-up.
Product & UX
An onboarding flow has one rough step
Smoothing every screen equally is a poor use of design time. Locate the single worst point of friction and fix that, then make the final screen feel like a small accomplishment rather than another form field.
Hiring
An interview had one weak answer in an otherwise strong conversation
An interviewer's later memory of the whole conversation leans on its most striking moment and how it closed. Be aware this can distort your own evaluation, and close interviews deliberately rather than letting them trail off.
Healthcare
A procedure or appointment is uncomfortable
Clinicians can measurably reduce a patient's remembered distress, and increase the odds they return for future needed care, by managing the intensity of the final moments, even when total duration or total discomfort does not change.
Events & Hospitality
A multi-day trip, course, or conference is being planned
The closing session or final hour shapes the review disproportionately. A strong close is often worth more than filling every hour of the schedule evenly.
Negotiation
A session ends on a transactional note
How you end a negotiation session, a small final gesture, a warm close, a moment of respect, can shape how the other party remembers the whole negotiation, separately from the actual terms reached.
Same procedure, same total pain. Different ending. Different memory.
This is a simplified illustration of the actual research design used in the original colonoscopy studies (see History & Accuracy below).
The finding has a specific research trail, not just a catchy name.
Expand for the studies behind the rule and where its predictive power runs out.
Who actually formulated the peak-end rule?
Barbara Fredrickson and Daniel Kahneman described it in a 1993 paper as part of a broader "snapshot model" of remembered experience: the idea that memory retains a few representative moments rather than an accurate running average of an entire episode.
What did the colonoscopy research actually show?
Donald Redelmeier and Daniel Kahneman studied patients undergoing colonoscopy and lithotripsy procedures in 1996. Retrospective pain ratings tracked the intensity of the worst moment and the final moments far more closely than they tracked total procedure duration or the moment-to-moment average. A 2003 follow-up by Kahneman, Redelmeier, and Joel Katz found that patients whose procedures were deliberately extended by a few minutes, with discomfort eased before the scope was withdrawn, rated the overall experience as less unpleasant and were more likely to return for a future scheduled procedure.
Does the rule hold for every kind of experience?
No. Researchers have identified real boundary conditions. Multi-day experiences, such as vacations or hotel stays, do not compress as cleanly into a single peak and a single end. For expensive or high-stakes purchases, first impressions can outweigh the peak or the ending. A 2008 study found the rule was a real but imperfect predictor of remembered value, and individual differences matter: in food-preference studies, "restrained eaters" show a noticeably weaker peak-end effect than unrestrained eaters.
Simple does not mean foolproof.
Smoothing the middle instead
Evenly improving every step of an experience is a poor use of limited effort. Find and fix the single worst moment, or amplify the single best one, before you touch anything else.
Ending on logistics
Closing on a form, an invoice, or "please fill out this survey" wastes the highest-leverage moment of the whole experience. Close on something resolving and human first.
Assuming duration matters as much as intensity
Cutting an experience short without addressing how it feels at the finish rarely improves memory of it. A slightly longer experience with a gentle end is often remembered better than a shorter one that stops abruptly.
Treating it as license to manipulate
Engineering a good ending to paper over a genuinely deficient middle is a design choice with real ethical limits, especially in healthcare, service recovery, and any consent-sensitive context.
Missing a peak that isn't at the end
The rule cares about the peak wherever it falls, not only the finale. Some experiences have their most intense moment in the middle, and that moment still needs deliberate attention.
Applying it uniformly to long, multi-episode experiences
A multi-day trip, a year-long project, or a multi-session course does not compress into one peak and one end as cleanly as a single short episode. Each distinct episode may need its own peak-and-end treatment.
Some situations resist compression into a peak and an end.
Use full-experience measures when...
the experience spans many days or many distinct episodes; a single peak-and-end story compresses a vacation, a course, or a multi-month engagement poorly.
Use cumulative outcome metrics when...
you are optimizing for something measurable and additive, such as total pain relief delivered, total value received, or safety outcomes. Remembered pleasantness is not the same as actual benefit.
Use direct honesty over "ending well" when...
the ending is being engineered specifically to disguise a genuinely deficient middle. Managing memory of a problem is not a substitute for fixing it.
One model rarely does the whole job.
Build a peak-and-end redesign prompt.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Your AI-ready prompt
Turn this mental model into a reusable AI skill.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply the Peak-End Rule, what questions to ask, what evidence to seek, what mistakes to avoid and how to structure its analysis.
Build your own Mental Models Skills Library
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
For readers who want the nuance.
Is the peak always a bad moment?
No. The peak is whichever moment is most intense, positive or negative. A single delightful surprise can be the peak of an otherwise ordinary experience, and it will be remembered disproportionately too.
Why does duration matter so little to memory?
Researchers call this "duration neglect." Memory does not store a running average of an experience over time; it stores a small number of representative moments. Because duration is not one of the moments stored, it contributes far less to the remembered evaluation than people expect it to.
Does the peak-end rule mean the experience itself doesn't matter, only the memory of it?
Not for every purpose. If your goal is a decision that depends on memory, a return visit, a referral, an interview score, a review, the peak and the end are what you should manage. If your goal is the actual outcome delivered, such as a health result or the amount of real value created, remembered pleasantness is a separate thing worth tracking on its own.