Imagine two versions of the same registration process. In the first, everything goes smoothly, but at the very end the form throws an incomprehensible error message and forces the user to re-enter their password. In the second, there's a minor hiccup choosing an account category along the way, but the final screen greets the user by name, confirms success, and immediately suggests a first step in the product. Which process will be remembered more favorably? Psychology points clearly to the second — and not because it was objectively easier.
What is the peak-end rule?
In the early 1990s, Daniel Kahneman and his colleagues ran a series of experiments that changed our understanding of how people evaluate past experiences. In one of the most well-known studies, participants submerged their hand in painfully cold water under two conditions. In the first, the hand stayed in 14°C water for 60 seconds. In the second, the same 60 seconds at 14°C were followed by an additional 30 seconds during which the water temperature rose slightly. Rationally, the second condition means more suffering — it lasts longer. Yet when asked which condition they'd prefer to repeat, most participants chose the longer one.
Kahneman drew a simple but powerful conclusion: people don't evaluate experiences as the sum of all their moments. Instead, our brains take a shortcut — they primarily remember the most intense moment (the peak) and the very end. This heuristic, dubbed the peak-end rule, proved remarkably robust, confirmed by subsequent research in contexts ranging from pain and pleasure to everyday consumer situations.
Why does this matter for design?
As UX designers and researchers, we operate on user experiences — we map them, analyze them, and try to improve them. The natural approach is to eliminate all friction points one by one, evenly. That approach isn't wrong, but the peak-end rule suggests that not all moments carry equal weight in user memory.
A product can have a mediocre onboarding, an acceptable checkout process, and several minor annoyances along the way — but if the peak moment (say, the instant when users first see the result of their work in a tool) is truly satisfying, and the end of the interaction leaves a positive impression, the overall experience rating will be surprisingly high.
The reverse is equally true — a product with a polished interface that fails at a critical moment (a payment error, lost data, unclear order status) will be remembered through the lens of that failure, regardless of how well everything else worked.
Peaks we design (and those we don't control)
It's worth distinguishing between two kinds of peaks in user experience. The first are negative peaks — moments of frustration, confusion, or anxiety. These are typically what we identify in usability studies. A participant pauses, furrows their brow, tries to go back to the previous screen — we know something went wrong. We want to eliminate or soften these peaks because they'll dominate user memory.
The second kind are positive peaks — moments of delight, surprise, or a sense of competence. These moments are harder to surface through systematic research because they're difficult to trigger under controlled conditions. A user in a testing lab is completing a task — they're unlikely to experience genuine delight. But in real usage contexts, these peaks exist and carry enormous weight.
Examples of positive peaks aren't limited to spectacular animations or unexpected rewards. They also include the moment when a search engine perfectly understands the user's intent, the instant when a complex form turns out simpler than expected, or the situation where the system proactively solves a problem before the user even reports it.
The end effect — an underappreciated design element
While peaks tend to attract the attention of researchers and designers, interaction endings often remain neglected. Confirmation screens, summary emails, logout screens — these are places that rarely go through a thorough design process. That's a mistake, because the peak-end rule is clear: the final impression shapes memory with comparable force to the most intense moment.
Psychologically, we're dealing here with the recency effect — information processed last is more readily available in working memory and more likely to be encoded into long-term memory. When a user finishes interacting with a product, the last screen, the last message, or even the last micro-interaction builds the frame through which they'll recall the entire experience.
What does this mean for researchers?
Awareness of the peak-end rule changes how we can think about designing research and formulating insights.
First, when analyzing user experiences, it's not enough to count problems and rank them by frequency. What also matters is where along the user journey a given problem occurs. A difficulty at the very end of a purchase flow will have a disproportionately larger impact on perceived experience quality than the same difficulty at the beginning.
Second, when designing studies, it's worth accounting for the temporal and emotional context of interaction moments. Retrospective studies — where we ask users about their experiences after the fact — will inherently reflect the peak-end rule at work. That's not a flaw in those studies; it's valuable information about how users actually remember the product. But if we want the full picture, it's worth supplementing them with methods that capture experience in real time.
Third, when formulating design recommendations, it's worth consciously prioritizing peak and end moments. This isn't about ignoring problems elsewhere, but about recognizing that design investment in the end of a process can yield a disproportionately large return in perceived product quality.
The manipulation trap
Knowledge of the peak-end rule can tempt a straightforward manipulation — design a spectacular ending, and users will forgive shortcomings along the way. This approach is not only ethically questionable but also shortsighted. The peak-end rule describes a mechanism of retrospective memory, but users don't live exclusively in their memories. Real-time frustrations lead to product abandonment during the interaction — before the user ever reaches a carefully designed finale. A memory heuristic is no substitute for solid usability foundations.
Wrapping up
The peak-end rule is one of those cognitive psychology findings that connects elegantly with the practice of experience design. It's not a recipe or a framework — it's more of a lens through which to view user experience as a narrative with key moments and an ending, rather than a uniform sequence of equally weighted interactions. Understanding how human memory selects and distorts experiences allows us — researchers and designers alike — to choose priorities more wisely and hit closer to what truly shapes a user's opinion of a product.