How do you apply the peak–end rule to customer experience?

Map the emotional trajectory of the journey, find the current peak and end, then deliberately engineer a positive peak and a strong close. Fix negative peaks first, because a single low point disproportionately shapes the remembered experience.

Why does the end of an experience matter so much?

Memory weights the final moments heavily, and the remembered experience — not the moment-by-moment reality — drives whether a customer returns, renews, or refers. A weak ending can undo an otherwise good experience.

Insights · Customer experience

The peak–end rule in customer experience.

Q: What is the peak–end rule?

The peak–end rule is a cognitive bias, identified by Daniel Kahneman, in which people judge an experience largely on how they felt at its most intense moment (the peak) and at its end — not on the average or sum of every moment. Duration is largely ignored.

PiqueInsights · Neuro CX · ~6 min read

Customers don't remember the average of an experience. They remember its most intense moment — the peak — and how it ended. That single quirk of memory, the peak–end rule, quietly decides whether they renew, return, and refer.

What is the peak–end rule?

The peak–end rule is a cognitive bias identified by Nobel laureate Daniel Kahneman. When people recall an experience, they don't tally every moment. Their judgment is dominated by two data points: the peak (the emotional high or low) and the end. The duration barely registers — a phenomenon called duration neglect.

In Kahneman's famous colonoscopy study, patients whose procedure was longer but ended more gently remembered it as less unpleasant than patients whose shorter procedure ended at a painful peak. More total discomfort, better memory — because the ending was softer. The remembered experience, not the real one, is what shapes future behavior.

Memory is built on the peak and the end. Duration is forgotten. Design accordingly.

Why it matters for CX and product

Most experience design tries to make every step "good." That's expensive and, per the peak–end rule, inefficient. Your NPS, renewals, and referrals are driven disproportionately by two moments most teams never deliberately design: the emotional high point and the last thing a customer touches. If your onboarding's "peak" is an error message, or your cancellation flow is the smoothest part of your product, the peak–end rule is working against you.

This is also why usage data and NPS often disagree. A customer can use a product constantly and still rate it poorly — because the remembered feeling is set by a painful peak, not the hours of fine-but-forgettable use.

How to apply the peak–end rule (4 steps)

Map the emotional trajectory. Walk the journey and plot how a customer feels at each step — not what they do, how they feel. Anxiety, relief, delight, confusion.
Find the current peak and end. Where's the emotional high or low? What's the literal last moment — the confirmation screen, the empty state, the support sign-off?
Fix negative peaks first. A single low point disproportionately damages the memory. Removing a bad peak beats adding a nice-to-have anywhere else.
Engineer a positive peak and a strong close. Create one deliberate moment of delight (a genuinely helpful surprise, a moment of progress made vivid) and make the ending land — a warm confirmation, a clear next step, a small win.

Peak–end rule examples

Onboarding: the peak should be the moment of first value — made unmistakable — not a wall of setup. End the session on a completed action, not a to-do list.
Support: the resolution is the peak; the sign-off is the end. A ticket that's solved but closed curtly still reads as a mediocre experience.
Retail/e-commerce: the unboxing or confirmation is the end most brands under-invest in — the cheapest retention lever there is.
Cancellation: a graceful end makes people come back. A punishing one guarantees they won't, and they'll tell others.

Engineer the peak and the end

The Peak–End Arc is one of PiqueInsights' Neuro CX frameworks. We map your journey's emotional trajectory and design the two moments memory is built on.

Book the Neuro CX audit →

Common mistakes

Optimizing the average. Smoothing every step while ignoring the peak and end spends effort where memory doesn't look.
Neglecting the ending. The last moment is the most under-designed and highest-leverage point in most journeys.
Trusting live sentiment over remembered sentiment. What people feel in the moment and what they'll recall next week differ — and the recall is what drives behavior.

FAQ

Who discovered the peak–end rule? Daniel Kahneman and colleagues, through research on how people remember and evaluate past experiences.

Is the peak–end rule the same as recency bias? Related but not identical. Recency bias weights the most recent information; the peak–end rule weights both the emotional peak and the end.

How is this different from just "good UX"? Good UX reduces friction everywhere; the peak–end rule tells you where to invest for memory and loyalty — the peak and the end specifically.