Peak-End Rule

Overview

People remember experiences not by averaging all moments but by sampling two: the peak (highest emotional intensity) and the end. Everything in between — including duration — is largely discarded. This is the peak-end rule, from Kahneman et al. (1993) and Redelmeier & Kahneman (1996).

Global evaluation ≈ (peak intensity + end intensity) / 2. Experience design is not an averaging problem — it is a peak-and-ending problem.

Neighbor skills: Use after aarrr-pirate-metrics to place the peak in the right lifecycle stage; use anchoring to set expectation baselines peaks must exceed; pair with nudge-theory to smooth the path to the peak and ending; use probabilistic-thinking before designing peaks to estimate expected effect size.

When to Use

Apply when:

• Designing or auditing a multi-stage experience with a clear start and end (onboarding, service encounter, event, medical visit)
• NPS or satisfaction scores are lower than expected given average perceived quality
• A competitor with similar objective quality consistently earns higher recommendation rates
• Allocating limited resources across experience stages and need to know where to concentrate

When NOT to use: purely instantaneous interactions with no duration; real-time performance optimization (not retrospective rating); welfare/health assessments where experienced utility — not memory — is the correct measure; one-time required events with no competitive alternative.