Your customers don't remember their 14-day trial as 14 days of experience. They remember two moments: the peak (most intense moment) and the end (final moments). Everything in between? Largely forgotten.
This is the peak-end rule — one of the most robust findings in judgment and decision-making research.
The Meta-Analysis That Confirms It
Alaybek and colleagues (2022) published the definitive analysis: 174 effect sizes across multiple domains. The peak-end effect clocked in at r = 0.581 — a large effect by any standard. Both peak intensity and end-of-experience moments predicted retrospective evaluations far better than duration, beginning, or trough moments.
Duration neglect — the finding that how long an experience lasts barely matters to how it's remembered — was confirmed across the board.
The Colonoscopy Studies That Prove It
Kahneman's team ran a remarkable series of studies. In the most striking one — a randomized controlled trial with real colonoscopy patients (Redelmeier, Katz & Kahneman, 2003) — half the patients had their procedure deliberately prolonged. The scope was left stationary at the end, creating a few extra minutes of mild (not painful) sensation.
The result? Patients who had the LONGER procedure rated it as less unpleasant. They were also more likely to return for follow-up procedures.
Read that again. More time. Less pain remembered. Because the ending was gentler.
An earlier study confirmed the same pattern: retrospective pain correlated strongly with peak intensity and the final 3 minutes (both P<0.005), but not with the total duration of the procedure.
The Cold Water Experiment
In 1993, Kahneman's lab demonstrated this with a simpler setup. Participants immersed their hands in painfully cold water for two trials:
Short trial: 60 seconds at 14°C Long trial: 60 seconds at 14°C, then 30 extra seconds where water warmed slightly to 15°C
Participants chose to repeat the longer trial. More total pain, but the ending was slightly better — and that's what they remembered.
What This Means for Product Trials
If you're running a 14-day trial, you're not running 14 equal days. You're running two critical moments with twelve days of background noise.
Engineer the peak early (Days 4-7). This is when the user should experience the product's full power. The "aha moment" that makes them think "this actually works." Progressive onboarding — guided steps to that peak — produces 3x higher conversion than passive trials.
Engineer the ending deliberately (Days 13-14). The final experience before the payment decision will disproportionately shape how the entire trial is remembered. A personalized summary of what they built. A surprise bonus. A "graduation" message. Anything except a cold "your trial is expiring" wall.
Don't obsess over the middle. Duration neglect means Days 8-12 are maintenance, not make-or-break. Keep things running smoothly, but save your dramatic moments.
The Counterintuitive Lesson
The colonoscopy RCT proves something that feels wrong: a longer experience with a better ending is preferred over a shorter experience with an abrupt ending. Applied to trials: slightly extending the trial period while designing a genuinely positive final experience may convert better than a shorter trial that ends with a paywall.
The trial isn't 14 days. It's two moments. Design both of them.
References
Alaybek, B., et al. (2022). The peak-end effect. Organizational Behavior and Human Decision Processes. 174 effect sizes, r = 0.581.
Kahneman, D., Fredrickson, B. L., Schreiber, C. A., & Redelmeier, D. A. (1993). When more pain is preferred to less. Psychological Science.
Redelmeier, D. A., & Kahneman, D. (1996). Patients' memories of painful medical treatments. Pain.
Redelmeier, D. A., Katz, J., & Kahneman, D. (2003). Memories of colonoscopy: A randomized trial. Pain.