cognitive psychology
Peak-end rule
We don't judge an experience by averaging every moment of it. We remember it mostly by two points - how it felt at its most intense (the peak) and how it felt at the end - and we quietly discount how long it lasted.
The demo
Two everyday waits. Click through each one moment by moment - you'll feel how each lands - then I'll ask which you'd rather go through again.
Which would you rather go through again?
What this demo shows (text version)
You play through two waits. The first has five moments: it starts calmly, dips to a sharp low when a till closes, recovers, and finishes on a high when a friendly cashier hands you a voucher. The second has only three: it is shorter and moves quickly, but ends badly when the card machine fails and you're told to start over.
Asked which you'd repeat, most people pick the first - even though it was longer and had a worse single moment - because it ended well. The second was objectively less waiting, yet its sour final moment is what sticks. That's the peak-end rule: an experience is remembered by its most intense point and its ending, while its length is largely forgotten, which is why the last moment is worth designing with care.
One wait was longer than the other, yet you'd rather repeat it - because it ended well, while the short one ended sour. You didn't tally the minutes; you remembered the peak and the finish. That's the peak-end rule, with the duration and all the dull bits neglected.
It's why the ending deserves more attention than its length suggests: a checkout that drags but closes with a warm, clear confirmation beats a fast one that dumps you on a blank screen. The final moment gets remembered out of all proportion to how long it took.
My rule: find the worst moment you can't remove and the last moment you can always control, and spend your effort there. Smoothing the dull middle barely shifts the memory; a better peak and a better ending shift it a lot.