Cover art for Loss Aversion Is Smaller Than Your Pricing Page Assumes

Published: November 20th

Loss Aversion Is Smaller Than Your Pricing Page Assumes

The 2x multiplier is a borrowed number, not a finding about your users

Loss aversion = the idea that losing something hurts about twice as much as gaining the same thing feels good.

The problem: that multiplier does a lot of work on pricing pages, and it is shakier than the copy assumes.

Scarcity banners, countdown timers, "you will lose your streak". Most of it rests on a number from one 1992 paper. The argument about that number got sharper in the last two years.

You are not citing a law of the mind. You are citing an average of other people's experiments, run at other people's stakes.

Three designers at a studio wall covered in printed pricing pages and sticky notes, laughing as one of them points at a circled line while holding a stapled research paper
The best version of this meeting is the one where someone asks where the number came from.

1. The number is an average of the literature, not of your users

What teams ship: Copy written as if every loss weighs double.

Why it fails: The biggest review pooled 607 estimates from 150 articles published between 1992 and 2017. It put the mean at 1.955, with a 95% interval of 1.820 to 2.102. That is a summary of published studies, not a measurement of your users.

Ship instead: Treat 2x as a guess about your product, then test it.

2. A second review, the same year, got 1.31

What teams ship: One citation, repeated as settled.

Why it fails: A 2024 review of risky choice fitted the model to individual decisions and reported 1.31, with a 95% interval of 1.10 to 1.53. They also found that few studies had ever fitted it properly, and that much of the data was poor.

Ship instead: When two careful teams read the same literature and get 1.96 and 1.31, stop quoting a single figure in a design review.

3. Take out two study habits and the effect nearly goes

What teams ship: Loss framing everywhere, on the belief the effect is sturdy.

Why it fails: A 2025 re-analysis took the big review's own dataset, 84 papers and 163 estimates covering 149,218 people, and split it. Where the losses on offer were smaller than the gains, or where options were shown in order of size, strong loss aversion appeared. Where gains and losses were matched and unordered, the figure was about 1.07, and not significantly above 1.0.

Ship instead: Treat the effect as something your context creates, not something people carry around.

4. Small stakes are the weakest case, and a plan is a small stake

What teams ship: Loss framing on a nine dollar tier.

Why it fails: A 2025 replication across three studies found the figure rose with the size of the stake. Values for the 50, 100 and 150 euro gambles were all higher than for the 25 euro one.

Ship instead: Save loss framing for what is genuinely costly to lose. Accounts, history, finished work.

5. Half the old evidence is really about inaction

What teams ship: Streaks, expiry warnings, "do not lose your progress".

Why it fails: Gal and Rucker showed that the classic studies mix up losing with doing nothing. In one of their tests, 92% picked the status quo option even when the two options were equivalent and neither was a loss. Their reading: there is no general rule that losses weigh more.

Ship instead: If what you are really using is inertia, say so and design for it. A good default beats a fake deadline, and it keeps working after the user notices.

The claim on the slideWhat the evidence says now
Losses hurt twice as muchThe mean of published estimates is 1.955, with a wide interval
It is one settled numberA second 2024 review of risky choice reports 1.31
The effect is sturdyWith matched, unordered options it falls to about 1.07
Size does not matterThe figure grows with the stake. Small stakes are the weakest case
The endowment effect proves itLoss is tangled with inaction. 92% stayed put with equivalent options

An audit for your next pricing review. Find every line that assumes a loss weighs double. Ask whose number that is. Ask what the stake is in money, and whether it is small. Ask whether the real mechanism is inertia rather than fear. Then run the test you were going to skip, because the research never settled it.

Get the PDF