Embarrassment and low-stakes failure
Fear of looking foolish stops people trying, even when nobody is watching.
Embarrassment arises from a perceived loss of face in front of a real or imagined audience. It does not require the failure to be consequential, only for it to be visible or recorded. People avoid situations where a mistake could be witnessed, which suppresses exploration and help-seeking.
How it shows up in software
Shared and collaborative products make experimentation publicly visible, so people stop experimenting. Sending a message to the wrong channel, publishing a half-finished page or asking a beginner question in a public forum all carry face risk. In single-player software the imagined audience still operates: users avoid features they think they should already understand, and they do not read help because needing it feels like an admission.
Using it well
- Give people a private place to fail: drafts, sandboxes, preview modes, a personal space before a shared one.
- Make reversal fast and quiet. Editing a message without a prominent edited badge lowers the cost of getting it wrong.
- Write help content as though the reader is capable and the product is unusual, not as though the reader is behind.
- Do not broadcast beginner status. Badges, onboarding labels and empty-profile placeholders visible to others suppress participation.
Where it turns manipulative
- Using public visibility as pressure, such as showing colleagues who has not completed a setup step.
- Shaming copy on failure, including cute messages that imply the user did something silly.
- Publicly ranked activity in a work tool, which converts ordinary variation in workload into visible standing.
Where you have seen it
Figma
Draft files live in a personal space and become visible to a team only when explicitly moved or shared.
Slack
Message drafts persist per channel and editing is available after sending, so a mistaken post can be corrected in place.
Duolingo
Lessons are practised alone, and errors are corrected within the lesson before anything is reflected in a visible streak or league position.
What the research says
- Modigliani, 1971Mixed evidence
Manipulated public failure in a group task. Embarrassment tracked the perceived loss of esteem in others' eyes rather than the objective failure itself.
Small laboratory sample from the early 1970s. The audience-dependence has held up; the specific measures are dated.
- Dahl, Manchanda and Argo, 2001Mixed evidence
Studied embarrassment in a consumer purchase context and found the presence of others, even unfamiliar and passive, raised reported embarrassment and changed behaviour.
- Miller, 1996Mixed evidence
A book-length synthesis of embarrassment research, covering its triggers, its social function and the strategies people use to avoid it.
A synthesis, so treat it as an organised account of the literature rather than as independent evidence.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.