---
name: Social proof
slug: social-proof
category: pattern
status: published
created: 2026-08-21T00:00:00.000Z
modified: 2026-08-21T00:00:00.000Z
definition: Showing what other people did or said next to a decision, as counts,
  reviews, logos, or live activity, so the reader can borrow their judgement.
aliases:
  - name: testimonials
    source: ui-patterns
  - name: review count
    source: community
  - name: used by logos
    source: community
  - name: activity notification
    source: community
  - name: N people are viewing this
    source: community
tags:
  - commerce
  - perception
relations:
  contrastWith:
    - logo-cloud
    - fake-social-proof
    - leaderboard
  variantOf: []
  partOf: []
  seeAlso: []
implementations: []
sources:
  - title: "UI Patterns: Social Proof"
    url: https://ui-patterns.com/patterns
  - title: "GoodUI: Try Social Proof instead of talking about yourself"
    url: https://goodui.org/
demo: inline
exhibit: false
useWhen: the interface shows what others chose to steady your choice
---

Social proof is the interface saying "other people already decided this". Robert
Cialdini named the principle in Influence (1984): when we are unsure, we treat the
behaviour of others as evidence about what is correct, and the less we know the
heavier that evidence weighs. Interfaces spend it in a handful of shapes. A star
[rating](/rating) with the number of reviews behind it. A usage count. A row of
customer logos. An [avatar group](/avatar-group) of people already in the room. A
quoted testimonial. A live activity line saying how many others are looking at the
same thing right now.

The difference between proof that works and decoration that does not is specificity.
"12,400 teams" is checkable and "loved by thousands" is not, so the first reads as
evidence and the second as marketing copy the reader has learned to skip. A quote
carrying a name, a role, and a company is worth several anonymous ones. A rating means
almost nothing without the count underneath it, since five stars from three people is
weaker than four and a half from twelve hundred. Retail analyses of review data have
repeatedly found that perfect scores convert worse than very good ones, for the same
reason: a page with no criticism on it reads as a page that removed the criticism.

Relevance beats volume. Proof from people the reader recognises as being like them
(the same industry, the same size of team, the same problem) does more than a bigger
number from strangers, which is why logo walls are usually segmented and why review
excerpts are worth choosing rather than sampling. Placement matters just as much: the
work happens beside the decision, on the pricing card, the plan chooser, the checkout
button, and not on an About page nobody visits on the way to buying.

This is also a pattern with a well populated dark twin. Invented testimonials, stock
photography standing in for customers, seller-planted reviews, and counters showing
activity that never happened are all [dark patterns](/dark-pattern) and, in the United
States, are directly targeted by the Federal Trade Commission's 2024 rule against fake
reviews and testimonials, which carries civil penalties. The honest version of a live
activity line is easy to state: if the number is not a real count of real people, do
not show a number.