---
name: Fake social proof
slug: fake-social-proof
category: pattern
status: published
created: 2026-08-21T00:00:00.000Z
modified: 2026-08-21T00:00:00.000Z
definition: Manufacturing evidence that other people approve, through invented
  reviews, testimonials, or live activity messages nobody actually generated.
aliases:
  - name: fake reviews
    source: community
  - name: activity notification spam
    source: community
  - name: fabricated testimonials
    source: community
tags:
  - commerce
relations:
  contrastWith:
    - social-proof
  variantOf:
    - dark-pattern
  partOf: []
  seeAlso: []
implementations: []
sources:
  - title: "Deceptive Design: Fake social proof"
    url: https://www.deceptive.design/types
demo: inline
exhibit: false
useWhen: someone in Leeds just bought this, and nobody did
---

[Social proof](/social-proof) works because judgement is expensive and other people are
evidence. Four hundred reviews, a queue outside a restaurant, and a colleague's
recommendation all let a decision be made without doing the research, and interfaces have
borrowed the effect honestly for decades. Fake social proof keeps the shape and removes the
people. The reviews were written to order, the testimonial photograph came from a stock
library, and the notification saying someone in Leeds just bought this was generated by a
script from a list of names and towns.

The manufactured kinds have tells, and they are worth learning as a reader as well as a
designer. Activity popups cycle: watch one for two minutes and the same six towns come round
again, always a plausible interval apart, never at three in the morning, and never
mentioning a return. Invented review counts do not match the reviews on the page, because
the number is a decoration written by marketing and the list is a query. Fabricated
testimonials share a voice, since one person wrote them all, and they praise attributes
rather than describing events. Ratings that are bought cluster at four point eight, high
enough to sell and low enough to look real, and the distribution behind them has no long
tail of ones.

It sits in a family that fakes one input each, which is the fastest way to keep them apart.
[Fake urgency](/fake-urgency) fakes time, with a countdown that resets when the page reloads.
[Fake scarcity](/fake-scarcity) fakes supply, with a stock bar that is a random number.
[Sneaking an item into the basket](/sneak-into-basket) fakes a choice the buyer never made.
This one fakes other people. All are cases of a [dark pattern](/dark-pattern), and this one
is the one with the widest blast radius, because it degrades every honest signal beside it: a
reader who has learned that "12 people are viewing this" is fiction stops believing the
review count next to it, which was real.

Enforcement has arrived faster here than for most of the family. The US Federal Trade
Commission's 2024 rule on consumer reviews bans buying, writing, or misrepresenting them,
with penalties per violation, and the UK's Digital Markets, Competition and Consumers Act
puts fake reviews and undisclosed incentives on the same footing. That changes the design
brief in a useful way: the question is no longer whether an activity widget converts, it is
whether every claim on the page could be defended with a record. Honest versions of each
element exist and are usually stronger. Say how many people bought this month rather than
naming a stranger in Leeds; show the rating distribution rather than the average alone; and
publish the negative reviews, which are the single most credible thing a product page can
carry.