Fake Reviews: Scale, Evidence and Implications
Introduction
Online reviews now sit at the centre of consumer decision-making.
They influence purchasing behaviour, shape brand perception, and increasingly feed into search and AI-driven recommendation systems.
Despite this, there has historically been limited independent, cross-platform analysis of how reviews behave at scale.
TruthEngine® was established to address that gap.
Over the past five years, a team of PhD-level data scientists has analysed millions of publicly available reviews across multiple sectors and major platforms, using a combination of statistical, behavioural and linguistic techniques.
This page summarises the key findings from that work.
Core Finding
Across the datasets analysed to date, 50.15% of reviews exhibit suspicious or non-organic characteristics.
This figure is derived from a five-year, multi-sector analysis conducted by TruthEngine®’s research team, covering large volumes of publicly available reviews across leading platforms.
Observed ranges at sector level typically fall between 40% and 55%.
Interpretation of “Suspicious”
The term suspicious is used deliberately.
It does not mean that a review is definitively fake.
It indicates that a review, or a pattern of reviews, displays characteristics that are:
- statistically atypical relative to baseline behaviour
- inconsistent with historical patterns
- or aligned with behaviours commonly associated with review manipulation
Examples include:
- abnormal concentrations of reviews within narrow time windows
- repeated or highly similar linguistic structures
- reviewer accounts exhibiting suspicious behaviour
- atypical shifts in rating distributions
- clustering of reviews around specific locations, individuals or events
No single signal is determinative on its own.
Assessments are based on the presence and interaction of multiple signals at scale.
Scope of the Issue
Public discussion of fake reviews has often focused on clearly fabricated content.
The evidence indicates the issue is far broader.
Review environments can become misleading through a range of mechanisms, including:
- undisclosed incentivisation
- selective review solicitation
- suppression or filtering of negative feedback
- contributions from connected parties (e.g. staff or associates)
- coordinated or structured review activity
These behaviours can materially affect how a business is represented, even where individual reviews are written by real people.
Methodological Approach
The analysis underlying these findings is based on a five-year programme of research conducted by
a multidisciplinary team of PhD-level data scientists, combining large-scale data engineering with statistical, behavioural and linguistic analysis.
This includes:
- aggregation of publicly available review data across major platforms
- cross-platform comparison
- time-series analysis of review volumes and rating distributions
- behavioural analysis of reviewer activity patterns
- coordinated or structured review activity
- clustering and anomaly detection techniques
Assessments are made at a pattern and distributional level rather than through binary classification of individual reviews.
This reflects the fact that manipulation is typically observable in aggregate behaviour rather than isolated instances.
Regulatory Context
The regulatory environment has evolved significantly.
Under the UK’s Digital Markets, Competition and Consumers Act:
- fake and misleading review practices are explicitly prohibited
- responsibility sits with the business being reviewed as well as the platform hosting the reviews
- businesses are expected to take reasonable and proportionate steps to ensure authenticity
This responsibility is non-delegable.
Reliance on third-party platforms does not remove accountability.
In addition, historic reviews may create current exposure where they continue to influence consumer decisions.
Implications
For Consumers
Consumers may be relying on review information that is not fully representative of genuine customer experience.
For Businesses
Online reviews now carry implications beyond marketing, including:
- regulatory compliance
- reputational risk
- investor and due diligence considerations
- exposure to enforcement action
For the Market
The normalisation of manipulated or distorted review behaviour creates structural challenges for fair competition and consumer trust.
Limitations
It is important to note:
- the presence of suspicious characteristics does not establish intent or wrongdoing
- not all suspicious patterns result from deliberate manipulation
- findings should be interpreted as indicators requiring further review, not definitive conclusions
Conclusion
The evidence indicates that the scale of potentially misleading review content is materially higher than commonly assumed.
The issue is not limited to isolated fake reviews, but reflects broader structural dynamics in how reviews are generated, collected and presented.
As regulatory expectations increase and consumer awareness evolves, the ability to understand and evidence review authenticity is likely to become a standard requirement.
Regulatory Position: International Overview
Regulation of fake and misleading reviews is not limited to the UK.
Across multiple jurisdictions, laws already prohibit the creation, solicitation or presentation of reviews in ways that mislead consumers.
| Jurisdiction | Regulatory Framework | Position on Fake / Misleading Reviews |
|---|---|---|
| United Kingdom | Digital Markets, Competition and Consumers Act (DMCC) | Explicit prohibition of fake and misleading reviews. Responsibility sits with the business benefiting from the reviews. |
| European Union | Omnibus Directive | Requires businesses to take reasonable steps to ensure reviews are genuine and prohibits misleading practices. |
| United States | FTC Endorsement Guides + 2024 Final Rule | Prohibits fake reviews, undisclosed incentives, and reviews from employees or related parties. |
| Australia | Australian Consumer Law (ACCC) | Treats fake or misleading reviews as deceptive conduct. |
| Canada | Competition Act | Prohibits false or misleading representations, including fabricated reviews. |
| Germany | UWG (Unfair Competition Law) | Requires transparency and prohibits misleading consumer review practices. |
| France | Consumer Code | Requires verification of review authenticity and prohibits misleading practices. |
| Italy | Consumer Protection Code | Active enforcement against misleading review practices. |
| Netherlands | ACM Guidance | Requires authenticity checks and transparency in review publication. |
| Spain | Consumer Protection Law | Prohibits misleading commercial practices, including fake reviews. |
Across jurisdictions, the legislation is consistent: reviews must not mislead, and businesses are expected to take active steps to ensure their authenticity.
About TruthEngine®
TruthEngine® is an independent, cross-platform review analysis system.
It does not host reviews or monetise review traffic.
Its role is to provide objective, evidence-led insight into how review environments behave at scale.