Trustpilot rejects categorically false allegations by Grizzly Research

The Board of Trustpilot rejects categorically the allegations made in the report published by Grizzly Research. Their report contains factual inaccuracies and false claims, which were intended to adversely impact the company's share price. Trustpilot was not contacted for comment.

The report is built on a basic misunderstanding of Trustpilot's business model. It ignores publicly available information about our scale, policies, data, and enforcement, and makes a series of demonstrably false accusations, including:

●     It is false to claim that Trustpilot creates unsolicited review profiles to sell subscriptions. Over 97% of businesses on the platform do not pay for a subscription. Consumers and businesses create profiles for free. 70% of reviews on free profiles are 5 stars.

●     It is false to claim that fake reviews are managed differently for paying vs free businesses. All businesses - whether free or paid - must comply with the same Guidelines and are held to the same moderation principles.

●     It is false to claim that we have traded the integrity of reviews for revenue. We have c.200 people focused on ensuring trust and integrity across our platform and have removed 6.7m reviews from the platform so far this year. We have successfully brought lawsuits against bad actor businesses and review sellers.

We are considering all appropriate options in response to demonstrably false statements. Below we set out the facts about our business model and content moderation.

Trustpilot exists to help consumers make the right choice and businesses to build trust, grow and improve

Trustpilot operates a freemium business model where businesses have the choice whether to use Trustpilot for free or to subscribe. Businesses which subscribe are buying access to brand rights and tools that automate review collection and provide insights into customer feedback. There are 1.3 million businesses that have claimed their profile on the platform of which 3.1% pay for additional features. These features and our pricing plans can be found here.

The platform is global, open and hosts over 350 million reviews. This means consumers have the freedom to share their genuine experiences as and when they choose, and businesses can invite consumers to leave feedback, and respond at any time. Information about how Trustpilot works is in the public domain, and was not referred to in their report.

Any business has the right to claim its own Trustpilot profile for free. This enables them to receive and reply to all reviews. We transparently disclose data on all businesses reviews and whether they are a paying customer through the "How this company uses Trustpilot" link on every profile page.

Trustpilot does not create unsolicited review profiles in order to sell subscriptions. Profiles are created by consumers when they search for or want to share feedback about a business, or can be created by businesses who want to collect feedback.

Our guidelines and fraud detection technologies apply equally to paying and non paying businesses. The TrustScore calculation is identical for all customers. It is active engagement by businesses inviting and responding to reviews, rather than whether they are a paying customer, which impacts the TrustScore. The average TrustScore for a paying customer is 4.37 and for a free business that is actively engaging with the platform is 4.26.

Trustpilot is vigilant about moderation and is constantly developing new detection models

Trustpilot recognises that all review platforms operating at our scale attract bad actors. We take content moderation extremely seriously and are constantly evolving our processes to address bad actors. We have a team of c.200 people focused on ensuring trust and integrity across our platform. Their report suggests that we have at most 6 people, which is clearly inaccurate. Over the last 2 years, we've increased our investment in this process by almost 50%, putting millions of dollars into our people and technology, and we will continue to do so.

Our commercial team is entirely separate from our Trust function and is not able to influence the outcome of content moderation. All teams at Trustpilot, including our commercial team, undertake regular training around the importance of acting with integrity, and in accordance with our Trust principles.  

All businesses - whether free or paid - must comply with the same Guidelines. The same moderation standards apply and responding to and reporting reviews is free for all businesses.

Our Guidelines for Reviewers are explicit: users are welcome to use profile pictures but they are  expressly prohibited from impersonating somebody else. We provide the option for reviewers to securely verify their ID to help build trust and credibility with our community. We have a whistleblower form available to all users to report violating imagery, and when we have evidence of this we will act as set out in our Action We Take policy . We are constantly considering new technologies to detect this.

How the process works in practice

When a review is submitted to the platform it is analysed by our automated detection models. These don't just look at the substance of the review, but at hundreds of metadata points that are effectively a digital fingerprint of the reviewer. We are constantly iterating our models to detect new and evolving patterns.

As we set out in our 2025 Trust Report, in the last year we introduced automated guideline detection software. This technology has enabled us to more accurately detect and remove content which violates our guidelines. Since its phased introduction in August 2024 this software has been identifying and removing content which contains nonsense. When we introduce new models we prioritise the most recent reviews first, and then we backfilter over time.

We catch 90% of reviews we remove before they reach the platform. Technology is not perfect and continues to evolve, but we are dedicated to constantly improving. Any business or consumer can flag reviews that breach our guidelines and they can also submit a whistleblower report. This is then investigated by our Trust team.

Where we discover breaches of our guidelines, we take action in line with our Action We Take policy. We are transparent about the action we take and where necessary and appropriate we will apply alerts and warnings to business profiles.  Each year we remove millions of reviews and restrict large numbers of accounts that breach our guidelines and we have continued to scale up enforcement this year. Given the scale of the platform, even with our best efforts, things will get through which is why we enable people to flag reviews they believe are suspicious.

So far in 2025 we have:

●     Removed 6.7m reviews (+68%% YoY)

●     Issued 23,932 warnings (+35%% YoY)

●     Issued 6,128 formal notices (+22% YoY)

●     Removed 6,500 businesses (+60% YoY)

●     Implemented 9,694 consumer warnings (+18%)

●     Implemented 15,262 consumer alerts (+55% YoY)

In 2025 to date reviews removed for paying and free business profiles were as follows:

●     Free business profiles (by star rating): 14 per cent were negative, 85 per cent were positive and 1 per cent were neutral.

●     Paid business profiles (by star rating): 14 per cent were negative, 84 per cent were positive and 2 per cent were neutral.

We are an open platform and recognise that consumers may wish to leave feedback along different aspects of their journey. Recognising this, our review system is designed so that only a reviewer's most recent review for a business counts towards the TrustScore. The report falsely claims that multiple reviews from the same user contribute to the score.

Trustpilot has a strong record of improving regulation and protection for consumers

Review manipulation of any kind is illegal. There has been significant new legislation enacted to support this in the last 18 months, both in the US under the FTC's new rule and in the UK with the introduction of the DMCCA - Trustpilot has played an important role in shaping these new laws.

Trustpilot expressly bans review gating, also known as cherry-picking. Our guidelines prohibit businesses from only inviting happy customers, from incentivising reviews, as well as from using any kind of fake reviews. In line with our Action We Take policy so far this year we have taken action against over 330 paying businesses, terminating the paid subscriptions of 39, for review gating or fake reviews.

New information will often surface about a business after reviews have been written. We continually assess and analyse profiles and work with regulators, law enforcement and industry bodies to drive industry standards and collaborate in our actions against bad actors. The ways in which users and bad actors engage with platforms like Trustpilot is constantly evolving. We have a number of product development plans in the pipeline for 2026, including initiatives to reduce the risk of scam businesses.

The existence of black markets in fake reviews is not unique to Trustpilot and is one we are constantly fighting to reduce. Buying or selling fake reviews is illegal. When we identify that reviews have been purchased, we remove the reviews, sanction the business and, where they are a paying customer, terminate the relationship. We also report sellers and buyers to regulators or enforcement bodies. We have successfully brought lawsuits against bad actor businesses and review sellers and engage with the platforms the review sellers are advertising through. These efforts helped to highlight and focus the need on appropriate legislation.

We categorically refute the allegations in their report and have highlighted here some of the inaccuracies we have found.

Enquiries

Notes

Trustpilot exists to help consumers make the right choices, and to help businesses build trust, grow and improve. To protect the integrity of our platform there are clear rules and guidelines for all users to follow. These are set out in our Terms of Use for Businesses and Consumers  and Guidelines for Businesses and Reviewers. Our Guidelines and Terms are reinforced by our Action We Take policy.

All data provided in this report is for the period 1 January - 30 November 2025.