Trustpilot Case Study

“Pasabi is achieving an extremely high degree of verified accuracy in identifying bad actors on the platform”

 Anoop Joshi
Senior Director, Legal

Trustpilot Case Study

“Pasabi is achieving an extremely high degree of verified accuracy in identifying bad actors on the platform”

 Anoop Joshi
Director, Legal

Pasabi supports Trustpilot tackle fake reviews at scale

89% of consumers globally check reviews online before making purchases

Source: Canvas8 2020

Trustpilot is the one of the world’s most popular websites.

They bring consumers and companies together on their open platform to continuously share, collaborate, and improve. The power of the platform comes from being transparent to all consumers and companies and most importantly, independent of both.

Trustpilot is dedicated to being the most trusted review platform in the world. With an open platform where everyone has an equal voice, Trustpilot is committed to fighting fake reviews and platform misuse.

The Challenge

As one of the world’s leading review platform with over 3m posted to the platform every month, Trustpilot already has the technology to detect, investigate and take action against suspected fake reviews.

However, removing fake reviews is just one part of the challenge. The goal was to identify and remove misbehaving companies at scale providing consumers with a cleaner, more authentic experience.

Behind fake posts are companies and individuals that are trying to ‘game the system’ through:

  • Buying fake reviews from review sellers
  • Incentivising consumers to leave only positive reviews
  • Undermining the competition by posting negative reviews

Trustpilot needed to find a way to provide additional insight into misbehaving companies, strengthen existing tooling and capabilities and gather the additional evidence to allow the Content Integrity team in Trustpilot to take action.

With such a high volume of monthly posts, analysing the data manually was not an option. The ability to tackle the problem at scale was essential.

The Solution

Pasabi AI Technology Platform supports the detection of fake reviews, positive and negative bias and the identification of misbehaving companies.

Specifically our graph technology is able to draw links between individuals and groups of individuals by analysing hundreds of data points. The Pasabi platform filters out the unconnected posts and unrelated companies, leaving a smaller data set for more detailed analysis.

Network Analysis

The worst culprits tend to work together. Our cluster detection technology and bespoke algorithms spot patterns in the data to uncover the actions of review sellers and other groups working together. Our approach identifies where the enforcement teams need to direct their efforts to find the most prolific offenders.

Positive & Negative Bias

The Pasabi platform identifies companies attempting to ‘game the system’ or companies who are under attack from clusters of negative reviewers.

Not real data – for illustration purposes only

Making an Impact

Working closely with the Legal, Fraud & Investigations and Content Integrity teams at Trustpilot, we have been able to provide unique insights into their data, allowing them to gather the evidence they need to pursue taking the necessary action against misbehaving companies and individuals reviewers at scale.

The Results

  • Millions of reviews and profiles analysed monthly

  • Extremely high degree of verified accuracy in identifying bad actors

  • Thousands of enforcement actions taken

“Pasabi has been able to turn huge volumes of data into something that’s really useful from a legal team perspective.

Their user interface is clear and simple to use and is something we’re using to flag suspicious activity and take action at scale. This should also have a deterrent effect, if people think they’re going to get caught.

Being able to identify the scale of the issue through this use of technology has allowed the legal team to be even more proactive in mitigating future risk and in discussions with regulators.

We have been able to help them to identify where some of the challenge is coming from, as more become concerned with tackling fake reviews online.”

Anoop Joshi
Senior Director, Legal

Listen to Anoop discuss Trust and Reviews on the Trust Series Podcast

Looking to understand the scale of fabricated & biased reviews and companies trying to ‘game the system’ ?

Request an Online Reviews Insight Report and we’ll show you the true scale of your challenge and how to address it.

Request Online Reviews Report