Fabricated reviews: how fake review detection software is the answer

Written by Chris Downie

03 July 2020

Review platforms and eCommerce sites are fighting against an increasing tide of spam, scam, fabricated and biased reviews. 

Research by Bazaarvoice has highlighted the negative impact fake reviews can have on trust in brands and consumer purchase behaviour. The survey of 10,000 consumers suggested that 54% of consumers wouldn’t buy a product if they suspect it has fake reviews. 

Faced with the task of identifying and removing fabricated, biased and bought reviews, many brands have struggled to make real headway in tackling ‘bad actors’.  The scale and complexity of the problem is one that is proving difficult to solve at the human level.

Recent developments in machine learning technology applied to the challenge of fake content and reviews can help brands to understand the true scale of the challenge, identify patterns, surface fake posts and gather the data needed to start tackling the problem in a meaningful way, at scale.

At Pasabi we have developed fake review detection software (link: Pasabi software page) now being used by some of the world’s leading online review platforms, including Trustpilot. 

Let’s take a look at the key features of the software and the business outcomes that can be achieved.

Spam & Scam

From hacking services to insurance scams to pyramid schemes, unscrupulous individuals and groups are peddle their wares to an unsuspecting audience. The very presence of spammers and scammers on your website hurts your brand and challenges your audience’s trust in your service.

Key Business Outcomes 

  • Remove Personal Identifiable Information
  • Block persistent offenders
  • Improve the user experience 

Fabricated Reviews

fabricatedreviews_UI

Identify both suspects and individual suspicious reviews. Our analysis includes patterns of language and behaviour, building up a graph database, scoring suspicious behaviour, and bringing the most suspect activity to the top of the list.

Key Business Outcomes

  • Identify fake posts at scale
  • Remove fake posts, providing an authentic user experience
  • Identify groups acting together

Time Series Analysis

timeseriesanalysis

Understand the progress you’re making in reducing the number of fabricated reviews by tracking legitimate reviews vs suspicious reviews over time. For individual companies identify spikes and irregular behaviour for further investigation. 

Key Business Outcomes

  • Easily demonstrate to your business, progress on reducing fake posts
  • Identify companies under attack, or those looking to ‘game the system’.

Let’s tackle this problem together and give consumers the authentic user experience they deserve.

 

New call-to-action