Last week Amazon.com introduced a new proprietary system to regulate the ranking and placement of its product reviews. "The system will learn what reviews are most helpful to customers...and it improves over time," Amazon spokeswoman Julie Law said in an interview with CNET. Instead of relying on static algorithms, the new machine-learning platform will make data-driven decisions based on changing criteria to adjust the placement of reviews. The system is intended to promote newer reviews as well as reviews from verified Amazon purchasers and those that more customers identify as being helpful.

Last April, Amazon sued a number of online businesses for allegedly selling fake customer reviews of their products. The practice is known as astroturfing, masking the source of an online endorsement to make it appear as though it originated from grassroots consumer participation. A simple search of the website Fiverr, which allows users to offer services for five dollars, displays several offers of individuals willing to write online reviews and product testimonials for the purchaser. While it was not explicitly stated in the rollout of the new system, the emphasis on reviews by verified purchases indicates that Amazon is attempting to combat the practice of commercial astroturfing. 

The website Yelp, which publishes crowd-sourced business reviews, has already instituted a customer alert system to combat the practice. The system warns users that a business has been caught attempting to buy fake positive reviews from known sources and displays the warning on the offending organization’s page. However, the warnings only remain on the site for 90 days, allowing businesses to potentially circumvent Yelp’s attempts at enforcing transparency in their future attempts at purchasing positive reviews. 

The news of Amazon’s new customer review platform comes at the same time that the UK-based Competition and Markets Authority found that over half of UK adults use online reviews before making a purchase. The organization subsequently opened an investigation into fake online reviews in the UK. 

Last year, a Harvard study estimated that 15 to 30% of online reviews are fake, whether in the form of unacknowledged endorsements or attempts by organizations to discredit their competition. 

Also see: Our Growing Confidence in Online Shopping; Should We Be Concerned?