Amazon has continued its fight against fake reviews, as the online retail giant last week filed a lawsuit in Washington state court against more than 1,000 people who allegedly offered to hire themselves out as fake reviewers at $5 a review.
The suit casts light on what appears to be a growing practice of paying for fake reviews that masquerade as testimonials from ordinary people. The answer to these problems go beyond lawsuits; for the company to counteract the business of fake reviews, Amazon (AMZN) and other online retailers need to “get out the vote” among consumers to generate more genuine buzz in the marketplace.
While weeding out suspected fake reviews and putting a stop to those posting favorably for a fee is a good start, these steps don’t go far enough. In order to harness the participatory potential of social media, retailers need to encourage more authentic consumer conversations.
Bogus reviews are not only inherently untruthful, but they undermine the credibility of reviews in general. An avalanche of paid hype can bury comments from unbiased shoppers who tried a product and freely offer a star-ranking (one through five) and a review in their own words.
When marketplace reviews are manipulated, it exacerbates another problem: reviews tend to be written by a non-representative group of consumers. There is ample research that shows that reviewers often have more extreme preferences or extreme behaviors. Fake five-star and one-star reviews only make this problem worse. With its vast number of product lines—from books to clothing to vitamins—Amazon is unparalleled for reaching consumers. Reviews that appear on the Amazon site also wield influence on future consumer purchasing decisions – both whether to purchase and which items are purchased. Due to a lack of broad participation, many online items have very few customer reviews. For these items, a single five-star review can have a tremendous impact on consumer decisions.
Such influence, however, has apparently raised the stakes for product creators (e.g., book authors) to try to increase their average ratings to attract more attention, achieve a “best seller” designation, and encourage more purchases. Those trying to game the system may “recruit” reviews from family and friends or even pay for favorable comments and ratings.
There’s another concern regarding the influence and authenticity of reviews, according to research that I authored last year with Duncan Simester at MIT Sloan School of Management. Some of an apparel company’s best customers may write negative reviews about products that they have never purchased, the study found. While it’s unclear why, one reason is that these customers are trying to influence both what other customers buy and the types of products the company sells.
That research spurred us to also look for fake book reviews on Amazon.com. We found that book reviews without an “Amazon Verified Purchase” tag are nearly twice as likely to receive a one-star review (9.4% vs. 4.8%). The reasons lower-rated reviews tended not to be verified purchases could be linked to any number of reasons, including a disgruntled shopper or a competitor. Reviews that were verified by purchases tended to have higher ratings. One possible explanation was that authors or their friends purchased books from Amazon when submitting favorable reviews to inflate their ratings. Amazon’s lawsuit confirms the suspicions raised in the research that third-party firms offering Amazon Verified Reviews for a fee (which includes the cost of purchasing the book through Amazon) have been skewing the ratings higher.
Cracking down on fake reviews will help assure consumers that opinions are those of fellow shoppers and not just hype for hire. But, to me, the larger issue is increased participation, which will help to drown out the noise from fake reviews and their undue influence on consumer purchases.
Eric T. Anderson is a professor of marketing at Kellogg School of Management at Northwestern University and director of the Center for Global Marketing Practice.