Is Amazon’s Fake Review Response 94 Days Too Late?

Phony posts are largely very short-term campaigns

Reviewers-for-hire is a well-documented plague across the world of e-commerce. And networks of fake reviewers aren’t hard to find: Just search on Facebook for “Amazon Review.”

To explore this phenomenon, UCLA Anderson’s Sherry He and Brett Hollenbeck and USC’s Davide Proserpio, with the help of a team of UCLA undergrads, collected data on hundreds of product sellers buying fake reviews from the internet and then tracked what happened to ratings and sales of the promoted products.

The study began by documenting more than 20 private Facebook groups in which Amazon sellers seek fake reviews. “These groups average 16,000 members each and feature more than 500 posts per day from sellers soliciting reviews,” the researchers report in a working paper. They estimate that reviews were purchased on behalf of as many as 4.5 million products in these groups in the past year. More than 80% of the product sellers are from China, particularly from the Shenzhen manufacturing area, the authors write.

Opt In to the Review Monthly Email Update.

Amazon, it turns out, identifies and deletes fake reviews “at a very high rate,” the researchers say. But the online giant, so adept at rapidly getting goods to your door, is slow in handling this task: The lag from the start of a fake-review campaign to Amazon’s deletion of the posts averages more than 100 days.

That might seem relatively effective if the fake-review campaigns were a yearlong endeavor, but they’re typically short-lived, the researchers find. The campaign has effectively garnered its sales and folded its tent by the time the Amazon eradication program arrives.

Here’s how the authors describe the usual sequence of events after a faux-review campaign is launched:

  • Once hired, a reviewer buys the product so it can be listed as a “verified purchase.” Reviewers are encouraged to leave “lengthy, detailed reviews with photos to seem authentic and organic,” the study says. The “vast majority” of fake-review solicitations compensate the writer by refunding the cost of the product via PayPal after a five-star review has been posted. Sellers also typically cover PayPal’s fee and any sales tax. What’s more, about 15% of product sellers offered a commission on top of refunding the product cost. The average commission was $6.24; the highest observed commission was $15.
  • Soon after a seller starts purchasing fake reviews for a product, the number of reviews posted for that product increases significantly. The product’s average rating and share of five-star reviews also increase in this period. The average number of reviews posted per week rises by seven, roughly doubling the number of write-ups the product received before soliciting fake testimonials. The average product rating rises by about 5%, from 4.3 stars to 4.5 stars at its peak.

The median duration of these campaigns is just six days. And one to two weeks after these promotions end, “Both the number of reviews and average ratings start to decrease substantially,” the study says. By the fourth week following the end of a fake-review campaign the ratings results are even worse. “Products end up having average ratings that are significantly worse than when they started recruiting fake reviews,” the authors write.

Over the following seven-week period the share of one-star reviews as a percentage of all reviews rises from about 8.5% to more than 13%, on average. “The share of one-star reviews starts to increase considerably once recruiting fake reviews stops,” the study finds. The inference is that consumers who bought a product based on phony reviews begin posting low ratings to register their unhappiness with their purchase.

The boom in fake reviews online is “directly linked to Amazon’s decision to encourage sellers from China to sell directly on Amazon,” Hollenbeck said in an email exchange. Before that shift, which began a few years ago, Chinese manufacturers “produced goods for brand name companies in the U.S. who sold those products on Amazon,” Hollenbeck said. He believes that many Chinese manufacturers that now sell directly via Amazon have less incentive to maintain a strong brand reputation. “They cut corners and do things like buy fake reviews to boost sales,” he said.

Though it clocks over 100 days on average to delete them, Amazon does take aim at phony reviews, acknowledging that they threaten the platform’s image as an honest place to do business, the authors say. “In practice, this means that Amazon is not able to eliminate the short-term effects that these reviews have on product sales,” the authors say.

The preliminary results of the ongoing research suggest that phony reviews typically provide a short-term boost to product ratings and sales — but then give way to consumer disappointment and a barrage of poor ratings. “Overall, these results suggest that fake reviews deceive consumers into buying products that then turn out to be of lower quality than expected,” He, Hollenbeck and Proserpio write.

The study’s tracking of Facebook group activity occurred from October 2019 to July 2020. That led to a random sample of about 1,500 products whose sellers were found to be soliciting fake reviews. The three largest product categories in the sample group: “beauty and personal care,” “health and household” and “home and kitchen.”

Featured Faculty

About the Research

He, S., Hollenbeck, B., & Proserpio, D. (2020). The market for fake reviews. 

Related Articles

Uncle Oke's Pot shop store front with an open sign Research Brief / Retail

Small Enterprises Benefit from Economies of Scale in Marijuana Retailing

Washington state data: Profits higher, prices lower at multi-store firms

A woman with long hair hangs up a for sale sign Research Question / Real Estate

Middlemen: Matchmakers or Sources of Strategic Information?

Probing that question using a database of for-sale-by-owner home listings

An exterior night shot of a Dollar General store. Research Brief / Retail

How Dollar Stores Contribute to Food Deserts

Expanding chains drive out independent grocers, reduce access to fresh produce

YouTube on a tablet computer Research Brief / Consumer Behavior

Taking the Battle for Financial Literacy to Where the Eyeballs Are

Research by Bruce Carlin and Stephen Spiller suggests YouTube videos could help consumers make better money decisions