Chinese sellers have been using a number of underhanded techniques to boost their global rankings and profits on Amazon, The Wall Street Journal reported Monday.
Grabbing the high ground on a search results page is critical for any seller trying to rise above the din created by the more than 500 million products offered on Amazon, which is why some sellers feverishly search for ways to get an edge in the rankings.
“To get one of those spots, some sellers are ready to do anything — fake reviews, artificial sales and even bribes,” WSJ Clment Brge said in a video report.
Unscrupulous sellers use a variety of techniques to vault over competitors in Amazon rankings, according to the report. One of those methods is to undermine the reputation of a rival’s products. Negative reviews can cause serious damage to sales.
“Every day you wake up and the first thing you do is check in your reviews,” Amazon seller Meir Simhi said in the video report. “You want to know if you’re still in the business or not.”
It’s easy to manipulate Amazon’s automated system to hurt competitors, Brge maintained. Flagging a product as “dangerous” or “fake” can get it suspended immediately, for example.
That happened to Simhi two years ago during the Black Friday-Cyber Monday weekend.
“I wake up in the morning one day, I look at my computer and I see zero sales. The reason was my main products were suspended,” he said. “It happens all the time. For me, it can cost tens of thousands of U.S. dollars a year.”
Product reviews are an important component of getting a choice spot in Amazon search results. The more positive reviews a product receives, the higher it rises in the rankings, so unprincipled sellers will seed their products with manufactured positive reviews.
“There’s always been a fake review problem, but it extends beyond Amazon,” said Greg Sterling, vice president of strategy and insight for the Local Search Association, a not-for-profit association of companies engaged in local and location-based marketing.
“It’s a problem that’s existed since reviews began appearing online and continues today, especially now that there’s a lot of evidence that reviews are very influential in consumer buying and rankings,” he told the E-Commerce Times.
Although Amazon suspends sellers it catches generating fake reviews, that can be like a game of whack-a-mole.
“Many Chinese sellers have multiple selling accounts,” Brge pointed out, “so if they get banned from one, they’ll just switch over to another. And if that one gets banned, they’ll just make even more.”
Paralyzing Number of Choices
Just how influential reviews are to consumers was highlighted in a survey conducted by Cone Communications in 2011. Four out of five participating consumers reported changing their mind about purchasing a recommended product or service based solely on negative information they found online.
Eighty-seven percent of participants agreed that a favorable review online confirmed their decision to purchase a product, the survey also found.
Reviews not only point consumers to products, but also can determine if a product ever appears on a consumer’s radar at all.
“The way buyers rely on reviews has grown over time. For example, you can sort by reviews,” Sterling said.
“People can exclude products in a category that falls below a certain ‘star’ review threshold, so there’s a lot at stake for manufacturers, he noted.
“Consumers do that to narrow down the available choices, because in a shopping environment where there are thousands and thousands of choices, you need to find a way to screen your volume of options,” Sterling added. “Otherwise, it can become paralyzing.”
Another factor contributing to rankings is sales. Hucksters have found a way to game that metric, too. Companies can commission a third-party to buy their products and inflate sales figures.
“Some professionals told us they even sent empty boxes to random U.S. addresses and claimed genuine transactions,” WSJ’s Brge noted.
If all else fails, there’s old-fashioned bribery.
Amazon currently is investigating some of its employees in China to see if they’ve been accepting bribes to give advantages to sellers, Brge said.
Sellers have paid bribes ranging from US$80 to $2,000 to Amazon employees to delete negative reviews, restore banned sellers’ accounts, and obtain restricted Amazon data that will help them boost sales, he noted.
Working Hard to Foil Bad Actors
“We work hard every day to earn and keep trust from our customers and sellers,” Amazon said in a statement furnished to the E-Commerce Times by spokesperson Cecilia Fan.
“Bad actors that try to break that trust and abuse our systems make up a tiny fraction of activity on our site. We use sophisticated tools, including machine learning, to combat them, and we are making it increasingly difficult for bad actors to hide,” the company maintained.
“We block bad actors before they reach our site, and we work with sellers and law enforcement to hold them accountable by withholding funds and pursuing civil and criminal penalties,” Amazon pointed out.
“These bad actors show a flagrant disregard for our community, our policies, and in some cases, the law, and do not reflect the flourishing community of honest entrepreneurs that make up the vast majority of our sellers,” the company added.
Because of the volume of work needed to thwart bad actors, automation is a necessity, but it can be a boon for hucksters.
“Automation will lead to a lot of loopholes,” Howard Thai, who specializes in optimizing Amazon rankings, observed in Brge’s report.
“Their system right now is very vulnerable,” seller Simhi noted. “They don’t do enough, and many sellers are getting hurt.”
A Matter of Trust
More artificial intelligence may be necessary, not less, if Amazon wants to protect the innocent and nail the guilty, suggested Rob Enderle, principal analyst at the Enderle Group, an advisory services firm in Bend, Oregon.
“Amazon needs to implement deep learning-based protection for this, because the numbers are so high and the efforts to game the Amazon system so pronounced,” he told the E-Commerce Times.
“Otherwise, they are likely to tag — accidentally — legitimate sellers, and not catch the most aggressive and innovative cheaters,” Enderle added.
Amazon’s existing AI systems aren’t smart enough to do that, he said.
“One of the advantages Amazon has over its competitors is it has a very strong brand. People trust Amazon,” LSA’s Sterling observed.
“If this becomes a problem with too much visibility, it will impact Amazon’s brand, and it will impact consumer behavior,” he continued.
“Much like Yelp,” said Enderle, “if you lose trust in Amazon, you are likely to fall back on brick-and-mortar stores where you can again touch and test before you buy.”