The Amazon’s Choice endorsement is known to increase sales of the endorsed product by up to 300% on the Amazon platform but the label itself has attracted a lot of controversy recently as some experts are suggesting that Amazon’s Choice labels are given to products that aren’t really the best products available. Additionally, there are unconfirmed rumours that the Amazon’s Choice label can be won unfairly, with purchases linked to fake reviews counting towards the assessment of whether an Amazon’s Choice is received. Controversy has also emanated from Amazon’s failure to clarify the precise process behind how the accolades are awarded, with many suggesting that algorithms and robots decide which products are awarded the coveted label. This article will look at the Amazon’s Choice label and consider what implications the controversy has for Amazon and for consumers in general.
Amazon’s Choice
On the Amazon website, there are hundreds of similar products available so customers have lots of choice and options when it comes to selecting the product they want to buy. For instance, a simple search for “towels” returns more than 60,000 search results on Amazon, and some estimates suggest that as many as 10 million products are available for delivery on the website. While this competitive “arena” for products is, in many respects a positive thing, it is easy to see how customers can get overwhelmed by the choices available. Amazon’s solution to this problem was to create the Amazon’s Choice label, which is an icon that gets added to a product page to highlight a product and recommend it as endorsed by Amazon. The Amazon’s Choice icon is distinctively black and is highly visible both in search results and on product descriptions. Furthermore, the Amazon’s Choice option is the default option that Amazon chooses when customers use “voice” commands to search for products to buy on Amazon. As such, the label has a high value for sellers, as the endorsement is likely to increase their sales and the profile of their products considerably.
Amazon’s Choice was introduced as part of a wider initiative to enable customers to select the best products quickly. This initiative included allowing customers to search for items based on which were best sellers and which had the highest star rating. Amazon adds “best seller” badges to a variety of products, and this feature is updated on an hourly basis. This badge doesn’t take account of any other data other than sales volume.
Nobody really knows exactly how Amazon selects Amazon’s Choice products. Amazon itself states that “Amazon's Choice recommends highly rated, well-priced products available to ship immediately”, so it is clear that the label is created after analysis of how often a product is sold, how many reviews it has receives, what type of reviews it receives, its price and shipping options. Perhaps the biggest controversy surrounding the Amazon Choice label is the fact that nobody knows whether the label is awarded purely on the basis of data, or whether human judgment is ever used to make the selection as to which product to endorse. Experts suggest that human involvement in the addition of Amazon’s Choice labels to millions of Amazon products is highly unlikely, surmising that it is far more likely that the system is entirely based on automation and calculated according to an algorithm created by Amazon. The precise formulation of these algorithms has never been revealed, but because Amazon’s Choice labels don’t necessarily go to the cheapest or the most highly rated products, commentators have speculated that complex data like how often products are returned also factor into the acquisition of the coveted black Amazon’s Choice. The reliance of data is sometimes highlighted as the reason why the Amazon’s Choice award can endorse products that aren’t the best in their categories. Many consumer groups like Which? magazine has repeatedly pointed this out as being a major problem with the Amazon’s Choice endorsement, and current advice from Which? is to make decisions about products based on all of the available information, including a personal assessment of the reliability of all reviews and endorsements.
Amazon has been quick to address the problems and controversies surrounding their Amazon’s Choice feature. As of 2019, it is making it possible for third party reviews to appear on the site to endorse products. This is an important change as it allows for human driven product reviews to appear on the site and be assessed by buyers as they weigh up what products to buy.
Can We Trust Amazon’s Choice?
Amazon’s Choice was relatively uncontroversial when it was first introduced. It was simply an innovative way to prevent what was dubbed “choice overload”. However, controversy has mounted as people began to complain that the system wasn’t really selecting the best products.
One of the biggest indictments of Amazon Choice was a 2018 investigation by Which? magazine, which uncovered a “cottage industry” of fake reviews that were being taken account of by Amazon in its assessment of which products to give Amazon’s Choice endorsements to.
The investigation revealed just how common fake reviews were and how detrimental they are to customers and genuine retailers. The main motivation for retailers buying fake reviews is just how useful they are when it comes to boosting sales. Buying a fake review also addresses the problem that genuine reviews are hard to come by, despite how critical they are in a buyer’s selection of the product they wish to buy. A 2011 survey revealed that although 87% of customers surveyed stated that a positive review helped them decide to make their purchase, only 3 – 11% of people who bought products actually left a genuine review.
The investigation revealed a number of Facebook groups, which had been set up to refund customers for their purchases in exchange for a five-star review. Furthermore, honest reviewers were harassed and threatened with a refusal to provide the pre-agreed refund.
The Which? investigation was an undercover one where Which? investigators set up accounts on Amazon, PayPal and Facebook. In doing so they discovered several Facebook groups including Amazon Deals Group and Amazon UK Reviewers. A total of 7 of these groups had a collective membership of almost 90,000 members, and the groups were offering access to a scam, now called ‘incentivised reviews’, which was basically a system of where fake reviews could be ordered by retailers in exchange for free products, or cash incentives. These incentives were then offered to garner reviews that were not genuine. Some members of the groups were offering reduced rates for products, or free products in exchange for fake reviews. When contacted, the groups replied through private messaging asking that the investigator search for the product through Amazon, and purchase it, in the knowledge that they would receive a full refund once the review had been posted. This was to guarantee that Amazon would class the purchase as a “verified purchase”. Using the Facebook groups, the Which? investigators ordered products for example Wireless Bluetooth headphones and smart watches. Then when they attempted to leave 2 and 3 star reviews, they were refused the refunds they had been promised and messages demanding 5 star reviews were received.
After the investigation was complete Which? approached both Amazon and Facebook with the results. Both companies issued responses showing disapproval of the fake review scams, but their responses offered little explanation of how they would tackle the problem of fake reviewing. Amazon stated that their goal was to make the review process as useful as possible to the customer, and emphasised that fake reviews were not permitted on the website, nor was it permissible to exchange payment or compensation of any kind in exchange for a fake review. It also pointed out that sellers and customers found to have infringed their review guidelines faced action, which included possible termination of their accounts. Facebook was similarly general in their response, highlighting that fake user reviews were not permissible, and that reporting mechanisms for fake and suspicious content were available for use by Facebook users. Facebook reiterated that they encouraged the use of their reporting tools and that any user found to be in breach of their standards faced action including termination of their accounts.
As of 2019, Facebook has refused to shut down many of the Facebook groups involved in the fake Amazon review scandal. It states that it has shut down some groups, but others remain open while their investigation continues. Experts including Which? Head of Products Natalie Hitchins has described Facebook’s failure to act as very concerning. She has also been critical of the apparent failure of Facebook executives to proactively remove or suspend groups suspected of involvement in selling fake reviews. Which? Head of Products Natalie Hitchins has also suggested that removed groups have reappeared under pseudonyms and she has called on regulators to take enforcement action based on the results of the Which? investigation.
The aftermath of the scandal saw much criticism of both companies, with many suggesting there was a lack of accountability and that particularly in the case of Amazon, customers were being consistently misled in circumstances where not enough action was being taken to prevent this. Which? described the debacle as tantamount to “fake news” and called on customers to be aware of the fake review problem, and the problem of Amazon’s failure to act to remove fake reviews from their website.
The same controversy has dogged Amazon in the USA. In 2019, high profile Senators Frank Pallone, Jr. and Jan Shackowsky asked Amazon to publicly address the Amazon’s Choice debacle and explain what actions are being taken to protect customers from fake reviews and unmeritorious Amazon’s Choice endorsements of inferior products. They have written a letter to Jeff Bezos asking him to clarify exactly how Amazon’s Choice labels are acquired so that more public scrutiny of the system is possible.
In demanding an explanation the senators cited the investigation carried out by Buzzfeed into fake reviews linked to Amazon’s Choice labels. Buzzfeed damningly refers to the matters exposed in their investigation as “Amazon’s fake review economy”. Their investigation revealed an economy of fake reviewers getting paid to post fake products reviews on Amazon. The investigations anonymously identified a number of so-called “fake reviewers” who get paid to post reviews of products they have never tried. One, “Jake” who spoke to Buzzfeed on condition he would never be identified as he feared his Amazon account would be shut down, explained that he recently received an iPhone case in the post and posted a five star review of the product, despite the fact that he doesn’t even own an iPhone. He received $13 USD to post the review and he got to keep the iPhone case.
The whole operation operates under the guise of legitimacy. Prospective “fake reviewers” register on Facebook groups and other platforms and receive free products to try. They post reviews and they get to keep the product. However, the reviews this so-called “invisible workforce” are posting are at the very least biased, and at worst fraudulent, since people reading the reviews are being misled into thinking that these reviews are created by people who have genuinely bought and used the product and decided to post a review, on their own initiative. The reality is that this invisible workforce are rewarded for posting five star reviews, and many are pressured into posting five star reviews regardless of the quality of the products they endorse. The upshot is that people are being tricked into buying inferior products, and also genuine sellers who are trying hard to compete with retailers assisted by this army of fake reviewers are being disadvantaged. Amazon, although not directly complicit, is being criticised for not doing enough to shut down accounts linked to fake reviews and not doing enough to weed them out in the first place. In 2016, Amazon banned free items or discounted items in exchange for reviews, however the practice is still continuing under the radar. Furthermore, some critics suggest that hallmarks of fake reviews, for example reviews that contain so-called “unnatural language” are not policed on the Amazon site and not removed fast enough.
The Wider Problem Of Fake Reviews
According to Which? magazine 97% of adults who took part in a survey of 2000 adults agreed that reviews had influenced their choices when it came to online shopping. The Competition and Markets Authority or CMA has suggested that as much as 23 billion GBP of retail spend in the UK alone is influenced by reviews.
Which? went on to investigate the incidence of fake reviews and their research suggests that 31% of people surveyed had bought a product, based on its reviews, only to be disappointed by the product.
Which? has now issued guidance to people concerned about fake reviews, advising that it is possible to spot fake reviews on product information pages.
People are advised to consider the language used in the reviews, and consider whether the language used appears natural, and whether the reviews were too long or too short. Reviews that are repetitive should also be regarded with suspicion. People are also advised to check the dates that reviews are posted. If lots of reviews are posted within one short time frame, it is more likely that the reviews are fake. People are also advised to judge for themselves, using their own gut instinct whether the review is impartial or whether it makes valid and useful comments. People are advised to check the users leaving the reviews and look at other reviews these users have left. People who leave multiple reviews at 5 stars should be regarded with suspicion as this is one of the hallmarks of someone leaving fake reviews. People are also encouraged to look at the types of products reviewers have left reviews about. One reviewer who has left multiple reviews of similar products within the same product category could be a fake reviewer, as it is unlikely that a genuine user of a product has purchased many similar products within a short time frame. People are also advised to look for consistency in reviews as a whole. Does one feature of a product meet with contradictory praise and criticism? If so, this could be an indication of a fake review. People are also encouraged to check different sources for reviews, and again check for consistency across all of those sources. Is criticism and praise for various features consistent or is there a stark contrast? A stark contrast can be a cause for concern about, at least some of the reviews available.
Amazon Quizzed Over “Choice” Store Rating – What Now?
Faced with a clearly compromised Amazon’s Choice system, the pressure on Amazon to act to restore trust and faith in their brand and their endorsements is growing. Furthermore, the threat of an intervention by regulators is ever-present, and although Amazon has taken some steps to address the problem, for example removal of fake reviews, many agree that this isn’t enough. In failing to take meaningful steps to address the Amazon’s Choice debacle, Amazon risks losing customers to competitors like Shopify, who have recently announced plans to invest in their own fulfillment network to begin operations in 2019.
