Amazon.com is a customer-centric company. Few others have been as successful in achieving a public brand perception that is so aligned with their self-presentation. Customers are to Jeff Bezos, Amazon’s founder, as design was to the late Steve Jobs. From the time the company was founded, Bezos became known for his habit of leaving a chair unoccupied at meetings, to remind those present that the customer was the most important person in the room.

Putting the customer first has always been how Amazon justifies itself. Its ruthlessness in negotiating with suppliers and undercutting the competition is excused not in the terms of conventional business practice, but as part of the never-ending quest to perfect customer experience. On occasion, Amazon has not lived up to this ideal. But, by and large, its commitment to the interest of the customer is genuine and an indispensable component of its global dominance.

An alternative to the intelligentsia

Amazon started out in 1994 as an online book store. Globally and in India, books now constitute only a small fraction of the company’s sales, but Amazon had revolutionised the book business and continues to be the single-most powerful entity in the history of the industry. One of its enduring innovations is customer reviews, published alongside – and often given more prominence than – a book’s official promotional material. Before Amazon, word of mouth was the only alternative the traditional and so-called professional reviews for customers who wanted to know what others thought of a book before buying it.

Publishing unedited and unfiltered customer reviewers went against the orthodox logic of bookselling. After all, why would a retailer reveal that most people disliked a product? But it embodied Amazon’s customer-centric philosophy and was a harbinger of a broader two-decade trend that has seen critics and other gatekeepers edged out by the wisdom (or lack thereof) of crowds.

In terms of rigour, depth or fairness, Amazon reviews can rarely be confused with literary criticism, but they speak to two traits that book buyers share with consumers on the whole: a desire to make informed choices, and what could be called the fear of missing out or, more cruelly, herd mentality – if others are reading something, I should too (it is this second need that publishers appeal to when they label a book on its cover as a bestseller).

The fallout

For better or worse, Amazon book reviews are here to stay. But recent events have shown up the limitations of these reviews, at least in India, not in literary terms but in something that ought to cause Amazon great concern – their value to the customer. Amazon has always allowed customers who have not actually bought the book or other product in question to review it, as long as they have an account on the website and have made at least one purchase. It is this feature that has allowed trolls to hijack the pages of books they have not read, piling one-star review upon one-star review until the page resembles a pavement garbage dump.

The first victim of this strategy was Barkha Dutt whose book, This Unquiet Land, was subjected to an organised campaign of negative reviews. On a smaller scale, many of the titles in the Murty Classical Library of India series have been similarly trolled, as has Wendy Doniger’s The Hindus. The most recent, and egregious case is that of Rana Ayyub’s self-published work of investigative journalism – Gujarat Files: Anatomy of a Cover Up. Less than two weeks after its release, the book has already elicited 1,234 reviews on Amazon.in 844 of them one-star.

The campaigns against Dutt and Ayyub’s books have much in common. In both cases, the targets are women journalists – it is well-documented that women receive a disproportionate and especially vicious share of online abuse – who are perceived to be critics of Hindu nationalism and Prime Minister Narendra Modi. Ayyub is a Muslim– as is Dutt, according to one of the many pernicious online myths about her. And the overwhelming majority of those giving their books one-star reviews have neither bought nor read the books in question.

This last point is the only conclusion that sustains a thorough reading of the reviews in question – not an arduous task when so many of them are one word long. But it is reinforced by a comparison of those reviews marked “Verified Purchase” – indicating that the customer has, in fact, purchased the book on Amazon – and those that aren’t.

It's in the maths

Only three of the 844 one-star reviews of Gujarat Files – and none of the nine reviews that bear two and three stars – are marked Verified Purchase. By contrast, nine of 21 four-star reviews and 89 of 344 five-star reviews are by customers marked verified. Of the 101 people that Amazon knows who bought Gujarat Files, 98 rated it four or five stars. Yet, right at the top of the page, the customer is informed that 1,234 people have given it an average rating of 2/5.

The picture for Dutt’s This Unquiet Land is depressingly similar. Only 18 of 3,890 one-star reviews are by verified buyers. There are only 25 four-star reviews, but a larger number of these – 19 – are verified. A person who gave the book four stars is more than 160 times as likely to have actually bought the book on Amazon than the one who gave it a single star. For five-star reviews (40 our of 139 verified), the figure is more modest — the reviewer is 62 times as likely to have purchased the book on Amazon.

For Doniger’s The Hindus 16 of 18 five-star reviews are by verified purchasers, but just one of the 20 one-star reviews is.

By allowing anyone with an account to review any product, Amazon, in theory, allows customers the widest possible set of views. But when an agenda-driven mob attempts to crowd out legitimate reviews, ordinary customers lose. The books themselves can, perversely, actually benefit – Ayyub is now Number 1 on Amazon’s bestseller list, in part because the barrage of negative reviews have attracted both publicity and a counter-campaign in support of the book.

In the interest of customers

Amazon’s customer review guidelines are hilariously pedantic: A review can be deleted if it “comments on other reviews visible on the page (because page visibility is subject to change without notice)” or if it is written in a “foreign language”. But none of the guidelines apply to organised trolling, because unless there is a financial or personal conflict of interest, Amazon makes no provision for reviews written in bad faith.

These guidelines may work elsewhere – on amazon.com (US), most reviews tend to be verified and thus legitimate. Take two recent bestsellers: 2,615 out of 3,063 reviews for Paul Kalanithi’s When Breath Becomes Air are marked verified, as are 611 out of 730 for Jane Mayer’s Dark Money, a book by a female investigative journalist who is loathed by American right-wingers.

But in India the potential for abuse has already proved too great. Amazon is a retailer, not a media organisation: it has no obligation to provide a platform for unfettered expression that is designed purely to sabotage. Customers do not benefit from having to wade through hundreds of reviews from those who have never read a book to find the genuine opinions of others. Amazon should never delete or censor negative reviews – as long as they really are reviews.

Amazon has shown a willingness to modify global practices to suit Indian realities – by embracing, for instance, cash on delivery and the marketplace model. Online trolling of this kind is not unique to India, but arguably, nowhere else does it do such damage. In India, Amazon should either restrict customer reviews to Verified Purchases for all products, or reserve the right to do so when a book or other product is evidently the subject of trolling. Only then can it claim to be truly putting its customers first.