Is the Amazon Customer Review System Broken?

English: Studio publicity portrait of the Amer...
In principle, book reviews spur sales. But on Amazon, they don't seem to. Any author who's following the sales of his/her books can testify to this: when good reviews come in, they rarely signal a spurt in sales. Yet book reviews are needed to be able to use the better advertisers like BookBub that will not take an author's books unless a sizeable number of reviews can be shown, and particularly reviews from "authoritative" sources.

Add to this the now confirmed fact that e-book sales have gone very badly in recent months and Amazon's bottom line is showing it. David Streitfeld in an article in the Business Standard, (see here, it was then picked up by the New York Times) has drawn attention to the fact that "to secure its upper hand, Amazon disrupts its own model."

What Streitfeld is talking about is this: some $250 million in profits that had been expected by analysts somehow went missing last summer. Why? It seems Amazon practiced deep discounts and giveaways, offering free music, videos and e-books and that hurt profits. And of course Wall Street took note.  The argument is that Amazon is doing this on purpose to secure an ever-increasing share of the market - until it will be the only retail colossus on the scene.

But a lot of people have rushed forward with other reasons for the slowdown in Amazon sales, in particular attributing it to the long drawn-out Amazon-Hachette dispute (it last 6 months!) and continuing disaffection from once-loyal clients.

Whatever the reason, there is one thing that works poorly on Amazon and it's their customer reviews of books. As long as Amazon doesn't make an effort to organize it better, Amazon's hope to compete with Hachette or any other "Big Publisher" is doomed. Because traditional publishers have got their buzz-around-books model down pat: they get top reviewers to write in major newspapers and magazines, they organize big prizes like the National Book Award or the Puliter, all events that drive traffic and draw the public. All stuff that's closed to self-published authors and where authors published by Amazon imprints get little space, if any.

But we all know the arguments: This is the digital revolution that has enabled self-published authors to compete with traditional publishers. And Amazon rankings work to show who are the top sellers. And Amazon's system in the Kindle Store is totally democratic, driven by the customers likes and dislikes, allowing everyone to express his or her opinion and the sales numbers speak for themselves.

Until they don't.

You have books with thousands of reviews...and they don't sell or don't sell as much as you might expect. You don't believe me? Look here (a book with over 2,000 reviews and a ranking on Amazon above #6,000) and here (over 800 reviews and a ranking above #10,000). Of course, there's a high correlation if you check out the pages with books that have over 1,000 four-star reviews and above (see here) - there are nearly a million titles there, and the correlation is strong at first, as long as the number of reviews is above one thousand, but as you keep going through them, the correlation starts to fail you.

First lesson: you need at least a thousand reviews (more or less) to hope to sell steadily in the Kindle Store. But your book can still peter out, like this one, the Shadow of the Wind: over 1500 reviews and a ranking that doesn't reflect it, at over #5,000 - and that happens to be one of my favorite books, I highly recommend it, it's an extraordinary combination of dark poetry and suspense.

Second lesson: reviews on Amazon mean relatively little. In spite of Amazon's policing efforts (they've gone after "sock-puppet reviews" in a systematic way since the 2012 scandal), it is still a fact that a 5-star review can be written by a friend or by someone who has absolutely no idea of literature or more simply, doesn't know the difference between a good read and a bad one.  Ditto of one and two-star reviews.

Consider this one, about Elizabeth Taylor's performance (click here), a particularly juicy dyad of 2-star reviews; here's the screenshot:


As the friend who drew my attention to this said: "I'd like to know what Tennessee Williams would have said."

Indeed. And to think that the second reviewer ("Kona") is ranked by Amazon as a "vine voice" and "top 1,000 reviewer"... In short, someone whose reviews are appreciated by both Amazon and its customers. Someone who presumably has an "experienced" taste and a professional touch: you get up there in the Amazon Hall of Fame of reviewers by doing lots and lots of reviews and having lots and lots of people clicking that button which says that "they found the following review helpful".

So what is wrong? I am not going to go into what happens with reviews of other products on sale on Amazon, the endless electronic gadgets, apparel etc - I shall limit myself here to books (a product I happen to know something about).

And books require special handling. You can like or dislike a book but that is not enough to constitute a helpful review. The next person doesn't know you and may not share your tastes. So anyone doing a review should always explain the how and the why a book is likeable or detestable: that's only fair to whoever is going to read your review. And the reviewer needs to come to some sort of conclusion that is reflected in the number of stars awarded. You can't say something is absolutely transcendentally wonderful and then give it 3 stars because it's not the kind of book you normally read or like. If something is transcendentally wonderful, then it deserves 5 stars, full stop. There is an organic linkage between the value judgments expressed and the number of stars given.

Reviews are not easy to do.

In fact, since the 19th century at least, book reviews have been in the hands of literary experts, people who are both widely read and know how to express a judgment clearly. This is far from simple and not everyone can do it or has the time to do it. Professors of literature at universities can do this successfully, they have the time and in a way it's part of their job; best-selling authors can do this, and in general writers are good at this because they were all born readers first. You are never going to be a good writer if you're not an avid reader in the first place; and someone able to read critically, as Francine Prose has so masterfully explained in her book Reading like a Writer: A Guide for People Who Love Books and For Those Who Want to Write Them (HarperCollins, 2006) By the way, if you haven't read it, I highly recommend it!

On this blog, I have argued in the past (see here) that Amazon should provide help in guiding book reviews, ensuring that major points normally covered in a professional book review are in fact covered (for example, the setting, the development of characters, plot pace, language/dialogues etc). But if Amazon is not willing to help and possibly fears that this type of guidance would be viewed by its customers as an unbearable intrusion, then there is another way to do this.

Amazon already has in place this Vine Program for reviewers (I blogged about it here). But at the moment, the program covers reviews of any sort, and top reviewers tend to review anything they wish, and remarkably enough, the very top reviewers (highest rank) cover all sorts of products but no books!  No books at all, or very few books, and in a most desultory way, reaping in fact very few votes from online viewers.

What Amazon should do is establish a Vine Program for Books Only. Book review guidelines should be issued and reviewers would be able to maintain their rank only if they follow the guidelines. And, once the system is up and running, the ranking of reviewers could start to take place, in order to arrive overtime at 1,000 top-notch book reviewers. My guess is that those reviewers are likely to be literary types of all kinds, including bloggers who specialize in reading  books in given genres.

Once Amazon has got a Vine Program for Books going, they should consider revamping the book description page, separating customer reviews from Vine Program reviews. That would be very important. Because at a glance, you could read the reviews that you can trust, that you know come from people who love books, read them all the time and can talk about them in a meaningful way.

At that point, and at that point only, would book reviews start to make sense and jumpstart the famous book buzz everyone is looking for, readers and writers alike...  

Announcement: The World Food Programme desperately needs funds to continue its assistance to Syrian war refugees. Even the smallest donation, a single dollar, helps. It's Christmas time coming up, be generous! To donate online:  wfp.org/forsyrianrefugees or on the WFP site www.wfp.org.
This is what you will see, it is a totally secure site:

 

Comments