tl;dr: When there’s only one book retailer (and increasingly only one publisher) reviews are the only differentiator between books, yet they are being widely abused. At the same time there is a rapid race to the bottom of both content quality and marketing tactics. The utopian vision that Amazon’s self publishing tools presents is tainted and won’t get better until Amazon accept some responsibility, or until an alternative retailer can make a big enough dent in the market to matter.
A story for you. My wife has written a diet book. Claire is doing the diet herself, it’s working for her, and the text and recipes within the book are her own original work. And people are (or were) buying it.
Let me also state right now that, despite how it will doubtless sound later on, this is not intended to be a whinge about competition. As dispassionately as possible I’d like to discuss how Amazon, as sole retailer and publisher, also owns the only way consumers can discriminate between products, and how this is fundamentally broken. I’ve deliberately not linked to the books in this article to try and avoid making this anything other than a story about some stuff that happened. That this is based on Claire’s experience is a necessary fact, but not the point.
Claire’s book is published via Amazon’s KDP and Createspace services. This is an increasingly common way for authors to bring their books to market; between them they provide both digital and physical publishing tools to allow authors to self publish their work on the Amazon network of sites.
Private pressings (or “vanity publishing” as it’s sometimes disparagingly known in the trade) have always existed. Having invested long years of your life in preparing your magnum opus it’s obvious (if you have the resources, at least) that the shortsightedness of a closed, cliquey, distant few shouldn’t affect whether or not people have access to your book. After all, we all have a book inside us; self publishing allows those with a little cash to taste the pleasures of seeing their name on a dust jacket.
It used to be that when an author approached a specialist book binder they were gently talked down from a print run of thousands, while at the same time talked up into taking additional services such as proof reading, editing and jacket design; all the things that help turn a manuscript into a book, in fact. And so what might have been two thousand error strewn paperbacks stored in the spare bedroom for years to come becomes 100 or so beautifully bound and grammatically correct hardbacks for distribution to exactly the right bookshelves. The author’s goals are met, the printer’s margins have been increased and the practical realities of the dead tree publishing industry explained and (in most cases) understood.
Like with so many old school media industries a few years ago the Internet began to change all that. First there were print on demand services like Lulu, and then five years ago Amazon waded in with KDP (Kindle Direct Publishing) that allowed authors to sell digital versions of their books directly on the Amazon site. And then two years ago Amazon bought Createspace (a print on demand rival to Lulu). At this point the circle was complete; with a few simple tools an author can now get both paper and digital versions of their books for sale in the world’s largest bookshop, linked together, consistently listed and ranked.
As a long term resident of the Internet I whole heartedly approve of yet another dinosaur of an industry being disintermediated. Seriously, an industry where the recommendation is to send a copy of your manuscript to upwards of 50 people from whom you are warned you shouldn’t expect a response within 6 months is ripe for being torn a new one. Amazon saw that, and started tearing. And more power to them, frankly.
In this process we have disintermediated traditional publishers who took, on average, 75% of net receipts (this is important – it means the author bears the cost of discounting to retailers) and replaced them with a retail channel that offers 65% to 70% of list price (meaning that any discounting is borne by the retailer). A £7 paperback published by a traditional publisher will earn the author around £1 per copy. Traditionally published ebooks are sold at the same price and under the same contractual terms, so that £7 book as an ebook also earns the author £1. Published via KDP for £7 the book will net the author around £4.55 per copy. Sold as a print on demand paperback via Createspace it will earn about £2.10. And let’s not forget that due to the structure of publishing deals (net receipts again) if that £7 traditionally published book is sold in Tesco it might only earn the author 10p per copy sold.
So if you know, in your heart of hearts, that your book is not going to be one of the 10 a year that get the full marketing weight of the publishing house behind it (and that’s increasingly true for many major authors let alone new ones) then why go through the heartbreak of dozens of rejection letters when you could get straight to market via Amazon, sell your book digitally for £4 a copy and earn £2.50 a sale?
Of course, you’d be advised to buy freelance proof reading at a minimum (no one can proof their own work. No one), and maybe editing too. Proofing costs somewhere in the region of $150 for a medium sized novel (there are literally hundreds of people offering this service in the Kindle forums), so hopefully won’t break the bank. And jacket design isn’t quite as important as it once was, now that people only see a postage stamp sized thumbnail of the jacket before buying.
All very utopian, no?
The big question, of course, is how does the author market their work? 65% of no sales works out just the same as 25%. In the traditional publishing view of the world there is poster and TV advertising, weekly book supplements in the newspapers, placement on the 2 for 1 table in Waterstones and a myriad of other ways where the publisher’s resources ensure a route to market.
Self-published authors can’t tap into those routes. There are no posters, no adverts, no ready and willing newspaper columnists with a deadline and inches to fill and no two for one table upon which to place their books. Instead, after self publicity through channels like blogs, Twitter and Facebook authors are singly and utterly reliant on Amazon reviews.
And therein lies the rub. Amazon’s review system places very few controls on who can review what. You can’t, for example, leave reviews on the same item from different accounts that share a credit card, but that’s about it. That’s not much of a control though; all it takes is a quick ring round your friends to get the reviews up. Or if you’re too embarrassed to ask your friends you can buy them on Fiverr instead.
A couple of fake five star reviews to kick start the book isn’t too bad. Friends and family want you to succeed and are going to rate your book highly whatever. What’s killing the Amazon review system is the lack of control of negative reviews from competing authors and publishers. Consumers have to take reviews at face value, and a couple of scathing reviews completely destroy sales.
And here’s where we get to Claire’s experience. The book she’s written is about the “5:2 diet”. There’s been a bunch of publicity around the diet recently, following a BBC Horizon documentary and a couple of articles in the mainstream press (including one in The Sun by, of all people, Dom Joly). The big publishers move too slowly to get books out quickly, and when everyone suddenly wants a book about this miracle diet it’s only the self published authors who are there to service that market (take note, publishers, by the way). I’d argue that this is one of the many beauties of the KPD/Createspace system.
Claire’s book was receiving some nice reviews and selling very well until the negative reviews started appearing. And they were vitriolic: “Cashing in”, “Read the Times instead”, “rubbish”, “This book has been rushed and it shows”, “Nothing new”. It’s hard not to take those sorts of comments to heart. We spent a lot of time staring at the book and wondering how some people could interpret it the way they had, and why the reviews were so polarised. We certainly didn’t feel it was rushed, it wasn’t badly formatted, all the content was original, including calorie tables and tried and tested recipes, and the description of the product made clear that it was a short book.
It took us a while to work out what was going on, but eventually we twigged that other people had had the same idea as Claire to write a book about the diet, and saw negative reviews on her book as a way of boosting the sales of their own. At one point, with some lovely (real!) positive reviews of the book, she was selling close to 80 copies a day. Now that the book has been swamped with negative reviews (all completely at odds with the positive ones) she’s down to 10 copies a day and falling. At the same time less scrupulous publishers are riding high in the charts, bouyed both by the positive reviews they’ve left on their own book and by the negative ones they’ve left on Claire’s.
We’ve complained to Amazon. We’ve marked reviews as inappropriate. But nothing’s been done. And now we live in fear of another negative review hitting our main competitor, because each time a bad one lands on theirs, another negative one appears on Claire’s.
Of the ten 1 and 2 star reviews on Claire’s book 5 are not verified purchases. (Of the verified purchases one is a genuine complaint about a misprinting by Createspace, one is from a nutter, and one is by the author of the “rival” book, but more of that later.) Given that the book is published by Amazon and only available through Amazon it seems strange that such a high proportion of reviews are not verified purchases; the user leaving the review has to explicitly choose to remove the Verified Purchase message from their review.
Indeed, if we look at other KDP/Createspace books where there are lots of reviews (I used books with more than 30 reviews) and where the reviews are relatively consistent (i.e. not polarised) we can see that on average only between 10 and 15% of reviews are marked as not verified purchases. On books where the reviews are polarised the reviews show the same characteristics as those on Claire’s – a disproportionate number of either the positive or the negative reviews are not verified purchases.
To continue this train of thought we can examine the reviews on the rival diet book. Here we see that none of the 5 star reviews are verified purchases (and all were written in the same 48 hour period), yet both of the one star reviews do come from verified purchases.
And it’s not just the unverified reviews that can be untrustworthy. One of the negative verified purchase reviews on Claire’s book is from a user that calls herself “Crafty Mama”. She used to have a five star review on the rival book until it disappeared (one assumes because she tried to leave another review on the same book from an account that shared a credit card). This user has reviewed very few books but appears to have very strong opinions about Ukelele books. The Ukelele book she likes is published by the same publisher as the rival diet book. The one she doesn’t is not.
There’s also an unverified 5 star review left on the rival book by “Cooking Mama” who has only reviewed one other book, published by the same publisher and being a book to which “Crafty Mama” has uploaded customer images.
With most fake reviews you can’t prove anything. If, like Cooking/Crafty Mama, you do it consistently there’s an easy to follow trail, but there’s no real way for a consumer to tell if a one off review is fake or not. And it’s not like most people would even notice those with the obvious trails – it’s just too many clicks.
I can only assume that if this is happening to Claire’s book (a pretty niche title) it’s happening to hundreds if not thousands of others as you read this.
Ideally the vision of self publishing books via Amazon is that every author gets a fair crack of the whip in an open marketplace. One of our friends (the lovely Sophie, whose second novel is out now) had huge success with her first novel last year, selling tens of thousands of copies via KDP, and it’s that type of story that Amazon would have us buy in to. What we aren’t hearing so much about (yet) is the vast morass of books produced by unscrupulous authors/publishers who are unafraid of violating copyright or plagiarism, or of damning competitors with fake reviews.
Where the Amazon self publishing tools should really be helping is in the ‘long tail’ of low volume books where publishers are less and less inclined to take an interest. Unfortunately it’s exactly in these low volume books where the reviews are least well policed. There’s been considerable noise and bluster recently about established authors posting both positive and negative fake reviews, but down the sales ranks and in the niche subjects no one’s paying any attention. Yet it’s there that thousands of authors are trying to eke out a living with no external support.
Because the review system is so baked-in to the Amazon experience I would argue that it has to be managed very closely – certainly more closely than it is now. Amazon’s success with the ubiquity of the Kindle is even more powerful than Apple’s early success with the iPod. Right now the self published book market doesn’t have an editorial voice or a recommendation system outside Amazon. Unlike, say, Pitchfork or The Quietus for music there isn’t an established independent voice (or voices) yet. The book social networks like Goodreads are a way off providing recommendations of the quality of Last.fm. And because getting content onto a Kindle by any other means than via the Amazon store is so cumbersome there’s little to no opportunity for direct sales via your own site or via an alternative service (like Bandcamp, for music). In every way we are more tied to Amazon and the Kindle than we ever were to iTunes and the iPod.
The utopian ideal of disintermediating the large publishers is turning into a filthy race to the bottom, shepherded by Amazon. Their apparent position of “as long as you buy something we don’t care” allows the unscrupulous to rise to the top in all but the very biggest genres, ruining the experience for authors and readers alike.
What could Amazon do to fix it? I think there are a couple of easy steps.
Firstly, if a book is published via KDP or Createspace they should only allow verified purchase reviews. These books are only available through Amazon, so any review that comes from an unverified purchase is either too close to the author or acting against them. It’s not a huge barrier to fake reviews, of course, but it stops a quick ring round your mates having quite the effect it does now. This would solve almost all of Claire’s book’s problems in one fell swoop, and I imagine would solve a lot of other people’s problems, too.
Secondly, they need to take the ‘flag this review as inappropriate’ reports seriously. If I can identify a proportion of the fake reviews simply from the public information available on the site imagine how easy it would be with additional information like account age, IP addresses and browsing history? Claire’s income from her book has dwindled to nothing since the negative reviews started piling on; I don’t think it’s too much to ask from Amazon as the only intermediary between author and reader to at least investigate these reports and remove the obviously fake reviews.
A couple of relatively simple algorithms would make this easy, I think. Viewed less then 4 items on the site? Only bought one item? Repeatedly view only the same one or two items? Repeatedly logging in from the same IP as other users with reviews on the same item? Always logging in from the same browser session as another user with reviews on the same item? Always logging in from the same IP, but with no cookies set (using private browsing)? Funded by gift card, rather than credit card? Reviewing items negatively/positively in the same chart as items reviewed positively/negatively by related users (shared credit card, IP, etc)?
If 4 or 5 of those 8 questions come up yes, just don’t publish the review. For a long time I had assumed that it was simply a case of the more reviews the better, but when upwards of 50% of reviews are fake this no longer holds true. Only 3 reviews, but all good quality, is far better than 6 good quality ones out of a total of 12.
At the time of writing Claire’s book has 16 reviews. We’re pretty confident that 9 of the negative ones are fake (and 2 more are just insane, but you can’t really adjudicate for that). At the same time, the rival book has 10 reviews of which we’re confident 6 of the positive ones are fakes (timing, lack of verified purchase, links to other books from the same publisher, etc).
Yet the only way readers can decide between the two titles is by this review system. It’s all a bit messed up.
Claire has just published a second edition of her book. What do we do if that starts getting poisoned by fake reviews too? Stoop to the same level and start slating our competitors? Or take some ineffectual moral high ground and watch sales plummet again? With the Amazon marketplace structured the way it is these are currently our only two options. Neither work for me.