An insightful article from NSFWG member Tim C. Taylor
This blog post should have been easy.
Over Christmas and New Year 2014/15 I self-published two novels that were workshopped at Northampton Science Fiction Writers Group. The first one, in particular, was reshaped in response to workshop comments. In the US, UK, and Australia, the books quickly reached the #1 and #2 spots respectively in the military science fiction bestseller charts. Then they topped the space opera bestseller charts too. That’s not just the Kindle-only charts, but includes paperback, hardback, and audio books too.
I was ecstatic.
Like I said, this blog should have been easy. All I had to do was give a brief woo-hoo, and congratulate NSFWG on their help and our shared success. Job done. Back to writing the next book.
But I hesitated. Could I really say my books were bestsellers without embarrassing myself and the writers group? If I use the term ‘bestseller’ am I a fraud? After all, my books were exclusively available through Amazon. They didn’t appear in the ‘real’ bestseller lists you see in Publishers Weekly, The Bookseller, and national newspapers. Seeing my books at the top of the Amazon charts, and seeing the sales figures come through so quickly that they were ticking up in real time — they were proud enough achievements. I don’t need to pretend I’d made it into the real bestseller charts. Like I said, I need to get on and write the next book in the series.
But, armed with actual sales figures for an Amazon-only book, I went looking at the mainstream bestseller charts to see how I compared with the big guys. What I learned left me utterly astonished. Now, I’ll say up front that I’m talking here about the categories assigned to my books: military science fiction and space opera. But in those subgenres, having started with the assumption that Amazon bestseller charts lacked the legitimacy of the real charts, I now realize I got that completely the wrong way around.
It is the charts produced through systems such as Nielsen Bookscan and publisher surveys that lack legitimacy.
Don’t believe me?
Here’s a fact to get you started.
In their 2014 end of the year reviews, the major publishers and their spokespeople have recently been touting the idea that eBook sales have stagnated. Yet the eBook sales statistics from Amazon aren’t counted when a Kindle book is sold, they come later from surveys of major publishers, and in a few cases from surveys of consumers. Or they don’t come at all!
Books sold by publishers who aren’t part of the survey are either ignored altogether, or their sales figures are extrapolated from publishers who are part of the survey.
Crucially, Amazon doesn’t share its sales figures. So if you buy a Kindle book, it normally only counts in industry statistics if it is published by a mainstream publisher. Since I self-publish, my books don’t count. And if you buy a book from a small publisher or self-publisher, your purchase doesn’t count in the statistics either.
But does that really make much difference to industry-wide pronouncements? After all, the major publishers have the book world tied up, don’t they?
Here’s a more honest statistic than anything you’ll hear through authoritative industry press releases. On 20th Jan 2015 (when I started writing this post) I checked all the books in the top20 of the amazon.com Kindle charts for the categories of military science fiction and space opera. I have little doubt (but no reliable statistics, which is the point of this article) that the amazon.com Kindle Store is the most important eBook retailer in the world, and by a long margin. For the statements of the major publishers about eBooks to have legitimacy, you would expect the vast majority of the Kindle titles on Amazon to be produced by the major publishers.
So how many of those titles in the top20 were from a major publisher? Go on, guess…
The answer is zero.
I looked at the space opera charts and got the same result. Not a single book in the top 20 was from mainstream publishers (unless you counted one that was from Amazon’s own imprint). Then I extended to the top 40. Not a single big publisher in the military sf top 40, and just one entry ("Ender’s Game") in space opera. Again, this is only my neck of the science fiction market, and these genres are probably an extreme case. Nonetheless, when the publishing industry tells you facts about eBook sales stagnating, bear in mind that in arriving at their figures, in most cases they are currently excluding every single one of the top20 bestselling Kindle books in military sf and space opera. [To be fair two eBooks in these top 20s did have an ISBN. That means their sales weren’t counted directly, or reported by the publisher (Amazon’s imprint, 47 North), but in one of the book industry surveys (BookStats), there would have been a crude attempt to extrapolate their sales. Respectable bestseller charts would still have ignored them.]
Now, I know I’m repeating myself, but I feel the need to point out that military science fiction is probably an extreme case, but the degree to which my genre is misrepresented by some industry statistics is jaw-dropping.
When they write their press statements about the state of the book industry, can you imagine the small print about methodology that the publisher organizations should have written?
In the case of military science fiction and space opera genres, we arrived at our eBook sales figures by excluding all bestselling Kindle titles, and will continue to ignore the buying habits of the vast majority of Amazon consumers, except on the rare occasions when they purchase a book published by one of our member publishing corporations. If the publisher decided to use an ISBN then we may attempt to guess sales figures by averaging every book sold by the publishers we talk to, and using that average to extrapolate the sales of all the publishers that we don’t talk to.
Getting the facts about how the various published bestseller charts are constructed is difficult or impossible to come by, because the details are often proprietary. Some national newspapers still use the old method of surveying bookstores. In the UK and US, an organization called Nielsen has a point-of-sale system installed at a majority of traditional bookstores. It gives a fair idea of the relative sales of books traditional books sold through brick and mortar stores, but how can it report eBook sales when only a tiny proportion are bought from the old retailers?
Here in the UK, the ‘official’ charts are produced by a magazine called The Bookseller. They take the Nielsen data for print books, and for eBooks they survey a panel of publishers. Those bestselling self-published Kindle eBooks aren’t on that panel. In the UK, when national newspapers and the BBC report on industry statistics, I expect that is the methodology used: ignore all self-published eBook sales.
The Bookseller’s charts explicitly exclude heavily discounted physical books and any eBooks under £2. A key retail price point for small and self-publishers through Amazon.co.uk (because it’s the minimum price to trigger a 70% royalty rate) is £1.99. I will leave you to draw your own conclusions…
Interestingly The Bookseller is about to change the rule that excludes the sub-£2 books. Is this an attempt to add legitimacy to their bestseller charts? The change came about because major publishers complained that some of their bestselling titles were occasionally on sub-£2 promotions, and they wanted those sales included in the bestseller charts. It seems bestseller charts are all about making money out of us by raising the profile of key titles. Nothing more.
This article isn’t an angry rant against the evils of the publishing cartel. I’m intrigued not angry. (Although I promise to get slightly vexed at the end). And what is intriguing me is the legitimacy of a slew of statistics.
Time for another statistic. For the first two weeks of January 2015, I have intimate knowledge of the Amazon.com bestseller charts for space opera and military science fiction, because I had two books race up the top20 to the #1 and #2 slots. Since then I’ve been drifting down. In other words I’ve got a shrewd idea of how to compare Amazon.com’s sales rank during that period with actual unit sales. It’s not precise, but it’s more accurate than any data possessed by Nielsen or the major publishers.
During those two weeks, for the amazon.com military science fiction top-20 I estimate the total unit sales of paperback + hardback + audio book + eBook for self- and independent publishers to be 10,000 books per day, of which a third were Kindle Unlimited borrows. For the top-20 amazon.com book sales from major publishers, I estimate an average of 100 per day. For space opera the majors fared better and there was one title that nearly reached the top-10. I estimate it sold an average of 200 per day over that period. What I haven’t done is count how many titles were in both charts. I’m going to estimate that about half were in both charts. Combine the two top-20 charts. That puts the major publishers averaging 300 sales per day over those two weeks, and self- and independents at 15,000.
In other words, the books that were counted in the ‘real’ bestseller charts were outsold by those excluded at a rate of roughly 50:1.
Up till this point I’ve considered this intriguing, but nothing more. Bestseller baubles are just so much froth. It’s interesting, but having fans who love, your books and customers who make purchases that put food on my family’s table — that’s more important.
Here’s the point where I get vexed. I’m now taking off my author’s hat and putting on my reader’s one. I have another interest in the bestseller charts: as a reader, like you. I’ve bought some of the books there because I’m a fan of the genre and enjoy reading.
It’s only natural to buy a book and want to see it do well in the charts. Just the same as you want your sports team to top their league, and your favorite band to get their album to the top of the music charts.
It’s like a vote. It’s not as important as a political election, but it feels like democracy in action.
It isn’t (at least I hope that’s not how our democracy really works!)
Most purchases I make are for eBooks and my most common price point is £1.99. They don’t count in my country’s bestseller charts. Sorry, pal, we don’t like your sort. Your vote doesn’t count. Now clear off!
If someone’s political vote in an election was ignored because they were black, or a woman, or gay, or expressed the wrong opinion or life choice… well, there would be hell to pay and quite right too.
Discounting someone’s ‘vote’ in a bestseller chart is infinitely more trivial than denying people political voting rights. All the same, it is irksome that the vote I express through my book purchases is ignored, that many of the authors I support and love are disqualified too. As I write this, my two new books have so far sold 25,000 copies. That’s thousands of individuals whose ‘vote’ was also disqualified.
Who is ‘they’ who disqualify the vast majority of the book-buying public in my genre? Unfortunately this is where my vexation dissipates because there isn’t a convenient villain.
Well, Amazon won’t share their sales data. I don’t blame them either; big data really is a key business asset, not to be given away lightly. I don’t blame the major publishers either for always trying to put a gloss on statistics to flatter their message. I held stock in a big publishing company until recently, and that’s precisely what I would want them to do. Nothing wrong there. Big publishers want to downplay the way in which self- and small-publishers have seized the market from them in some genres. I get that, although their statements about eBook statistics — the ones that forget to tell you that they don’t count all Kindle bestselling titles — well, that feels a little brazen, shall we say.
So there is no clear villain, and having journeyed through the tangled thicket of publishing ‘statistics’ in my genre, I can now happily go back to ignoring them as discredited, and having gotten an answer for the one question I asked in the first place.
Can I say with a straight face that the Northampton Science Fiction Writers Group workshopped a #1 international bestseller in military science fiction, despite it only being sold through Amazon?
The answer is yes.
Turns out my wording was wrong. It’s a legitimate #1 not despite being an Amazon bestseller, but because it was an Amazon bestseller. The discovery I’ve made this week is that, in my neck of the industry at least, the Amazon bestseller charts are the only ones with any legitimacy whatsoever.
I recommend AuthorEarnings which takes a detailed analysis of Kindle books on amazon.com, applies surveyed data to map Amazon sales rank to estimated sales, and then presents interesting conclusions. Also, this quarter, is a lot on the subject I’ve been writing on: the legitimacy of official bestseller charts and other statistics.
The statements made are based on a mixture of the author’s memory of Amazon bestseller charts combined with his actual sales statistics and bestseller rankings over that period. I might have failed to spot more traditionally published books in the lower reaches of the bestseller charts, but I don’t think so because I was studying them intently every day. I made a statement that self-publishers outsold major publishers by 50:1. If I’ve gotten my facts badly wrong, maybe it was only 10:1 (although it’s equally possible that it was greater than 50:1).
I don’t claim accuracy. I don’t need to because all I’m trying to do here is answer whether the Northampton Science Fiction Writers Group can legitimately claim to have helped shape a bestseller. No one in the world actually has the data to state how many books were sold across all platforms in all formats, but on the way to answering my question, I stumbled across compelling-enough evidence that: (1) In my genres on Amazon, self- publishers outsold major publishers by a very large margin over this period, and (2) industry statistics on eBook sales come predominantly from publisher surveys, not retailer point-of-sale data, and so ignore most or all self-published titles, and (3) the first two points kind of go together powerfully.
And while we’re on the topic of being honest about statistics, let’s state the obvious. I’m only talking about a single 3-week period, because that is the period when I have good quality information about how many books were being sold by bestsellers in my genre. Better quality information, I might add, than the wider book-publishing industry. However, I studied the amazon.com bestseller charts for military science fiction on many occasions in 2014, and can state with utter conviction (though no statistics) that the top20 was dominated by self-publishers and Amazon’s own imprint, 47 North. Gut feel? For all Kindle military science fiction books sold through Amazon during 2014, I would be astonished if authors published by major publishers accounted for any more than 1% of those unit sales. On the basis of January 2015, that figure is probably closer to 0.1%.
Oh, yes, sometimes people say, but you’re only talking unit sales. Gross sales tell a different figure. During this period, your first book was on sale for 99 cents. That’s practically giving it away. You can’t seriously count that as a sale comparable with a $5.99 paperback.
Oh, yes I can. Gross sales don’t put food on the table: author earnings do. On a typical contract, an author with a mass market paperback selling at $5.99 will make maybe 35 cents per sale (Assuming 40%-45% is taken by retailer and/or distributor. Author earns 10%-15% of the remainder and pays out 15% of that to an agent). For a $4.99 eBook, I’m really not sure. Probably a little more, maybe as much as 60 cents. That’s a guess. My first book is 99 cents, but on average my earnings per sale were a lot more than the author with the $5.99 paperback and more than my guess of the $4.99 eBook too.
What about Kindle Unlimited book ‘sales’? This is Amazon’s new book subscription service: a Spotify for eBooks. I include those in my statistics, and in my genre they are very popular. Detractors might say they aren’t really sales at all. Each Kindle Unlimited borrow in December 2014 of my 99 cent eBook, that was read for at least 10% of its content, earned me $1.25. Compare that with the 35 cents the poor traditionally published author earns with a $5.99 paperback. So far this month, my first Human Legion book has had 3,613 borrows through Kindle Unlimited. Those subscribers will have paid at least $36,000 to Amazon for this month’s subscription (it costs $10 per month in the US and around $13.25 in the UK). That’s real hard-earned money from real book lovers. They may not be traditional sales, but they have already become an important part of the publishing industry, and earn authors (So far. I can only hope it continues…) considerably more than mass market paperback sales through major publishers. So, yes, I count Kindle Unlimited!
For more information on the books mentioned, check out Tim's website here