Tue, 23 Apr 13 at 14:44 | 14 Comments
The long tail
In 2012, the average e-book earned its author less than 300 dollars.
And that’s the average, pushed higher by a small number of very, very successful books. The median result is likely even lower.
Now, this figure is an estimate. Though the 800-pound gorilla — that would be Amazon — sells the vast majority of e-books, the company is extremely secretive about actual numbers. Individual authors can see their results, but aggregate figures? You might as well ask for the President’s nuclear launch codes.
Now and then Jeff Bezos lets slip a small, isolated data point. The most recent was in January, when he noted that ebook sales had risen 70% — but was that unit sales or revenue? And what was the total? No further details.
Nonetheless, we can use a range of data from disparate sources to come up with a good guess. I’ve detailed these sources, assumptions and calculations below. If you’re interested, take a look and see what you think.
Of course, many ebooks aren’t intended to make much money. Forget free books, which are left out of these calculations entirely (include them, and the average would drop even lower). But other types of books will never see big sales: assembly-line reprints of out-of-copyright “classics,” auto-generated Kindle spam, vanity publications, one-offs intended as lecture-circuit lagniappes, and so forth.
Still, at $300 each, you’d have to publish 48 books per year just to make minimum wage — if you could write them in all in your waking hours.
Amazon, by the way, is doing just great in this model. A couple dollars per book isn’t much for the individual author, but with two million titles out there, the money adds right up for them.
And maybe that’s why they’re so closed-mouthed about the numbers.
Methodology & Sources
Amazon dominates, so we’re going to look at just Amazon ebooks. Given the fungible nature of the marketplace, however, there’s little reason to think results would vary with B&N, Apple and the others included.
This is complicated, so I’m including the spreadsheet here, if you’d like to follow along.
First, we calculate the US share of worldwide book sales. 2012 global figures aren’t out yet, but 2011 is good enough; it probably hasn’t changed enough to affect the proportion substantially. So: 2011 book sales were $27.2bn (Association of American Publishers, AAP) and US sales were $6.7bn (also AAP), for a US share of 24.6%.
2012 US book sales were $7.1bn (AAP), and we can extrapolate for a 2012 worldwide number of $28.8bn.
Key assumption: The problem with all “official” statistics is that they miss independent, self-published books. AAP’s figures rely on reporting from approximately 1,000 publishers; Bowker tracks ISBNs, which many self-publishers don’t bother registering (they use Amazon’s ASIN number instead). However, I’m going to assume that the significant difference is in e-books — ie, that self-published printed books are an immaterial part of the total.
So let’s figure out how much money Amazon made from “officially counted” printed books. Bowker Market Research reports that AMZN has a 27% market share, of a total pie that is 78% printed (non-ebook). Applying these percentages to the total $28.8bn above, we estimate Amazon’s total (worldwide) 2012 revenue from printed books at $6.07bn.
Now, one of the very few statistics Amazon actually reports is “total media sales,” and in 2012 they amounted to $19.9bn. However, this figure includes DVDs, CDs, music and streaming media. To determine what all this non-book stuff adds up to, we need …
Another Assumption: That the ratio of “home entertainment” (as this category is known) to “books” revenue is the same nationally and at Amazon. After all, why not? Amazon is probably the biggest player in both anyway — except for streaming, which is dominated by Netflix. So let’s remove the streaming revenue. According to the Digital Entertainment Group, total US home entertainment sales were $18bn in 2012. Of that, “almost 30%” was digital distribution, or $5.4bn, leaving $12.6bn net.
Separately, Netflix streaming revenue in 2012 was $3.6bn. Removing that from total streaming — the $5.4bn we just just calculated — leaves $1.8bn for everyone else. How much of that did Amazon have? Let’s call it 1/3, assuming that Hulu, a larger player, makes twice as much as Amazon [see Note]. That leaves AMZN with $0.5bn. Could be a few hundred million more or less, but that won’t make much difference to the final results.
OK, so total media sales for all companies (less streaming) is $12.6bn. That’s 1.77 times US book sales (of $7.1bn), or expressed another way, 63.9% of the of media sales (including books but not streaming). Applying this ratio to AMZN’s $19.4 (media sales less streaming), we find that everything in Amazon’s figure that isn’t books equals $12.4bn.
Thus total Amazon book sales in 2012 are estimated at the difference, or $6.98 billion.
Meanwhile, remember that we determined Amazon’s printed-book revenue to be $6.07bn. Subtract that from total book revenue of $6.98bn, and we’re left with their 2012 Kindle book revenue: $910 million.
Take a second to marvel at the number. Once more:
2012 Amazon ebook revenue: $910 million
Amazon NEVER even hints at this, and we figured it out! Amazing! (if I say so myself.)
Moving on, there are currently 1,933,163 ebooks in the Kindle store. Limit the search to “free ebooks” and we find 59,720. For the purposes of this analysis we need to ignore the free books, so that means a total of 1,873,443. Divide this into total revenue (again, $910m) and we get an average revenue per Amazon ebook in 2012 of $484.
The last step is to estimate out how much the authors are making from these sales. What we need to know is how many of those 1.87m titles are priced under $2.99 (35% royalty) vs. $2.99 and up (70% royalty). I just counted the number of <$2.99 books in the Kindle Top 100: 28. This is higher than last year (21) which makes sense, given increased downward pressure on prices; let's call it 25 percent. Do a little arithmetic, and we see that of that average $484, the author is getting $297.
Note: If book sales are tricky, determining Amazon’s streaming revenues is even harder. Many (most?) customers watch streaming video under their Prime membership, rather than pay-per-view. Proper cost accounting would allocate the appropriate portion of Prime membership fees to media sales. Does Amazon do this? Who the hell knows? That’s waa-a-ay more detail than is available in their public financials. Fortunately, as mentioned, the exact figure doesn’t matter much. You can run your own sensitivity analysis on the spreadsheet if you want.