How big is the market for self-published books?
Thanks to Amazon’s policy to non-disclosure, there is a vast black hole where all the data should be. There’s also the matter of definitions. The explosion in self-publishing is the most visible expression of a larger and I suspect even more profound industry shift: the expansion of what publishing actually is, and who and what can call themselves a publisher. Any organisation, author or brand can now publish, more or less effortlessly, and I think it will not be long before we see collectives emerging that are better able to purchase goods and services. All this new activity represents micro-publishing, and represents an entirely new business model to mega-publishing.
I decided to take a top-down approach and try to produce some indicative figures. It was extremely tricky and while I’m not particularly confident in the specific numbers, I am confident in the orders of magnitude.
Initially I looked at the PwC figures, which suggest that the overall US eBook market in 2014, on purchased priced basis, was $5.69bn dollars, a figure that is boosted by business, academic and professional titles, many of which have a higher RRP.
This doesn’t correlate well to other data and is almost 9.5 times higher than similar data for the UK. That said, a little recognised fact is that the overall size of the US trade publishing sector, based on purchase price, is something like 7.5 times larger than the UK 
By way of comparison, The Association of American publishers produced data about trade sales for Q1 2014 which stated that revenues from eBooks were $405m in Q1 , which would annualise to approximately $1,700m for 2014. But this figure represents publisher’s invoice value. I assumed these were retailed at 53% off the DLP  to arrive at a figure of $3,445m (£2,240m) for the size of the US ebook market for trade books, based upon purchased price.
(This figure is contradicted here , which implies that US publishers annual receipts from ebooks were $3,000m, which on a purchase value equates to more like $6,340m. but I think at some level it’s a similar data set)
That would leave a $2,250m pa (£1,463m) ebook market for business, academic and professional titles, or about 40% of the total e-book market, which sounds about right (if a bit high).
If AMZ have 65% of the US trade ebook market, that would give Amazon $3,445m (£2,240m) x 65%, or $2,240m (£1,445m) of revenue from sales of trade ebooks. To make this exercise work, I have got to assume that this figure includes self-published authors.
This correlates quite well to Amazon’s reputed total book sales revenue of $5,250m (£3,410m)  for 2014, representing 42% of the total. This figure must include self-published authors. Remember too that this is only the trade component of those sales.
So, 2014 total trade e-book revenue in the US, based on consumer purchase price, was £2,240m. In the UK, it was approximately £400m. As this figure is provided by the publishers, it can’t include self-published work. But it gives a total of £2,640m for 2014.
The overall reported sales, based on purchases price, of Anglo-American publishers, including BAP sales, in 2014 was $38bn (£25bn) . As the figure above is around 11% of this £25bn, I think it’s in the right ball park.
Let’s go back to Amazon.
In the US, their e-book sales, based on purchase price, were £1,445m.
I think Hugh Howey’s analysis that 31% of sales by purchase price come from self-published authors is the best we have and I’m entirely persuaded that the non-ISBN piece of the eBook business is huge (people in trade houses may find this incomprehensible, but it turns out you don’t need ‘em.)
If Howey’s 31% is right, that would mean that the income to Amazon.com from self-published authors in 2014 was £1,445m x 31% or £450m. If the trade sector as a whole in the US generated £2.240m in purchase price revenues, at 31% this figure is approx £700m.
The UK is harder still to fathom, because the £400m figure seems only to represent established publishers. So let’s assume that the UK e-book market for Amazon is 25% that of the US (£360m in the UK, or 90% of £400m, versus £1,456m in the US, or 65% of £2,239m) and plump for a self-published figure that is 25% of the American one. That is £700m x 25%, or £175m.
That would give a total figure for Amazon in the US and UK, for self-published trade ebooks, based on the purchase price paid by the consumer of £875m in 2014.
This looks like a pretty huge number. How many authors are involved in this production?
Amazon have said that the total of number of e-books in their catalogue in 2014 was 2m (up from 1.7m in 2013). Therefore the maximum number of authors is 2m. Many authors write a lot more than one book though.
Smash words have said that they have 70,000 authors producing 250,000 titles. This equates to about 3.6 titles per author.
This is the best I can do. It equates to 555,000 authors in the Amazon e-book universe.
Given the new title output of trade publishers in anglo-america this figure seems way too low. But I’m going to pluck a number out of the air and assume that half of these authors are self-publishing. (I’ve assumed that the vastly long tail of trade houses cancels out the productivity of many self-publishing authors, a big guess)
That would equate to 278,000 authors.
If 98% of proceeds go to 2% of the authors, that would mean that the public is paying £860m to Amazon for the work of 5,500 self-published authors.
That equates to an average of £150k in sales from the top 2%, leaving £55 to the other 98%.
If Amazon give 70% of this figure to the author, this means that average income from Amazon revenue is:
Top 2%: $105,000 (£68,000)
Remaining 98%: $40 (£26).
(I can’t think about VAT at this point)
In terms of Anglo-American split (AMZ US 75%; AMZ UK 25%) that equals:
Top 2%: £79,000 (US); £21,000 (UK)
Other 98%: £30 (US); £10 (UK)
Adjusted for market share (AMZ 65% US; 90% UK)
Top 2%: £121,000 (US); £23,000 k = £144,000
Other 98%: £45 (US); £11 (UK) = £56
To my slight surprise, these figures don’t look completely mad, though I suspect the income of the top 2% is materially overstated.
Can there really be 5,500 self-authors in Anglo-American an average of £144,000 in 2014? I suspect the power law of 98/2 is about right, but that the actual income is something like half the stated figure, which would imply many, many more self-published authors than 278,000.
That said, trade houses published 184,000 titles in the UK alone in 2013; the US published it was 305,000 , so the 5,500 figure doesn't look out of kilter from that perspective. It’s also a tiny percentage of the combined adult populations.
Some giant leaps, but the headline figure of £875m makes a kind of intuitive sense to me.
 http://www.newyorker.com/magazine/2014/02/17/cheap-words; and http://www.forbes.com/sites/jeffbercovici/2014/02/10/amazon-vs-book-publ...
Mundy is the founding director of @TMA_agency, representing authors and helping brands publish. TMA sells subsidiary rights for members of the Alliance of Independent Authors (ALLi). He is also to be executive director of the Samuel Johnson Prize. He founder formerly founder and chief executive of Atlantic Books.
Don't forget to weigh in on the survey. And see you in #FutureChat.