The Nameless Horror

Maths: "ebooks are 10% less to produce"

Copying this from @LucaVeste’s Facebook stream, because the typing was lengthy and the numbers interesting. This article suggesting that, for a publisher, an ebook is only ~10% less expensive to produce than a print book has been doing the rounds. You might be forgiven for thinking that this means the sale price of each should only be 10% (or less) different to make the same profits for all concerned, but this isn’t - I think - the case.

Disclaimer: Some 14 years ago, I worked at Gardners Books, the UK’s biggest book wholesalers, doing goods-in in their warehouse, and my wholesale price mark up numbers are based on what they used to be then. As far as I know they haven’t changed, but just so’s you know. (There’s a further disclaimer about publishers doing their own distribution way down at the bottom, along with the likely effect this has on the numbers.) Obviously, I also earn a living writing.

The point to remember when reading this article is this: publishers don’t get 100% of the sale price you and I pay for print books, and 10% less doesn’t mean a 10% RRP drop.

A £10 print book needing to pay off £100k in production costs will probably need to sell not 10,000 copies but somewhere between 20,000 and 27,000 (ish) copies as wholesalers pay publishers 35%-50% of the cover price on average per book, and though some publishers sell direct presumably to avoid this AFAIK the old way is still mostly how it’s done. Of a £7 mass market paperback, the publisher therefore gets £2.50-£3.50. Let’s assume they price to make a small profit on each sale because they’d be in deep financial schtuck if they didn’t, and assume that, at the high print run, mass sale, cheap deal end, a book costs £2 to make all told (even allowing for royalties being based as I recall on sale price, not wholesale price).

In addition to the print and distribution costs, they also need to factor for returns and pulping because of the deeply strange way books are sold, and I don’t know if that’s included in the 10% figure, but let’s run with it. To get the same profit then from an ebook they’d need to sell it to wholesale (Amazon, who I would imagine demand wholesale-equivalent percent-of-RRP since they do with everything else) at £1.80, and then factor their profit margin per unit, for a total £2.30 price sold to wholesale. At “30% of sale price is profit for someone” (depending on how much of a take Amazon wants to make; for agency books it’s set obviously) that’s a £3.30(ish) ebook for you, the consumer, as compared to a £7 mass paperback, or a slightly more expensive (£4-4.50) ebook that wasn’t likely to sell as many (possibly then plus a VAT extra, 20% of price to wholesaler; Amazon’s in Luxembourg and I can’t remember how VAT translates for export) .

The “10% cheaper” thing initially seemed like a £7 book would be £6.30 in digital, but that ignores the usual supply chain mechanics. The actual sale price would be half that of the print version for the publisher to come out the same.

(Of course, Amazon doesn’t need to make the same profit or markup on its ebooks as compared to its physical ones (even though it no longer tries to stock everything in-house) since their wholesaling overheads on digital books amount merely to server space and tech support spread across the hundreds of thousands of ebooks they store, so for any given book, functionally zero cost. They can, with non-agency deals, discount more heavily (comparably) than they can on physical books, driving the (comparable) price downwards if they so want. No doubt then leading to attempts to gouge that £2.30 that goes to publishers. (And boosted by extra income they get for co-op promotional shite and all the other buggering around that publishers have to pay for.))

Now the second disclaimer: We’re assuming my maths is correct and the state of distribution is roughly what it was (Gardners was putting its second warehouse floor on when I worked there 14 years ago; they now cover half an entire industrial estate complete with airport-style skyway with book conveyor belt over a dual carriageway, so I assume they’re not about to go bust unless as a result of burgeoning ebook sales - I could easily be wrong though).

I know that one of my publishers used, or partially used, their own distribution system to sell to stores - because it broke a couple of months before I had a book coming out and one of their other authors threatened them with legal action for lost income (I’m guessing Penguin, therefore; everything went wrong there) - but if there’s still a wholesale trade in play, that’s the mark up, and that mark up goes when you switch to digital. Considering that warehousing and customer order-filling is quite labour intensive, I wouldn’t be too surprised if those publishers who do do their own distribution wouldn’t have similar costs (say another couple of quid a book) because while Gardners have done very well for themselves out of it - largely in the early days by massively exploiting their workforce, true - I can’t imagine it’s a hugely profitable business overall on an individual per-book-sold level, so costs for (say) Penguin would be about the same - less complex throughput but much more stock warehousing.

Edit 19/5 @ 10am: On Facebook, Vincent points out that price can vary enormously according to expected sales because even at reduced earnings, publishers can still make their money back, and also points out the infinite shelf life of ebooks. Both are true.

That said, the first is true (to the extent of a publisher’s cash reserves being able to support it in the meantime) in print as well, and the second is sort of true in so much as in print the vast majority of a book’s sales - barring an unlikely lucky boost from some outside source (a movie gets made, a following develops somewhere else that then feeds back into an originally nothing-doing territory a la Dan Brown, etc.) - happen in the first month to six weeks or so. This is when discovery is at its highest, in stores, from reviews, from advertising, and after the initial surge it drops away quite sharply. While digital doesn’t have quite the same “front of store table display vs. back of store on a shelf” eyeball-attracting variance, the same ‘co-op’ promotional tactics are employed - and paid for by publishers - and I would be hugely surprise if there wasn’t a similar curve to commercial ebook sales. (Not so much perhaps with self-published books in the early days, where the slow clawing-out of a market niche I imagine has, if anything, the reverse.)

The following is hopelessly simplistic (it ignores hardback for a start) as there are many, many differences between the different revenue and cost streams for publishers, but for our purposes, hey…

If a book costs £100,000 to make, edit, print, etc., and anything (minus percentages to third parties) beyond that is profit, and a publisher expects to make £2.00 back per unit in terms of covering their costs, they’re expecting (presumably) to sell 50,000 copies. They crunch some numbers and (I assume) figure they can still get those sales with 50p profit for them thrown in, and thus negotiate to sell to wholesale at £2.50 per unit. The book sells as planned (ha!), some on the longer tail, most in the first six weeks, and earns the publisher £125,000.

There is nothing at all to stop them thinking they can sell not 50,000 print copies @ £2.50 but 500,000 copies @ £0.25. Their total earnings are the same. This applies in print just as it does in digital.

In practice, they don’t; if that book shifts 500,000 copies and earns them a big, fat profit, that profit in part pays for the many other books they print which don’t earn back. There’s an old saw about the movie business that only 1/10 (someone correct me if I’m wrong and it’s 1/100) films earns money for the makers, so they need a few big, reliable blockbusters, as well as the by-luck surprise hits, to cover all the costs of the other things they make. But if they don’t make the other films, they’ll miss out on the 1/10, some of which can be very big indeed, and one flop can ruin them. Book publishing’s not quite as mental as the movie business but I wouldn’t be too surprised if the ratio of profitable:not was similar.

And while they can bank on an infinite shelf life, they can’t put their earning-out point too far down the road because they have ongoing costs and if they don’t continue to meet them they could go out of business before they hit that point. I don’t know where it gets set (and again, hardback income throws our basic little model for a loop anyway), but it won’t be and can’t be very far ahead.

While the price-per-unit could vary a lot according to expected sales, in the old days at least the percentage-of-RRP terms they’d offer to wholesale varied mostly (but not exclusively) between 35% (more niche commercial titles) and 65% (for really big numbers stuff). Assuming those numbers are the same, and I’d be surprised if they’ve changed much, it seems that however they do their maths, a publisher wants to get back ~£2.00-2.50 per successful book, more for smaller ones, in order to feel happy. That’s why I based my figures on that as a “cost”.

Assumptions, assumptions. As they say in Monty Python & The Holy Grail, it’s only a model.

KDP: licensed to ill

While we’re talking Amazon, and while Apple and publishers are locked in somewhat ludicrous court battle over agency pricing and collusion (as if they needed to collude to agree that Apple’s long-standing standard practice was more attractive than Amazon’s shit-on-a-stick; that’s another argument for another day), let’s get legal for a moment.

This isn’t going to turn into an Apple vs. Amazon post - really - but I need to mention this for context. A while ago, when iBooks Author was released, half the internet briefly took up arms against perceived wrongs in the iBooks licensing agreement, believing, wrongly, that it enforced a single-point-of-sale rule on books listed through iBooks. Apple quickly amended the license to emphasise that the restriction applied only to .ibooks files created with iBA, and the controversy died.

But what of Amazon’s license agreement for Kindle Direct Publishing? There certainly hasn’t been the same level of uproar over it - indeed, most KDP authors seem happy with their terms. Odd, this, because it’s a very strange beast.

Most of us know about the “Matching Competitor Prices” rule in the KDP pricing terms. The rule has changed somewhat over the past year; it used to be a rule in the license itself that you guaranteed not to list something for sale cheaper anywhere else, and if Amazon found you had they could drop the price to beat it.

(As per this similar dissection of the agreement, the old version of Section 4 was:

You must set your Digital Book’s List Price (and change it from time-to-time if necessary) so that it is no higher than the list price in any sales channel for any digital or physical edition of the Digital Book.

But if you choose the 70% Royalty Option, you must further set and adjust your List Price so that it is at least 20% below the list price in any sales channel for any physical edition of the Digital Book.

That clause has since been dropped.)

It all happens now in the new section 5.3.2:

To the extent permissible under applicable local laws, we have sole and complete discretion to set the retail price at which your Digital Books are sold through the Program.

You provide a list price, but if Amazon need to sell cheaper “to match a third party’s sales price for any digital or physical edition of the Digital Book” they can do so. (Whether or not this affects your royalties depends on whether it was a 35% (it doesn’t) or 70% (it does) royalty book.)

So what exactly is a “Digital Book” as defined by this license? Apple very neatly confined their more draconian requirements to the .ibooks file created by their software. Is this something similar? (If you’ve never used KDP, you upload a .mobi or .epub file.) What’s Amazon’s take on this?

Badly defined.

This Agreement is a binding agreement … with respect to your participation in the Program and your distribution of digital content through the Program (all such content, “Digital Books”).

Content. But what, exactly, is the content by which a book is defined? If you sell a short story through Amazon for $0.99 (or for free via Amazon Select), and then include that same story in a collection at a higher price, does this give them leave to match the collection to the lower price? (Not that they need your permission, as the agreement makes clear, but they pay lip service to justifying such decisions in terms of competition with either other parties or other national versions of Amazon.)

(While we’re at it on content, let’s not get into the business of their actual content guidelines, which make the recent Smashwords erotica kerfuffle look tame. (And again I’d point you to the third part of the ePublish a Book article for why it’s especially bad.)

To use an analogy, if you sold a regular, a Blu-Ray, and a collector’s super-special director’s cut special edition of a movie under the same terms, is the content the same? The latter two certainly share elements of the same content, but the Blu-Ray “contains” massively more data (which, if a film is a series of images means a lot more “movie”) and the special edition has all sorts of extra material. In both cases the content of what’s on the disc is very different. Each, one might say, should be a different Digital Book under the license terms. What constitutes an “edition” of content isn’t defined.

(They might argue that both were different editions of the same thing, and with at least some justification - but my point is that the definitions they use are missing or totally inadequate. And open to abuse by Amazon, while if they monkey with your pricing it’s your earnings that take the hit.)

So far, so economics, and maybe I’m making too much of it. It’s not the weirdest part of the agreement, though. That’s this:

7 Confidentiality. You will not, without our express, prior written permission: (a) issue any press release or make any other public disclosures regarding this Agreement or its terms; … [a lot of stuff about Amazon Confidential Information which means, apart from anything else, anyone discussing their sales figures without prior written consent from Amazon is technically in breach of contract] … Without limiting the survivability of any other provision of this Agreement, this Section 7 will survive three (3) years following the termination of this Agreement.

This agreement, as previous linked, is a publicly available document. A publicly available document that its signatories are forbidden from discussing under what is effectively a lengthy NDA or face termination of their account (or, in theory if the expense was justified, legal action) if Amazon object at all because they’re breaching their contract.


I’ve never had anything like that in a contract I’ve signed with a publisher. There’s no such clause in the contentious iBooks Author license. Quite why it’s in here in the form it is… well, I’m baffled. I’m also completely unable to think of any happy, cuddly reason to have it there in any form at all.

(Bear in mind that this is an agreement which - in theory at least (I doubt it’d survive a court challenge) - can be changed for anything other than rights & royalties at any time, effective immediately, without Amazon needing to notify you directly; the onus is on the user to check the terms haven’t changed, and continual use of the program signifies acceptance of all changes whether you’ve noticed them or not.)

This is the glorious author-friendly, reader-friendly, digital future. Hooray.

By foolishly insisting on DRM, and then selling to Amazon on a wholesale basis, the publishers handed Amazon a monopoly on their customers—and thereby empowered a predatory monopsony.
The always excellent Stross on Amazon. If you haven’t read this, you should.

The same cannot be said of society. In the UK we are all, as subjects of Her Majesty the Queen, bound by a legal framework out of which we cannot opt.* Thus where there should be proper equality is in the structure of legal marriage. If there is no difference between a civil partnership and a registry marriage except in the gender of the participants, then there really should be no need for two separate legal entities. This, I believe, is where the government is trying to get to with its recent consultation paper that has stoked up so much controversy.

As for religious marriage? Well, as far as I am concerned, it should have no legal significance whatsoever. A church wedding is a matter between two people and their god - and no-one else. Even the witnesses are really only there for the free booze. After your bridesmaids-and-confetti extravaganza, if you want to enjoy the different rights our society affords to couples over single people, then you should hie yourself down to the registry office, where you can be officially recognised as a couple regardless of the contents of your respective trousers.

Mr Oswald is eminently sensible.