After I recommended that the major publishers drop mandatory DRM from their ebook products, I realized that my essay had elided a bunch of steps in my thinking, and needed to reconsider some points. Then I realized that it’s not a simple, straightforward argument to make. Consequently, I ended up writing another essay, although I’ve tried to summarize my conclusions below.

First, my conclusions:

1. The rapid current pace of change in the electronic publishing sector is driven by the consumer electronics and internet industry. It’s impossible to make long term publishing plans (3-10 years) without understanding these other industries and the priorities of their players. It is important to note that the CE industry relies on selling consumers new gadgets every 1-3 years. And it is through their gadgets that readers experience the books we sell them. Where is the CE industry taking us?

2. Dropping DRM across all of Macmillans products will not have immediate, global, positive effects on revenue in the same way that introducing the agency model did …

3. However, relaxing the requirement for DRM across some of Macmillans brands will have very positive public relations consequences among certain customer demographics, notably genre readers who buy large numbers of books (and who, while a minority in absolute numbers, are a disproportionate source of support for the midlist).

4. Longer term, removing the requirement for DRM will lower the barrier to entry in ebook retail, allowing smaller retailers (such as Powells) to compete effectively with the current major incumbents. This will encourage diversity in the retail sector, force the current incumbents to interoperate with other supply sources (or face an exodus of consumers), and undermine the tendency towards oligopoly. This will, in the long term, undermine the leverage the large vendors currently have in negotiating discount terms with publishers while improving the state of midlist sales.

Now the details:

1. Anticipating the future of ebook reading technology

(Background note: I have a computer science degree from 1991, which is a bit like having an aeronautical engineering degree from 1927. But I’ve spent a chunk of time working as a computer journalist, and I try to keep up to date.)

First, a note on the changing technology. The consumer electronics industry relies on selling everybody shiny new devices every 12-24 months for their revenue. Margins are narrow and R&D costs are high. They also have an interest in maintaining a floor under the price of their products by adding new features to justify the upgrade treadmill, because thanks to Moore’s Law, the electronics sector is trapped in a permanent deflationary cycle. So I believe that any forward-looking publishing strategy needs to consider the impact of this endless device churn on consumers, and their likely response.

Because the devices our consumers own mediate their experience of the ebooks we sell them.

First, the hardware:

It’s my belief that today’s e-ink ebook readers are doomed to obsolescence within a short period — 2-3 years possibly, 5 years probably. This is because the power consumption of LCD displays is dropping and their quality is rising. e-ink devices are inherently incapable of displaying video, are lousy as web browsers due to the screen refresh time, and if you use them to play audio or do any intensive processing (such as running apps) their battery life drops towards that of a regular LCD-equipped tablet. They’re essentially single-purpose devices, competing in a market with general-purpose devices. Their only advantages are battery life and readability in direct sunlight, both of which are under threat. So it’s my belief that general purpose tablets (and big-screen smartphones) will drive e-ink readers out of the mass market within 2-5 years, just as smartphones killed off your 2003 Palm Pilot.

Secondly, the software:

The two current tablet/smartphone market incumbents are iOS (Apple) and Android (Google). (Microsoft is making a come-back attempt with Windows 8 Mobile, but is fighting an uphill battle.) These are essentially competing software platforms, like MacOS and Windows in the late 1980s. However, just five years ago, none of these platforms existed; the market was dominated by PalmOS, Symbian, and Microsoft’s dead PocketPC platform. I therefore conclude that it is a really bad idea to make assumptions about the devices customers will own in even 3 years’ time.

In the tablet/smartphone world, DRM is supported at the application level. B&N (Nook), Amazon (Kindle), and Adobe all provide readers that run on incompatible DRM standards. Even when the file format is the same (ePub) the DRM prevents files from, say, the Adobe Digital Editions system from being read by a rival’s reader.

In the absence of DRM it is trivially easy to convert ebooks between file formats — as easy as opening a word processor file on a different machine, if not easier. Amazon’s continued use of a non-epub file format on the Kindle platform does not mean that Amazon could not, very rapidly, shift to supporting epub files; all that would be needed would be a software update pushed to their Kindle customers’ readers. In fact, Amazon acquired a software company specializing in epub reader software — Lexcycle — in 2009.

The main effect of DRM, from a platform vendor’s perspective, is to lock end-users into their platform in perpetuity. (Amazon, as both a retailer and a platform vendor, has leveraged this very effectively to give their retail channel a whip hand.)

2. Which sectors will respond positively to less DRM

Macmillan sells a variety of products (trade and mass market, audio, ebooks and paper books) into a variety of wholesale and retail channels, who in turn sell the products to the reading public.

The reading public is not a monolith, and the products Macmillan sells are dissimilar. Some books are unique and non-interchangeable, while others are treated as an undifferentiated commodity by their consumers. One large customer segment buy 1-5 books a year, usually bestsellers for recreational vacation reading. At the opposite end of the scale, 20% or fewer buy 20-150 books a year, typically midlist titles. The former group supplies mass sales, but the latter group supports the midlist and supplies diversity. A one-size-fits-all approach to the reading public is therefore unlikely to satisfy everyone.

For example, the 1-5 bestsellers-a-year people: previously they bought from airport bookstores and WalMart, to read once then discard; we expect them in future to buy ebooks to read once then discard. They will probably use a work-issue tablet or smartphone running a free Kindle or Nook app, rather than buying a special-purpose e-reader, and delete their books after reading. They couldn’t care less about DRM. They will probably stick to one well-known online retail supplier. You are absolutely right about there being no benefit from dropping DRM in this sector.

3. Who gains? And why?

The voracious 20-150 books/year readers are a small but significant market segment.

These people buy lots of titles. They frequently have specialized interests which they pursue in depth, and a large number of authors who, although not prominent, they will buy everything by. They frequently re-read books, and they are disproportionately influential on other customers because they enthuse about what they’ve read. They’re particularly common in genre fiction. Previously they bought paperbacks and hardcovers from specialist genre bookstores or, failing that, from large B&N/Borders branches. They will go to whatever retailer they can find online, and they find DRM a royal pain in the ass — indeed, a deterrent to buying ebooks at all.

There is a pervasive assumption that ebooks are disposable literature. But to the voracious readers, this is not the case. Currently it’s hard for many people to build up collections of books due to space constraints — nevertheless I know many SF fans (of the kind who read 50-150 books a year) who have turned their homes into libraries. They will be the tip of an iceberg once ebooks become mainstream; why discard an ebook when you can file it and come back to it in 10 years’ time and it takes up no space?

For such people, filing and tagging their collections is a major issue. And so is portability. It’s true that if they own an iPad they can have an iBooks app full of books purchased from Apple, and a Kindle app full of books from Amazon, and a Nook app full of books from B&N. But those apps are, thanks to DRM, data silos — you can’t cross-check to see if you bought book 3 in a series from Apple and book 5 from Amazon without a lot of fiddling around.

Platforms age and die. This summer, Microsoft is turning off the DRM servers for Microsoft Reader. This means that people who bought Microsoft Reader ebooks over the decade since 2002 now find that their ebooks are trapped inside a rapidly ageing, obsolescent slab of plastic and glass. In another 5-10 years, 95% of those books will be unreadable because the machines they’re locked into were designed by a CE industry obsessed with the 2-3 year upgrade cycle — they’re not durable. This is actually one psychological driver for piracy — people who have paid for a book resent being expected to pay for it again due to an arbitrary-seeming lock-in onto an aging piece of hardware. From their point of view, honesty is being punished.

There is no guarantee that B&N will stay in business, or that Amazon won’t discontinue support for older Kindle files, in the not too distant future. This is something that the hardcore readers cannot help but be aware of, because it has already bitten them in the past, if they bought a Zune, or a Palm Pilot, or any number of other devices.

If Macmillan drop DRM on ebooks typically bought by these people, it sends a signal: “you can continue to read these ebooks in future using whatever platform you want”. Converting a DRM-free ebook between ePub and Amazon’s Kindle format, or any other current ebook format, is as easy as saving a Microsoft Word document in Rich Text Format, or as a web page: there’s an app for that. Moreover, all the DRM-equipped reading platforms support importing non-DRM’d ebooks.

So, from the point of view of a particular subset of Macmillan’s customers — the hard core genre audience who read many, many books — removing DRM would be a major benefit and would probably generate immense goodwill.

(This is leaving aside the point that, if a trend towards relaxing DRM becomes established — as happened in 2009 in the music download industry — the first mover will reap considerable public relations benefits and news coverage in the short term.)

But is there a business case for doing so?

4. Effects of removing DRM on the supply chain

(Firstly, I’d like to note that the Macmillan experience with dropping the mandatory requirement for DRM on audio books can’t be taken as a useful indicator. The main retailers of audio books, Apple and Audible, refuse to ship DRM-free audio books. Therefore DRM-free audio books remained essentially unavailable to the public.)

The main effect of DRM on the supply chain is that a consumer who buys DRM-locked content is locked into the supplier who supports that type of DRM. A non-casual reader with a couple of hundred ebooks on their Kindle can’t easily leave the Kindle walled garden. (I emphasize, DRM is the only thing that keeps them there: converting Kindle ebooks to ePub is trivially easy in the absence of DRM.)

This situation plays to the benefit of the largest incumbents in the retail sector. Currently we have gone from a near-monopoly by Amazon to a near-cartel among Amazon, B&N, and Apple. The independent bookseller sector is struggling to deal with ebooks.

It’s instructive to take a look at how the independent retailers are failing to cope. Powell’s have a large online store, and are quite successful with paper products. However, if you want to buy ebooks from them, they offer you a menu of DRM silos — Adobe Digital Editions, Google Ebooks, and so on. If you want to buy ebooks from Powell’s, you have to grapple with registering your device with them, so that the ebooks can be locked to your reader. This forces customers to jump through a bunch of technological flaming hoops; it’s easier for them to give up and point their web browsers at Amazon or B&N instead. And the results have been so poor that Google seems to be withdrawing from the retail market, at least to the extent of giving notice to quit to their larger retail affiliates (Powell’s included).

If Macmillan titles did not have DRM, then customers would find it much easier to buy books from independent retailers like Powell’s — or other small bookstores. DRM-free ebooks can be imported easily into whatever ebook reading device a reader already possesses. It will then be possible for bricks and mortar retailers and small online retailers to get a toe in the door and sell ebooks competitively. In short, it will lower barriers to entry into the retail supply chain, which in the long term is advantageous to publishers.

Another angle is that dropping DRM gives readers some assurance that their ebooks will remain accessible, even if they change reading devices and apps multiple times over the next decade. It also allows them to merge ebooks from different sources into a single collection, simplifying their reading experience, and to confidently purchase from smaller retail outlets.

As noted earlier, consumers change e-reader devices frequently. Within 5 years we will be seeing a radically different electronic landscape. Unlocking the readers’ book collections will force Amazon and B&N and their future competitors to support migration (if they want to compete for each others’ customers). So hopefully it will promote the transition from the near-monopoly we had before the agency model, via the oligopoly we have today, to a truly competitive retail market that also supports midlist sales.

(Why this will support the midlist: currently Amazon have swamped the midlist among ebooks in a sea of self-published rubbish. It’s impossible to find anything worth reading in the Kindle store that isn’t a very obvious bestseller. This offers an opportunity for specialist bookstores to offer a curatorial role. I believe the voracious genre consumers are picky enough about what they read that they dislike Amazon’s slushpile approach, and will preferentially shop in better organized outlets.)

Other thoughts

I don’t expect dropping mandatory DRM to have an immediate positive impact on sales. However, it will permit small retailers to compete and specialize in a market they are currently locked out of by network externalities. Right now, there is a window of opportunity for smaller resellers: Amazon’s inclusion of masses of self-published material in the Kindle store has made it impossible for heavy consumers to browse it effectively. Smaller bookstores may be able to gain a strategic edge by curating their content, providing quality control on reviews, and other tactics we can’t predict at this time. This is, I emphasize, speculative — but I believe saving the smaller resellers is key to diversity in the retail side of the market, and will further support the midlist (which is threatened right now by plummeting mass market sales and the difficulty authors experience in reaching their audience).

To the extent that piracy is an issue, I think the horse is well and truly out of the stable and over the horizon; bolting the stable door and adding chains and padlocks hasn’t worked to date, either in print publishing or in music and film publishing. However, I would recommend considering a switch to watermarking. Watermarking doesn’t prevent copying, but makes the original source of a copied file easy to find, which is a deterrent to piracy. This appears to be the current best practice in the music industry (in the iTunes store, all music downloads are watermarked), and they’re a few years further into the era of internet distribution than we are.

Dropping DRM is probably not going to have a significant effect on the bestsellers, but I will note that J. K. Rowling’s move into ebook territory is DRM free; presumably the rampant levels of piracy around her work was seen as a pre-existing condition, and anything that might convert pirate readers into paying customers was seen as giving Pottermore an edge.

Finally, if going DRM-free is a trend, it may be to Macmillan’s advantage to be seen to be a front-runner. Removing the requirement for DRM from specialist imprints marketing primarily to the voracious genre readers would be a useful experimental step: I will confess to a personal bias here, but I’d love it if Tor was allowed to sell my novels unencumbered by DRM — I could personally use that as a strong marketing angle. (Like many younger writers, my major point of contact with my readers is my blog — I typically get 12-14,000 readers per day, and provide them with a community for discussing my work and asking me questions — based on direct feedback I’m fairly certain that dropping DRM would allow me to generate additional ebook sales and point my readers at a more diverse range of retailers.)