Web dev at the end of the world, from Hveragerði, Iceland

Today is not tomorrow (or, how to beat Amazon)

The world doesn’t look the way it should. From a publishing perspective, the ebook platforms are too fragmented, ebook creation and design is too complicated, expensive, and limited, and sales are too dominated by a single platform that has a history of being aggressive towards its suppliers.

And readers are getting screwed. Their notes and highlights are treated like proprietary information, owned and locked up by the platform owners. Publishers seem to be colluding to raise prices. Books are often riddled with errors, even those from big publishers (see point about ebook creation being too complicated above). The books they buy are locked into a single platform.

Amazon is well on its way towards dominating the ebook market, but its platform has several weaknesses that not only threaten Amazon but also threaten the ebook industry should Amazon dominate.

The line of argument is as follows:

  1. Amazon’s Kindle platform has inherent flaws that lead to serious imbalances.
  2. Those imbalances threaten Amazon first and foremost and grow with the platform.
  3. If Amazon grows to dominate a mature ebook market, those imbalances will be big enough to both damage Amazon and threaten the ebook market.
  4. Amazon could end up dominating an ebook market that more resembles the modern day comic book direct market than a mass medium: a shrinking niche industry that caters to a limited number of expert readers.

I like Amazon. Without them, there wouldn’t be an ebook market. I’ve been a customer of theirs for as long as I’ve been online. But I honestly think that the path they are on threatens both them and the ebook market on the whole.

So, despite the linkbait title I don’t intend this post to be an anti-Amazon screed, but a vision of a more dynamic ebook market.

(B&N and Kobo, on the other hand, are only threatening themselves. I can’t understand their somewhat hostile attitude towards ebook design and ebook developers, and I completely fail to understand how anybody would think that their past behaviour was in any way a good idea.)

Amazon’s weaknesses

  1. No margins (mostly due to hardware)
  2. No moat
  3. Subsidised hardware
  4. Integrated silo
  5. Specialised user-base (expert readers)


Even investors who are bullish on Amazon think that their margins are getting much too tight. And, if you look at the company with a critical eye, the picture looks even more negative.

If we concentrate on ebooks, after the DOJ settlement, Amazon is entering a world where they have no margin on their device sales (in fact, if you tally engineering, sales, and development costs, its hard to see how they aren’t directly subsidised) and their margin on ebook sales, slim as it has been, will contract as they are likely to offer the biggest discounts on their biggest sellers (bestsellers).

(There is an argument to be made for discounting the long tail instead of the head, but I won’t get into that here.)

Amazon’s problem is that investors won’t tolerate margin shrinkage in the long term and their stock price is massively overvalued as it is. If this continues for a few more years, Amazon will end up in a place where a single negative incident or a few bad quarters could result in a massive share price collapse.

Which in turn would force Amazon to try and expand their margins.

The problem is that there isn’t much scope for margin expansion in the Kindle platform as it’s designed. The hardware’s main selling point is being cheap. The ebooks are almost inevitably going to be sold under the agency system, long-term, leaving Amazon with a limited ability to control prices. And ebook prices are being driven down by an abundant supply of small publishers, driving ebook retailer margins down with them.

Raise the price of the hardware and superior general purpose pieces of kit like phones with big high pixel density screens and iPads become much more cost competitive.

The only way publishers can reasonably be squeezed on price is by piling additional and spurious costs onto their distribution deals, essentially trying to blackmail them into give Amazon more money.

Which nets them chicken feed and motivates the entire industry to act against them (taking care not to collude this time).

Or they can try and become the exclusive publisher or distributor to as many books as possible and then squeeze the self-publishers dry.

Which is equally self-destructive.

I believe Amazon’s current path leads them to either destroy their own standing in the market by alienating their suppliers or to cannibalise the entire ebook market once their shareholders have turned on them.


A moat in this context are structures and tactics that prevent customers from abandoning the platform.

Amazon doesn’t have one. Not really. DRM does give it the power to keep books locked to their platform but the only way that translates into user lock-in is if Amazon Kindle devices outsell retina display iPads and high pixel density mobile phones.

Reading books is so much nicer on the new iPad than on any Kindle. In six months to a year, Apple will have sold more of the new iPad than Amazon has ever sold Kindles. High pixel density phones (iPhones, premium Android phones) outnumber Kindles several times over and make up for their marginally worse reading experience (which is debatable) with vastly superior portability.

These competing devices are better hardware and better platforms. The screens are better. The OSes are newer and superior. The App Store selection is much richer (neither Google nor Apple lock their ebook competitors out of their app stores).

And they all have a Kindle app.

So, when a user decides to switch to a general purpose device for their reading, they don’t need to move their library off from the Kindle platform. It’s always there in the app. They can switch from ereader app to ereader app and won’t really experience any lock-in because all of the books are accessible on their devices. Only the most anal of pedants will see the inability to move their library from app to app as a deal-breaker.

And, of course, Amazon doesn’t control the ebooks it sells, not really, and they have, through their actions, motivated publishers to take desperate actions (collusion, etc.). If publishers give up on their own DRM requirement, or even begin to demand no DRM, what little remains of Amazon’s protective moat around their users will evaporate into the wind.

Note to Amazon: pissing off your suppliers before you have a lock on being a monopoly or monopsony is extremely risky.

Not having a moat and not controlling prices in the long term means that Amazon needs to compete on customer support and publisher services. They can’t compete on hardware. They can’t lock their customers to their platform. They don’t control prices, not in the long term. Amazon risks the future viability of its ebook business with its current tactics.

Aside: mobi isn’t a defensible strategy, long-term

Once both Amazon and their EPUB-based competitors have both migrated towards a fully web-based stack for their formats (proper support for HTML and CSS), the similarities between the formats mean the defensive advantages of mobi are limited.

But Amazon can’t avoid transitioning to open web-based technologies.

Amazon’s behaviour means that most publishers have a vested interest in not giving them any advantage. That means that new format features will only be used if either the EPUB platforms match them or if they are too trivial to affect competitiveness.

And even if Amazon could add features that were so desirable that publishers would jettison their qualms about competitiveness, they would still have to be implemented in a way that is based on web standards, which in turn means that EPUB can implement them, which means they aren’t a defensible advantage.

KF8 will, over time, become de facto identical to EPUB because the non-EPUB features just won’t be used by publishers or authoring tools.

Long-term, Amazon is best served by switching to EPUB and competing within the format in the same way that browser vendors compete (vendor extensions, etc.).

Subsidised hardware

I’ve said several times before that the no-margin hardware as an ebook distribution strategy is extremely risky. It’s insane for Kobo and B&N, but it’s risky even for Amazon.

This strategy:

  • Forces them to compete with firms who both make a profit on selling hardware and so can invest more in R&D.
  • Increases their software development costs because they now have to support and develop a much larger set of apps and OSes than if they just focused on software clients.
  • Limits their flexibility and ability to execute. Software-only competitors have to worry about fewer OSes, fewer platforms, and fewer apps.
  • Increases their support costs because doing hardware is tricky and their no margin, low cost, strategy gives them an incentive to use cheap components.
  • Lowers their margin even more because electronics are expensive to ship. They aren’t making money on the devices and probably losing money on fulfilment.

A software-only competitor would be much nimbler and be able to match Amazon’s buying and reading experience at much lower costs. Prices would remain the same because after a couple of years we are likely to be back in the agency system. As a result that future competitor would be able to do more than Amazon and still make more money that can be ploughed back into growth, research, and development. That competitor’s user-base would easily be as large as Amazon’s because, at that time, the number of people with highly readable phones and tablets will dramatically outnumber the number of Kindle devices in existence (even more so than today).

Amazon’s gamble is that they can convince people to buy a Kindle as well as a phone and a tablet. The only other way Amazon can compete is if they can deliver a tablet of their own that both outperforms its competition and is cheaper.

(It has to outperform its competition to make up for the limitations Amazon puts on the app market by locking its competitors out.)

Even if Amazon does manage to release and keep releasing high quality tablets, they will still never attain the majority in phones and tablets that they have in bespoke ereader devices. The inevitable dominance of general purpose hardware will even the playing field and strengthen software-only competitors.

Integrated silo

Another one of Amazon’s weaknesses is vertical integration (something that is commonly thought of as a strength).

You don’t compete with an integrated product strategy with another integrated product (like Palm did when it tried to compete with Apple with WebOS). You compete with a modular ecosystem, which is the reason why, despite the criticism you might levy at the user experience, Android has managed to compete with iOS.

A modular ecosystem is weak where the integrated product is strong, but it’s strong where the integrated is weak. It is also more flexible and much more likely to target non-consumption, grow into new segments, and foster variety.

Every decision Amazon makes about its bookmarking, notes, and highlights system has to consider a diverse set of hardware clients in addition to their software clients.

A format change involves a half a dozen app and OS updates instead of just a couple of submissions to a couple of app stores.

Any proposal for change needs to consider whether their integrated hardware can even support it and when to cut off older generations of hardware and OSes. A software-only competitor has some of the same problems, true, but to a much more limited extent because they aren’t supporting multiple generations of bespoke devices.

The integrated silo is both expensive to maintain and slows Amazon down. If its competitors can deliver an experience that is as good using a modular platform, they will be rewarded with lower costs and much greater flexibility.

A specialised user-base

Today, Amazon owns the expert reader. These are the customers that deliver a large proportion of the publishing industry’s revenue and profits, but they aren’t the source of growth.

The analogy I’ve been using is the comic book’s direct market. This enabled comics publishers to cater to their expert readers, those with an intricate knowledge about past continuity and a burning passion for the medium.

But, nobody else went into comic book stores. Kids don’t grow up seeing comics on the newsstands or in bookstores anymore. Comics ceased in the late eighties, early nineties, to have a presence in our public spaces and renewal of the reader-base halted.

Publishers discovered new ways of squeezing money out of their regulars, but put no thought to where future customers would come from.

So, today we have a situation where bestselling comics sell an order of magnitude fewer comics than bestsellers did in the nineties (100 000 copies versus a million versus millions, plural, in the eighties/seventies/sixties).

(We are lucky in Iceland to have the Nexus store which has a much greater selection and is much more accessible than most comic book stores, but they are the exception rather than the rule.)

Novels could easily fall into the same trap. Publishers are raising prices to libraries and, in the UK, cutbacks are threatening their very existence. Bookstores are disappearing from malls, retail parks, and the high street. Retailers like WH Smiths focus more on stationary and general goods than they do books, and what few books they have are dominated by celebrity bios and churned non-fiction. Like comics before, novels are slowly disappearing from the public sphere.

But, having captured the expert reader populace, Amazon is in the position to squeeze the market dry as it slowly fades away over the next twenty years. They won’t have the margins to expand novel reading and the Kindle is in many ways as unfriendly to casual readers as a comic book store is.

The Kindle as a device is a shibboleth for expert readers and as such drives casual readers away. ‘Why should I buy a Kindle? I only read a couple of books a year.’ It’s a symbol for a clique they know they won’t ever be a member of.

A free app, on the other hand, is a no-brainer decision for most people. They’ll download it, just in case.

But none of that addresses the problem of renewal once novels are removed from the public sphere. Comics have taught us that you can’t rely on parents training their children to love a medium, it needs to be instilled through exposure. The industry needs strong, healthy, and vibrant libraries. Ask any adult expert reader and they’ll all rave about how much time they spent in libraries as kids.

(Forcing kids to read dusty tomes in school is also probably the most effective way of turning them off reading for life. If that’s all we have left then novels as a medium has only a couple of decades left before it fades into obscurity.)

What can publishers and ebook retailers do?

  1. Focus on general purpose hardware. Only four platforms matter: iOS, Android, Windows desktop, and Mac OS X. If you want to beat Amazon, you need complete iPad support. Pick one enemy, don’t take on both Amazon and Apple.
  2. Modularity. Separate retail and ereader apps. Enable easy transfer of ebook libraries. Avoid DRM because it directly counteracts modularity. Offer generic notes and highlights services as APIs. Use the Open Publication Distribution System to let the reader buy an ebook from any retailer and have it automatically delivered to any ereader app.
  3. Target non-consumption. Shorter books. Fragmented books. New kinds of texts. New forms. Experiment and innovate in ways that Amazon can’t. They are locked in through their integrated silo and can’t move on new forms.

General purpose hardware

A firm whose only purpose is to make hardware is much more likely, in the long run, to be successful at it than a firm who sees hardware as a cost centre.

An app-only ereader doesn’t need to chase hardware trends like the current drive to retina displays and can focus their energies on just the client.

The idea behind Readium, collaborating on an open-source reading system, is exactly the right path to take.

(Readium is an open source project that unfortunately has closed source governance at the moment. See for example this comment on a pull request I made. Debates and discussions on Readium’s direction are taking place in private emails, not in public like they should be in an open source project. Readium needs a public mailing list where all of the decisions can be discussed and archived.)


Separate all concerns. Make it so that highlights/notes, the rendering engine, and sales and fulfilment can be separate services run by separate organisations.

A reader should be able to use Readmill’s social bookmarking with the Bluefire Reader and be able to buy from Kobo or B&N and have the book automatically download in any reading app (Bluefire, Aldiko, Readmill, whatever). Each of these organisations will find it hard to compete with Amazon if they are forced by a publisher’s requirement of DRM to implement an integrated system, but Readmill certainly has the capabilities to create a vastly superior social highlighting/note-taking experience, Bluefire has the talent to create an excellent reader, and Kobo and B&N know how to do retail.

Drop DRM. Demand that retailers implement open standards for distribution and open APIs for notes and highlights. And an Adobe monopoly on DRM is just as bad and destructive as an Amazon monopoly on ebook retail. An Adobe tax on every sale is in many ways even worse than an Amazon monopoly.

All of the building blocks are here, right now:

The only thing publishers need to do to enable this is to give up on DRM.

Then they can go collaborate with various vendors on implementing these open standards for distribution, fulfilment, and formats.

Target non-consumption

This is the one that seems the trickiest, but really isn’t. In a way, it’s likely to emerge naturally once a more open and modular ebook marketplace has been implemented.

But the basics are simple: Try new stuff. The economics of ebooks allow for a much greater variety of both form and content and we don’t know if there is a market for them yet.

This is an extremely lucrative tactic. Discovering an an entirely new and untapped market is a recipe for high margins and disgusting profits.

Experiment. Experiment. Experiment. Try new distribution venues. New tactics. New ideas. Work with new people. Try different things.

This is the very definition of an asymmetric bet. It doesn’t cost you much to play, but the rewards have the potential for being huge.


Don’t take action to strengthen Amazon’s existing competition. Create an ecosystem that is open for innovators who can grow fast. Make it really easy to become a reseller of your books. Create standard agreements and standard distribution platforms. Make it so that a small ebook retail startup can be up and running with a full catalogue with only a couple of days of work. That way they can focus on selling, the ereader, the ecommerce platform and not negotiating with you.

Set your distribution system up as an API. Make it your goal that a new ebook retailer should be able to download a standard agreement, set up the financial side of the distribution deal, and get access to a full, but secure, OPDS catalogue of your available ebooks in less than an hour.

You don’t want half a dozen ebook retailers competing with Amazon. You want thousands.

Agency pricing is not a bad idea in the long run, as long as it doesn’t harm retailers and distributors. That means no ‘Most Favoured Nation’ clauses.

With the right infrastructure in place, a group of specialist firms (retailers, ereader apps, distributors, publishers) will beat the tar out of the integrated silos.

Consider licensing your back catalogue to new e-only publishers, but do so only in big, genre-specific, batches. Give them an incentive to find value in titles that haven’t been given any attention in years and it gets you revenue from titles that weren’t earning.

Kobo and B&N are in for a very rough time

I think they’re toast. Amazon’s kamikaze margin dive is likely to take down any firm that is willing to dive with them.

Pay a lot of attention to Amazon’s next earnings report because it’s only going to get worse from there.

What should Amazon do? (Or, how not to get beaten.)

Amazon will be disrupted by a modular ebook supply chain, eventually. If its competitors are smart, it’ll happen this year. If they aren’t, it’ll still happen, just a few years later, when bespoke hardware like the Kindles have largely been driven out of the market.

I strongly think that this is inevitable. My only worry is whether there will be an ebook market left by the time it happens. That’s why I want it to happen sooner, rather than later.

Correcting for its weaknesses should be one of Amazon’s first priorities.

  1. Phase out the hardware funnel. It’s an incredibly expensive customer acquisition strategy with limited lock-in.
  2. Become the best place to sell books. Build out detailed publisher tools and give them access to detailed analytics. A publisher should be able to calculate their customer acquisition cost, per acquisition strategy, per funnel, per promotional campaign, from the statistics you give them. They should also be able to calculate their average annual revenue per reader from the data you give them. Bonus points for offering them ways to A/B test their marketing and sales tactics and materials.
  3. Make it really easy to make beautiful books. Something like a Kindle Themes service where people can upload their epubs and Word documents, pick a beautiful design, and download a ready mobi file. Make it so easy to use and the designs so good that people will be stripping the DRM off their epubs just to run them through the service and load onto their Kindles.
  4. Modularise the platform before somebody else does it. Open up Kindle highlights and notes through an API. Sell epub3 files alongside mobi files. Add epub3 support to the clients. Support the Open Publication Distribution System both as a retailer and in the client apps. Modularity goes both ways, enables people to join your platform as well as leave it.
  5. Compete on customer service and publisher service because the platform lock-in is an illusion and the no-margin strategy today will not lead to a high margin strategy in the future. The industry will implode much like the comic book industry did if you try.

I like Amazon. I’m a huge fan of its products and have always had a pleasant experience with their customer service, even when things have gone horribly wrong, they always came through. I love how they jump-started the ebook market from nothing, something they will never get enough credit for.

But, I sincerely believe that they must change their tactics before they do serious harm to the publishing industry and to themselves.

ETA: Mike Cane reminded me that without Sony’s pioneering work on ebook readers, Amazon would never have done the Kindle. So, even though Amazon were the first to jumpstart the growth of the ebook market, Sony was there before them.

You can also find me on Mastodon and Bluesky