Publisher or Retailer?

by Tim O'Reilly
January 2000

Lots of good responses to comment on here, so I thought I'd do it all in one message.

Rodent of Unusual Size <Ken.Coar@Golux.Com> wrote:

Linda Walsh wrote:

Your question implies the belief that the book author should receive a higher royalty percentage, as self-published eMatter authors do.

That's right, I guess I do. It seems to me that the ebook-specific costs are significantly reduced over the dead-trees version. One-time layout and 'print' run, no warehousing or shipping costs, no ink/paper/materials cost.

I think most people really overestimate the cost savings, and underestimate the cost increases. First off, manufacturing (including damaged returns) works out to 16% of net revenues for O'Reilly. Warehousing adds a few points more. Shipping we pass on to the customer. So assuming it all adds up to 20% of costs, a situation that kept the same proportions between author and publisher would increase your target royalty by 20%, or for most people, a point or two at most.

Now, what about the added costs? First off, you have to realize that the "manufacturing cost" is an average. Books with high page counts cost more than ones with low page counts, but more importantly, books with short print runs cost more than ones with long print runs. So there are very different effects in three possible eBook scenarios:

1. eBooks never really catch on. The incremental revenue is trivial, and neither authors nor publishers care too much what the split is.

2. eBooks seriously cannibalize print sales. What happens now is that print runs get shorter, and even though some of the copies sold (the eBooks) have no manufacturing cost, the manufacturing cost of the printed versions goes up. Depending on the ratios, the average cost to the publisher for each copy sold might even go up.

3. eBooks and print books are sold in a complementary fashion, and rather than cannibalizing print sales, eBooks represent incremental income. (This is an ideal situation for publishers and authors, which is why I'm so bullish on the idea that FatBrain put out, but hasn't executed on yet, of selling print and eBooks together for a blended price.) In this scenario, you can certainly make the case that the splits ought to be different for the eBook version, since it's "incremental revenue." But my question is who is making the investment to make that incremental revenue happen, the author or the publisher? As I'll explain below, at least for the foreseeable future, the publisher has significant added costs and little added revenue, while the author has no added costs. That's why, for now at least, keeping to current splits seems fair to me.

But more on those added costs. While it may seem trivial to produce an added e-book, and it probably is, if all you're doing is producing a PDF of an existing book, PDF is far from the dominant way to deliver eBooks. There are lots of possible ways to do it, ranging from developing online libraries and cross-book searching and annotation tools, to breaking up books in different ways at different price points. Any publisher exploring this area has added quite a few people, and paid lots of programmers and consultants to experiment with things that have not yet paid for themselves. Believe me, at the moment at least, eBook revenue is the most expensive revenue around for a publisher.

Further, there are significant added "distribution" costs, at least while the market is still immature. If online publishers were to come in with a universal system for delivering eBooks, putting them out would be as trivial as putting out print books, where there is a known distribution system. But with everyone rolling their own system, publishers have to deal with a hodgepodge of formats, business deals, negotiations, transfer and update methods, and so on. This isn't like getting a purchase order from Borders or Ingram, and just shipping the books.

There are also significant added sales and marketing costs. If someone like Amazon were to do what FatBrain said they were going to do, and simply offered an eBook option on the page for any book, in a seamless way, then the marketing would be automatic. But right now, anyone who wants to make eBooks happen (and isn't just distributing PDFs) has to put together:

• a hosting system
• online systems for security and transactions
• marketing

And since you're eMattering an already-published title, there would seem to be very few incremental editing costs, which would seem to almost completely discount the 'continuing costs' argument in your second and third paragraphs. If it were a new eMatter-only title those points would apply, but I don't see them being terribly relevant here.

If I understand your message, O'Reilly is taking the position that, "We don't know what effect e-publishing is going to have on traditional dead-trees publishing, nor what the margin profile is likely to be, so we'll play it safe and use the same royalty model.'

That seems like an incredibly conservative position for a ground-breaking house like O'Reilly to take. I'm also very surprised that it doesn't seem to acknowledge the author as anything resembling a partner in the enterprise. I would have expected O'Reilly to do something more along the lines of, 'This might be a lower-cost, higher-margin channel, author, so let's try a 20% royalty rate until we see how things develop. If it doesn't work out, we'll revert to the standard rate, okey?'

I'd put it the other way. Publishers are the ones who have to make an additional investment in order to make this happen. I'd think authors would want to incentivize them to invest and experiment, to develop this new revenue channel. It seems like an incredibly shortsighted position on the part of authors to be asking for a bigger part of the pie when the publishers are the ones sweating in the kitchen to make the whole pie bigger. Forward-looking authors might even say to publishers, "I'll take a lower royalty rate for now if you'll work to make my book successful as an eBook..."

Ann Navarro wrote:

If all sales were accounted for at the same rate, this argument would make more sense to me. However, we typically see different rates for foreign sales over domestic, licensing for translations, deep discount sales, and presumably some impact from bulk in the graduated royalty rate some authors enjoy as sales increase.

Therefore, why is it not possible to create an equitable distribution of profits obtained from this one channel? Authors certainly "pay" for the deep discount sales, shouldn't the "low cost" sales bring benefit back to them?

As a number of our authors noted, we don't do this kind of differential royalty accounting. And I think you'll find, if you look closely, that publishers who do this kind of thing do it to increase their take relative to authors. In fact, most of the channels where publishers pay lower royalty rates are precisely the kind of incremental, low-cost channels that many of you seem to see eBooks as. You can debate (as I have) the erosive effect of book club sales on regular sales, but if you believe in book clubs at all, they are money that comes to the publisher with no effort at all. Yet the royalty rate is normally considerably less.

The one case where I think that differential rates are really justified is with foreign translations. After all, you have to compensate a whole new group of people for work that they do. And in fact the costs of doing a translation are often higher than the cost of doing the original book. So splitting the royalty between the original author and the translator makes sense.

But if you start trying to slice and dice royalties by channel, you can get into really crazy situations. For example, sales to chain bookstores are easier and more profitable than sales to independents, since by dealing with one buyer and one accounting department, you can sell millions of dollars worth of books. Should sales to independents get a lower royalty rate? Etc. etc. Royalties, like print costs, are a game of averages.

At O'Reilly, we manage the averages, and we try not to look too hard at individual cases. That's why our publishing program is interesting: we don't, for example, only try to publish bestsellers, and ignore the small but important books, which are less profitable but nonetheless worth doing. Optimizing for profit doesn't optimize for value.

If indeed these sales became a significant portion of the overall sales count, the issue could be revisited. But at this stage, it looks at first glance to be, well, cheap --- not a word normally associated with O'Reilly publishing efforts (nor is the idea that this occured without the author's knowledge).

I'm sorry you feel that way. I don't think it's "cheap." I think we're trying to educate the writing public about the real costs and the real opportunity. I've been working very hard to get eBook prices up to the point where they are economic for authors and publishers.

"David L. Rogelberg" wrote:

In the end, regardless of logic, competitive market forces between publishers and authors will determine the royalties authors receive. While we hope author compensation is fair, fairness is not what determines author compensation. Authors receive their royalties because that's what publishers have to pay to get the content. It's not an act of altruism.

If one publisher takes an overly cautious approach to author compensation, then other publishers might see this as an opportunity. If all publishers take a cautious approach, then we'll likely see new publishers.

I agree completely. But I will also argue that someone who gives rates that are too high without understanding the costs may well find themselves unable to compete. That's the point that much of the discussion seems to miss: it's not authors who need to be incentivized to innovate and develop this market. It's publishers, retailers and distributors. The people who need to make the investment need to be encouraged to do so. Right now, the prevailing attitudes of authors that eMoney is going to be growing on the eBook trees is actually a disincentive to publishers to move away from the current situation.

The point is that before "an overly cautious approach to author compensation" bites anyone, someone will have to be making some money. Higher nominal rates might attract some people, but until someone is actually having success selling eBooks, the negotiation for the appropriate splits between authors and publishers isn't going to happen in earnest anyway. And again, my point is that if you guys want publishers to invest aggressively in this area, you can't expect them to be paying out extra dollars to authors for the privilege.

Pete Loshin wrote:

[much interesting discussion of royalties and costs of digital content deleted]

There has been a lot said about the costs involved in publishing online, but nothing said about the other side of the economic equation. The one thing I can remember from ECON 101 is that the optimum price for a product is set by the intersection of two curves: one tracking the price at which the seller wants to sell and the other tracking the price at which the buyer wants to buy.

I'm sure that's not exactly right--it's all kind of hazy, quite frankly, being a vague memory from about 25 years ago, but it's definite nonetheless: as the seller, my cost of goods doesn't make a damn bit of difference if no one wants to buy my product at *my* price.

Let's take an average $30 book, sold at that price both online and off.

For $30, I can get the dead-trees edition, which I can read anywhere there is enough light, which I can annotate, and which I can share with friends, loan to colleagues, and sell at a yard sale.

For the same $30, I can get the electron-only edition, which I can read only on an appropriate device costing at least a few hundred bucks. I can't annotate the digital version, and there may be restrictions on printing--though to print out a paper version, I'll have to commit maybe $5-$15 to paper, toner, and electricity. I'm also probably prohibited from copying or loaning or reselling the book by a software-style license (or by some form of copy protection--which adds a significant part of the price of the book). Oh yeah: if my hard drive crashes and I didn't back it up and I want to get the book again, I'll probably have to pay another $30 for it.

It doesn't matter whether it costs the publisher $15 or $50 to publish electronically, if consumers aren't willing to pay for the product.

I think Pete is completely right about this. Chances are good that eBook prices will be considerably lower than print book prices. As I've put it in another context, authors may want to lay claim to some of the publisher's presumed cost savings, but the public has already laid first claim on them...

Part of the "willing buyer/willing seller" equation is the perception that eBooks *should* be cheaper, which is why I spend so much time arguing for the fact that it ain't necessarily so. But Pete's argument is even more compelling. In the end, what Warren Buffet calls "Mr. Market" is going to tell us how these things should be priced, what the public will pay to publishers, and what publishers will pay to authors.

IMO, online for-pay content must be unique and/or offer additional value over the off-line version.

Or be an easy, incremental buy, which (once again) is why I liked that FatBrain bundled sales approach so much. I hope Amazon does something like that (and doesn't try to patent it :-) if FatBrain doesn't follow through.

I do think that there will be two main species of eBooks:

• incremental copies, presumably sold at a deep discount, mostly as a part of a bundled sale with a print book. This is the way we'll buy PDFs a la eMatter of books.

• online libraries. Whether through someone like NetLibrary, ibooks.com, or a direct publisher offering like the system Linda Walsh is going to be rolling out for O'Reilly this spring, these libraries will provide additional services above and beyond the individual book.

Digital copies of books, which eliminate the value of having the book online, are worth far _less_ than the hard copy version because they offer the consumer less value. The only benefit of buying a digital copy is speed: instead of waiting for delivery tomorrow, you get it right now. The big variable is whether that single benefit (which may not be so great for those with slow Internet connections) can outweigh all the costs of buying digital books.

That's not true. A lot of people who need high-value reference material want to carry it online, rather than on paper. With the growth of mobile computing, portability is important. This is why O'Reilly's CD Bookshelf products have been so popular. But note that the per-book cost is much lower than for print books (we've managed to get the price we need by small-scale aggregation), but not so low as to trivialize the value of the eBooks.

But still, your point is very good, Pete, namely that until we come up with an eBook value/cost proposition that makes sense to the consumer, a lot of this discussion is moot.

"Ben Sawyer" wrote:

The real use of ematter in my opinion is for good material on very specific topics that runs under 100 pages, preferrably under 50 pages. I think good books on critical topics could be broken up into smaller pieces and sold a-la-carte in such a way. I've got dozens of small 50 pages documents I'd gladly pay $10 or $30 for if they existed.

This is certainly true, but it begs the question of marketing. How are you going to find that small piece that contains just what you are looking for? Unlike the web, we don't have search engines to PDF-format eMatter, and we don't have mechanisms for publishers to do marketing of this kind of document.

This is where FatBrain has done the industry a disservice in the way they've set up the eMatter system. They've effectively set themselves up as the "publisher" for eMatter, without taking on any of the additional work that publishers normally provide, namely filtering (i.e. the proposal process), editing, and marketing! Ideally, they should have set up eMatter in some way that kept their role as distributor/retailer distinct from the publishing role, rather than setting up this promise that authors could somehow bypass publishers...without making clear that someone (and in the eMatter approach, that means either FatBrain or the author) is going to have to take on these additional tasks.

At the end of the day, we need universal standards for eBook publishing that can be adopted by a variety of online and print retailers, while what FatBrain has done has been to try to set themselves up as a publisher. I'm OK with competing eBook formats--that's not the issue. The issue is that by blending the role of publisher and the role of distributor/retailer, FatBrain has delayed the time when we can just put an eBook into the market, and have it appear on any online retail site.

Meanwhile, publishers need mechanisms for marketing and distributing eBooks that leverage their existing channels and methods for getting the word out.

For example, the other day I was looking for a good 20 page document on how to send high quality audio using ISDN lines. Including the equipment to buy, and how to properly setup and manage the ISDN line. It's a fairly simple process but I wanted some better details than I was finding and it took me a day to locate the right equipment manufacturers - a day I gladly would have paid money to not spend online looking for ISDN Audio Adapter manufacturers.

Another example, I've been writing a chapter on encoding online audio which is better than the stuff I've written before or seen written elsewhere. It covers a lot of the top products and details differences, etc. It's for part of another book I'm doing but on its own I know it could probably sell a number of copies for $5-10 as an extended chapter. This is the type of stuff that ematter could be good with. It's just going to take the right kind of publisher who can identify with a group of authors the types of small focused documents that will be successful under this type of publishing process. It also requires you to have a way to market the specific documents, in the case of the encoding chapter there are a number of online sites devoted to online music and webcasting that would probably link to the product if I put the effort in to email them about it.

I agree with Pete that large scale full books aren't as conducive to ematter style publishing but there is a whole new segment of documents and technical materials that could be valuable and profitable in this publishing model. However, the problem I see is marketing this effectively. The product development and marketing have to be done right before anything is written. In order to work I think an ematter style document has to have:

• A good focused topic that is extremely underdocumented
• Be able to be covered in under 100 pages, preferably 50 or less
• Have a ready audience that is easy to reach and market to via low cost means on the Internet
• Not take the author too long to write and properly edit/prepare for publication

All of these are great targets for eMatter books. I am just not sure how good a job FatBrain will do of "publishing" them. If you have ideas like this, bring them to O'Reilly :-) We may be able to put them out in both print and online. After all, there's no reason even a print book needs to be a 1000-page doorstop. And we're certainly trying to develop more outlets for short documents via the O'Reilly Network and our upcoming eBooks effort. But unlike FatBrain, we'll be picking and choosing topics, and doing our best to market the stuff we choose to publish.

This is a great discussion! Thanks for all the responses.

Publisher or Retailer, Part 2

Chris MacAskill wrote:

Our vision for eMatter is to introduce a new and complementary distribution channel, primarily focused at stuff that's longer than magazine articles and shorter than books.

...

There has been no economic venue this century for mid-length pieces--essays, white papers, short stories... All publishers reject them. Gone are the days when Charles Dickens could make A Christmas Carol popular. It simply doesn't fit the economic models of magazine or book publishers. But I believe the Internet makes distribution of these titles possible. Customers download and print PDF files by the millions today. Why can't we charge for them if the content is good, we protect the content against illegal copying, and we pay the publisher or author a royalty?

Just to complete the circle (and not just to maintain good relations between Fatbrain and O'Reilly :-), let me say that I think that this is a very worthy goal, and one that I support. In fact, it's a dream of mine as well. Many of my original books were "too short" for normal publishers. I've also launched several journals (The X Journal, The W3C Journal, and now The Journal of Linux Technology) as well as a whole bunch of technology web sites, because I'm aware that not everything fits into the price and packaging of books. Our whole focus in the O'Reilly Network (a network of ad-supported websites like xml.com, mySql.com, etc.) is to generate revenue from material that isn't yet ready for the book covers. I'd love to find more outlets for short material.

My biggest issue with eMatter is around the marketing language that has framed it in your appeal to authors, and some of the ambiguities you've introduced as a result. Here are some of the things that bother me:

• By saying that you "pay a 50% royalty" you are implying that the relationship you have with your authors is not the relationship of a distributor/retailer to a publisher, but the relationship of a publisher to an author. If you said "you set the price, and we keep a 50% commission" or "you set the retail price, and sell it to us at a 50% discount", you'd be making clear that you aren't performing the functions of a publisher, and that, far from being a far more advantageous deal than that offered by traditional publishers, you are in fact offering much the same deal on ebooks that you have worked out with publishers in print. Instead, you used language that implied that traditional publishers were taking advantage of their authors, and that this new medium was a great opportunity to redress that inequity.

  This may seem like hairsplitting, but the words we use shape the way we think, and the wrong choice of words can set in place the wrong business model, with ramifications that can take us years to untangle.

• As we discussed on the phone when you first briefed me about eMatter, the fact that it's a proprietary format also creates ambiguity about whether you are a publisher or a reseller. I'd ideally like to see you commit firmly to being a reseller, not a publisher, and one great way to do that would be to make the eMatter tools and software available (perhaps as open source) to anyone who wanted to use them. That would really allow authors to be online "self-publishers", to sell the eMatter from their own website if they liked, or via a variety of competitors. You won't own all the pie, but you'll own a piece of a bigger pie than if you set up a format for which you are the only outlet.

Ultimately, I think that eMatter (as currently positioned) won't work, not because there isn't a market for shorter pieces, but because you've set yourself up as the only source for material in this format. At the end of the day, we need something that is an industry standard. And I have a hard time seeing how that can come from any one of the online resellers, because anything put forth as proprietary by one is not likely to be adopted by the others. (I suppose it could eventually be pushed through in a version of the Betamax/VHS wars, by strategic alliances among a group of resellers who banded together against some other group, and who were able to make some high value content plays, but I'm not sure where that fits in your business plan.)

One of the things that works best in the publishing industry is the loose coupling between the various layers in the industry. Anyone can self-publish a book by going to their local print shop. They can hawk it on the streetcorner, or via any one of a universe of intermediaries. The very indirection and inefficiency of the market is what gives it its elasticity and low barriers to entry.

If eMatter is to succeed, you need either to become a true publisher (and thus work to find other outlets for the material on behalf of the author), or else commit to being a retailer, and let the author or intermediary publisher resell the same material, in the same format, through other outlets.

As to the issue of the 50% we pay versus the lower rates publishers pay... If I were a book author and I had a book-length manuscript to publish, I would expect to make more money by signing with O'Reilly and getting a lower royalty than I would by self-publishing eMatter. Why? Because O'Reilly's brand will give it credibility, and O'Reilly will add value through marketing and editorial contributions. They will make sure it sells through bookstores and they'll promote it at trade shows.

We pay 50% precisely because we don't offer what publishers do, so authors expect more. Besides, the self-publishing programs at Amazon and MP3.com made the royalty structure a forgone conclusion.

As noted above, my problem isn't really with the amount, but with the characterization in your marketing. When I met with your product manager, Dan Rush, the other day, he asked why what you were doing was dissimilar from a publisher selling books from their own web site. Isn't that also blurring the boundaries between publishing and distribution, he asked? Why is it OK for the publisher to act as a retailer but not OK for a retailer to act as a publisher?

My answer is simple: when publishers sell books from their web sites, they don't do so using the following marketing message: "Why give 50% of the cost of your books to a retailer? Why not buy direct from the publisher at half price and save all the distribution costs?"

Your eMatter self-publishing pitch might be expected from someone who thought of themselves as a competitor to professional publishers, but it doesn't sit well from someone who expects to derive the bulk of his revenues from those self-same publishers.

Does eMatter compete with books? I don't believe so. The movie studios were beside themselves over videotapes when they first came out. But people still went to the theaters as Blockbuster expanded. The studios sold more movies. New distribution channels nearly always turn out to be additive.

Counterexamples abound. TV did compete with radio. The internet does compete with TV. And in my observations, widely available free online information does compete with books. (It competes even more with magazines, as the demise of many a computer trade journal will testify.) Does an ematter version of a book diminish the sales potential for a print version of a book? Probably not, if the two are coordinated, as, for example, hardback and paperback books are coordinated. (You don't publish the mass market paperback at the same time as the trade hardcover, for instance.)

I don't think we know yet what the "right" coordination mechanisms are for online and print information. I'm hopeful that we can find them, and can continue to make the provision of high quality information a profitable endeavor for everyone participating on this list.

Is Fatbrain a publisher? We don't want to be. We want to be eBay, connecting buyers and sellers.

Yes, and eBay takes a commission. It doesn't pay a royalty :-) (But I'm beating that one over the head.) eBay is actually a pretty good model for you guys, but if it's the model, I'd argue that you're taking too big of a commission :-) eBay wouldn't get very far if they charged 50 cents on the dollar for each sale.

(Playing devil's advocate: You might grow the market more quickly if you had eBay-like commissions and let publishers and authors get closer to their current margins with a price that was substantially below the list price for print books. (In your current model, you ask the author/publisher to supply any discount from the retail price of an equivalent print product, and take your 50% commission from that price, when it would make more sense for you to either take a smaller commission or to do discounting on your side, out of your commission.))

Anyway, I hope there are no hard feelings over this. We're inventing a new industry, and the more we argue about things, the better chance that we'll air some good ideas that will get us all moving in the right direction.

Thanks for taking my lambasting so well, in the spirit of a substantial argument, and not turning this into a flame war!

Publisher or Retailer: Part 3

Linda Laubenheimer wrote:

In light of Stephen King's experiment, it seems much more reasonable to charge less rather than more if you want to gain market share. Digital content, even books, is fleeting: why should it cost minimally less than the more permanent version? Think about video rental vs video sales instead of rock concerts vs album sales.

$x per Kbyte for DL, or on media with $x/K content + $y media & burn charge + $z shipping & handling. Author gets ~50% of the DL(content) cost. Format should be readable on any browser, with "decrypt" for a given publisher being a plug in.

If you look for 50% royalties, you are making a statement about the nature of the channels your book will be sold through: i.e., it will go direct to the consumer. Stephen King (and a few others) are a special case. They already have the name recognition to go direct. But does the average author? No.

50% royalty assumes that there is one level of middleman, perhaps--a publisher or retailer. But eventually, you will have three or four levels, just as you do in the p-book world. Publishers aggregate authors for retailers (who otherwise can't deal with all of the flood of product, and need some trusted sources to parse the flow for them.) Heck, even established publishers have to compete forcefully for the attention of retailers, and wholesalers exist to aggregate smaller publishers for retailers; retailers aggregate customers for publishers.

Middlemen of various kinds exist because of the fundamental mathematics of the situation. Right now, we're in an anomalous situation where there are relatively few ebooks, and consequent high visibility. Now imagine millions of ebooks, and tens of millions of readers. It gets ugly fast.

We saw the same rhetoric in the early days of the web. It was going to level the playing field, disaggregate all the middlemen, and make all e-publishers equal. This was true for a short time, while there wasn't much content on the web. Then we saw the emergence of a new class of middleman--catalogs and search engines. And note, if you aren't found on the first couple of pages of a search, you might as well not exist for a lot of people. Links help, but just as in the physical world, high profile publishers have emerged. If you write a tech opinion piece, it's far more likely to be read and linked to if it appears on slashdot or salon.com than if you put it up on your own random web page.

You might say that even if this is true, there are such savings from ebooks that royalties should still be much higher. I don't agree. As I've noted in previous postings, the savings from paper, print and binding are on the order of 15-20%, depending on the publisher, and the ADDED costs, are on the order of 100%. Building or buying the technology for eBook distribution is more expensive at this point than printing. Sure, you can throw up a PDF, but building something that really works for the medium is harder than that. You need mechanisms for e-commerce, for hosting the content, for continuous uptime, for customer service for readers who can't figure out how to make it work.

It's always harder and more expensive to do something on the front end, before norms are well established. I know this from painful experience. I started out publishing in 1985, with a print run of 100 copies of two little pamphlets. I was sucked into an existing system of retailers, distributors, existing pricing models, established ways of reaching customers. It took time, but the issue was never in doubt. By contrast, in 1992, I launched the first commercial web site (GNN, the Global Network Navigator, which I later sold to AOL). GNN was the first "portal," the first site on the internet to carry advertising, and the template for a lot of what followed. No one knew what to make of it. We had to do everything ourselves. We commissioned the first ever market research study of the net to convince advertisers that the net was going to amount to something. We had to evangelize the net itself. A very expensive proposition. Yahoo came along, got VC money, and outspent us. We sold/partnered with AOL, who screwed things up (on the way to being very successful nonetheless). But it wasn't pretty, and it wasn't easy.

Because of the expense in establishing a new market like eBooks, I wouldn't be surprised to see author royalties go DOWN, not up. Sure, individual authors will become publishers, just like I did in the print world, and Mitch Waite did, and Andy Shafran did. But the exception doesn't prove the rule. The rule will be that eBooks won't take off until there is a rich ecology of middlemen--all of whom will need to take a cut in order to provide the essential services they offer.

In short, when you're thinking about eBooks, be sure to leave room in your business model for lots of parties to have a piece of the pie. The system that enables the richest ecology of partners will likely be the one that wins, not the one that tries to collapse the system into winner-takes-all. Have you guys seen the miserable sales results from eMatter? I tried to warn Chris McAskill that it wasn't going to work, for the reasons I outline here. A slight shift of the business model, to enable more players, would have made all the difference. Maybe they'll get there yet, but if not, I see lots of other people, like Lightning Source, who understand that the eBook marketplace will require a lot of the same types of players as the pBook marketplace.

Copyright © 2009 O'Reilly Media, Inc.