The Usenet rec.art.sf groups are currently having one of the recurring discussions on pricing of E-books and other related issues.
First, let's look at how Ebooks impact me, today, personally:
Baen is one of the few sane publishers in this respect. I get a much larger percentage of the price of an Ebook than I do of either a hardcover or mass market paperback. The difference is such that for an E-book sale of a novel I get ~$1.50, I think, while for the paperback I get $0.50 or so. For a hardcover, I get $2.50 or so, but the price of a hardcover is much higher. So AT BAEN, the sweet spot where you pay the least and I get the most is the Ebook.
HOWEVER -- businesswise in terms of how much booksellers want to stock my books, the Ebooks don't count much if at all. Hard copy has to sell. If Ebook becomes the dominant business, that will change a lot of things.
Now, on to pricing:
A lot of the argument on both sides tries to assume some aspects of the business working a given way in order to arrive at their conclusion. Neither side should do this, but both do. So here's MY analysis.
1) Mass-market paperbacks, as a class, make money. This is inarguable, because if paperbacks typically lost money, they would no longer be being printed. Moreover, there are quite a few books which NEVER have hardcover releases, so it cannot be assumed as a general case that a lot of the costs of book production are borne by the hardcover and the paperback is getting off easy. Thus, the current mass-market price is sufficient to support the entire production and distribution costs of the book and leave a reasonable profit for all concerned.
2) To determine a reasonable cost for an E-book, therefore, it is necessary to eliminate the costs peculiar to the hardcopy version of the book, and then add in the costs peculiar to the electronic version, and that should be the price point.
First, the largest single cost added to the book price because of its nature is the distributor --> retail markup. This is what the COVER PRICE is based on, and it assumes a roughly 45% markup from Distributor's price. Thus, the $8.00 paperback you find in the store cost that store $8.00/1.45 or about $5.50.
The distributor of course gets to mark up the books; so that $5.50 is higher than the minimum book price. I'm going to assume the distributor's overhead is a LOT lower than the overhead for the retailer, and guess their markup at around $15% (which is probably lowballing, but I suspect it's close). This makes the distributor's cost $5.50/1.15 or about $4.80.
After that come the printing and distributor costs. The book probably costs a relatively small amount to actually produce physically -- I think about $1.00, but let's be conservative and assume the actual physical cost of the book is $0.50.
This takes you down to about $4.30. I'll assume there are some other minor costs associated with the book and make that a nice round $4.00.
THIS IS THE BASE COST OF MAKING THE BOOK.
The price of a PROFESSIONALLY PUBLISHED Ebook, therefore, cannot go below about $4.00, or someone in the CRITICAL path of publication will have to go.
Now, we have to add costs for the Ebook production and distribution. There are some -- make no mistake about that.
First, you need storage space. Not being technically savvy in this area any more, I don't know exactly how many copies of a given file you need on your servers to support lots'n'lots of downloads, but let's take the worst-case (which would also be for business a best-case) scenario. For each book, you need enough capacity to support the downloading of the book fast enough to, let's say, earn out the author's advance in one month (as the highest sales of a book come in the first few weeks of release). Let's further assume that the average author's advance is ~$10,000 (I don't think so, but maybe if you count the advances of the top earners). This means 20,000 downloads a month (remember that right now we're assuming that the author gets ~$0.50 per copy) or about 28 an hour. If we assume it takes a full minute to download a book, we could double this and STILL only need one copy of the book live and available. But to be on the safe side, let's assume you keep 10 copies online at all times.
Let's say you release 20 titles a month, that's 200 files of, perhaps, 10MB each (assuming fancy artwork, no significant compression, etc). And let's say you want to keep your entire back catalogue available; rights revert to authors after 5 years unless you're still selling books (it's more complex than that, but bear with me). So let's say that this averages out to 10 full years of production. So that's 10 (years) *12(months per year)*20(titles per month)*10(copies per title)*10(MB per file) = 240GB.
A few years ago, I'd have been impressed. However, my laptop -- far from the top end -- has about half that much. Checking Tiger Direct, I see 500GB drives available for $49.00. Let's have a five-disk RAID setup to make sure there's no failures, that's $250.00
Downloading bandwidth will be a bit tighter, but still, you can figure that the most recent titles are eating the majority of your bandwidth. Say half your total bandwidth goes to the current month, so your total bandwidth requirement for the server will be 2(times the 50% to get you to the total)*20(number of titles per month)*10(megabytes per minute, under my doubled assumption a couple paragraphs back) = 400MB/minute, or a little less than 7MB/second. This is about the limit of an 802.11g WiFi modem. By modern standards, it's not much at all.
Supporting this downloading, of course, will require an IT department, a secure shopping/payment application, secure servers, and so on. HOWEVER...
... much of that support is included in the basic cost, because your publisher will already HAVE to have its own IT department and network. They would also be paying people in charge of dealing with distributors and so on who would no longer be needed, or whose efforts could be reallocated to supporting the Ebook market. Perhaps you would need to add one more full-time employee to administer the E-sales end of things, but electronic sales technology is a well-known, solved problem which is very inexpensive, reliable, and easy to handle.
Most of these costs are spread out over all of the titles, also. Each individual book has to pay for its slice of bandwidth, tech support, and so on. To be fair we'll spread the costs over one year (one year ROI), so... let's say you'll buy new equipment for this. $1,000 for a new server machine, $250 for the RAID setup, $50/month for additional bandwidth over your normal corporate needs for $600, and $200,000 for your Ebook person ($100k/year salary, burdened -- let's pay him or her well), plus $1,000 for "other costs" -- maybe you buy a REALLY good secure system. This comes to about $203,000 more or less.
Per book, that's divided up over 20*12 or 240 titles times 20,000 copies per title or 4.8 million books, or about $0.042 per copy.
So to meet all costs of selling the E-books yourself, it looks like you could add a nickel to the cost.
Now, one of the issues around Ebooks is that people don't like the idea of the author only getting fifty cents per copy. Let's say that we want the author getting a bigger cut -- say as much as he makes on a hardcover. That's another $2.00.
Of course, the publisher may be concerned -- especially as the market shifts more toward EBooks and people can potentially be copying and distributing it themselves -- that they need to make up their costs faster. So they add on another $0.25.
So you have $4.00+0.05 or $4.05 as the base number which would maintain the status quo.
Increasing the author's share and adding a cushion for the publisher risk brings us to $4.05+$2.00+$0.25 = $6.30. Tack on another $0.70 for miscellaneous additional costs (maybe more advertising, additional employees, more insurance, etc.) and we reach $7.00
In conclusion: I would guess the lowest a PROFESSIONAL Ebook can sell for would be around $4.00 or so, otherwise something, somewhere, will be getting shorted.
The HIGHEST reasonable price would be slightly below current standard paperback prices.
I don't see ANY reasonable assumptions which would justify prices even half of current hardcovers. (except the "E-ARC" type which Baen does, in which you're not paying for the book per se; you're paying to read it before other people can, and do so in a convenient electronic format.
There IS one additional joker in the pack: Digital Rights Management (DRM). Apparently, some of the most effective DRM schemes are extremely processing intensive (requiring more expensive servers and more complex data processing), AND are licensed such that the user PAYS A FEE PER EACH PROCESSED COPY OF A B
Baen doesn't use DRM, of course, so that doesn't figure in.
So there you have it, my own off the cuff but, I think, reasonably accurate analysis. EBooks cannot -- if you want to (A) pay the author, and (B) have the work professionally edited, proofread, illustrated, etc. -- be sold for $0.25; but they also shouldn't be going for over $10.00 a pop, either.