Monday, September 28, 2015

Whither eBook Subscription Services

Last week marked the announcement that the eBook subscription service Oyster will be shutting down in early 2016. This summer, Entitle—one of the original three eBook subscription services—quietly closed its doors. Scribd, the third of the group, had to backpedal from its promise of “unlimited” books per month to throttle the usage of its romance readers. Since the original three commenced operations, Amazon entered the market with its Kindle Unlimited. Google has hired the folks from Oyster and so is presumably considering a subscription service as part of its Google Play.

So the field is changing and Amazon plays a big role. But, are subscription eBook services sustainable?

The Economics of a Subscription service

The basic business equation still holds: Revenues – Expenses = Profit


In a standalone subscription service, revenue comes primarily from the monthly fees users pay to enjoy the service. Kindle Unlimited charges $9.99 a month. Oyster charges $9.95 and Scribd charges $8.99. Multiply the monthly fee by the number of subscribers and you have revenue. Unless you can sell ads along the way, or sell your subscription list, or monetize something else, that’s your revenue. To keep things simple, we’ll assume revenues consist solely of subscriber fees.


To operate, the business has to have a website designed to collect memberships, present a searchable catalog, record and deliver selections. While there are some variable costs involved in a subscription service, most of these operating expenses are fixed costs. One subscriber or a million they will occur, so it is important to grow your subscriber base quickly so the expenses decline as a percentage of revenue.

The second major expense are the acquisition costs. Publishers (and authors) want to be paid if someone reads their book.

Subscription services must negotiate with publishers, distributors and, in Amazon’s case, indie authors, concerning their compensation. Oyster and Scribd generally paid publishers something very close to what the publisher would earn by selling the book through an online retailer.

Under the Oyster/Scribd model these are variable costs. The more books lent out, the larger the expense. The more expensive book lent out, the larger the expense. We’ll discuss Amazon’s model in a bit.

Who would buy an eBook subscription?

For a reader the equation to calculate savings from a subscription service is:

Subscription Fee – [(Number of Books I think I’ll read in a month) X (average cost of book)]

Let’s say the average cost of eBooks purchased without the subscription is $2.99. The reader is a winner at 4 books a month (3 Scribd), loser at three (2 Scribd) or fewer. If the average cost drops to $1.99, then it takes six (five Scribd) eBooks to “win.” At $5.00 it only takes two books to be ahead.

Strategies Suggested by the Profit Equation

(1) Feature less expensive books. Free are best. $0.99 are very good. $1.99 good, $2.99 okay and anything more is pricey.

Look at the subscription catalogues and you’ll find they are crammed with “classics” that happen to be out of copyright (and therefore virtually free to the service)

You will see a very limited number of current, higher-priced books from the Big 5 Publishers. They are simply too expensive. I suspect those that are in their catalog provide the publisher with much lower royalty rates—the eBook equivalent to mass-marketing to Walmart or Costco.

(2) Pray people do not read too much.

Consider the profit formula and how it applies to fitness center memberships. In January in the flush of New Years’ promises, lots of people make the basic calculation that they will win by purchasing a yearly membership. And then by the end of January many stop going. These are gravy memberships. Revenue exists, but no variable expenses. That overestimating consumption phenomena may happen for book readers as well, but unlike the gym membership, they probably will not give up reading books entirely. Even if they have an off month or two, the only bar preventing them from restarting to read a lot again is finding time. For a gym membership there’s a psychological barrier of anticipating the physical pain necessary to get back in shape and the physical barrier once the individual actually restarts.

Subscription services try to limit reading by not providing all the books the subscriber would normally like to read. New best sellers are rarely offered because they will cost the subscription service too much. If people spend time reading those in paper form instead, it saves the service money and cuts down on the total books read on the subscription.

When people read a lot, the subscription service loses money. Scribd found itself in that situation this summer regarding their romance readers. Since they could not limit the number of books selected by romance readers, and they were not willing to increase the subscription price, they cut the number of romance novels available in the service. Drastically cut. They kept the freebies and eliminated the expensive books. Some of those in between remained. Smashwords estimates Scribd cut 80-90% of Smashwords romance titles.

(3) Pay publishers and authors smaller fees. With limited exceptions, Kindle Unlimited (KU) does not offer Big 5 Publisher books. Its offering is largely populated by its own imprints and indie author publications.

For indie authors, Amazon creates a pool of money—it determines the size—and allocates that pool to authors based on the number of pages read. Their previous practice had been to allocate the pool based on number of “downloads.” They found this encouraged gaming of the system whereby authors would split a book into four parts, so a 200-page book becomes four 50-page books, earning four times the income for the author.

Note that Amazon determines the pool size, which from an author’s perspective means Amazon determines the per page revenue. The indie author’s choice is to join the program or not. For the first month of this new payment system’s operation, July 2015, KU paid $0.005779 per page. For a 300-page novel that means $1.73. For August the payment per page dropped 11% to $0.00514, and the same fully read 300-page book would earn the author only $1.54.

Notice that if that 300-page book were priced at the low end of Amazon’s preferred range of $2.99 to $9.99, the royalty for a book purchased would be about $2.09.

Amazon has structured a model where the author subsidizes the subscription service. I’m sure Amazon will argue that the author will make it up in volume, but how can you know, and what is to prevent Amazon from settling on a much lower rate in the future, say $0.001 per page so our 300-page book now earns the author a paltry $0.30?


The Scribd model as currently constructed does not hold economic water. It is too easy for subscribers to determine if they are saving money or not on the service. There will be a small percentage of subscribers who are losing money by participating and are insufficiently motivated to stop their subscription, but they can’t make up for the costs of heavy users.

Amazon can control its costs by defining how much it reimburses indie authors, a large percentage of the KU offerings.

But consider Amazon’s approach to the Audible subscription service. It has a fixed monthly fee, but for that price you get one “free” audio book. The rest of the catalog is discounted. As long as the consumer was going to purchase at least one audio book a month, the customer is “ahead.” Amazon’s costs for additional downloads are offset by additional revenues. Although the audio books are discounted, I’m betting those lowered costs cover the royalty payments plus profit.

Scribd could move to a similar model for books. Say, you get four a month for free. The rest you can have for a discounted fee.

If Google Play (or Apple for that matter) decides to enter the business, they will bring deep pockets. They don’t yet have the indie author network as Amazon does. But what would happen if they offered better royalties than Amazon? It could prove interesting, yes indeed.

~ Jim

This post originally appeared on the Writers Who Kill Blog

Sunday, September 20, 2015

The Allure of a Mystery Series

Panel at 2015 Left Coast Crime

One of my great pleasures of mystery fan conferences such as Malice Domestic and Left Coast Crime is the opportunity to talk with a wide variety of readers. When I get to chatting with a reader I usually ask about favorite authors to compare notes. After a while I’ll ask about how they approach series.

Writers and publishers like series because of the long-tail effect: if someone reads one book in the series and enjoys it, chances are good they’ll read another in the series, and another and another. Each new addition to the series not only has the potential to sell to fans, but bring in new readers who will ultimately want to read the entire series.

A couple of years ago my better half, Jan, and I were attending Malice Domestic. We wanted to read all of the books nominated for Best Contemporary Novel. Jan discovered that Julia Spencer-Fleming’s Through the Evil Days was part of a series. Rather than just read that book, Jan wanted to read the entire eight books in order. So she blitz read her way through the series in about two months’ time (while reading the other new-to-her books as well).

It turns out Jan is not alone. A significant percentage of people I’ve talked are like her. (I wish I had kept an accurate account so I could tell you the exact percentage.) They strongly prefer to read series novels in order—some so much so that they will not read a series out-of-order! Can you imagine how long it would take a reader new to Sue Grafton to catch up on Kinsey Milhone and the now twenty-four alphabet books starting with A is for Alibi and ending currently ending with X? Not going to happen, right?

Well consider these Kindle Book rankings for Grafton’s series (early morning 9/4/15):

X -- #19
W is for Wasted -- #1,603
V is for Vengeance -- #4,603
U is for Undertow -- #8,886
T is for Trespass -- #9,164
C is for Corpse – #7,685
B is for Burglar -- #6,760
A is for Alibi -- #3,137

Not only does Grafton have a top twenty hit two and a half weeks after its release, she has seven other books in the top 10,000 Kindle sellers: the previous three and the first three. People are catching up if they’ve missed a few books, and people are starting at the beginning. This long tail is why publishers like successful series.

To allow that piling on effect, publishing contracts were (and often still are) for three books.

And the three-book contracts are, I suspect, why I have found another phenomena amongst many mystery readers. They won’t start reading a series unless there are a sufficient number of books published. The oft-stated reason goes something like “I don’t want to fall in love with an author and then have to wait a year for the next book.”

When presented with the Catch 22 situation that if no one buys the first books in a series, there won’t be more books, the next response is something like, “I want to make sure the series will be there.” Particularly with small presses and self-publishing they don’t want to invest in a character for only one or two books. From my sampling of folks, those with this attitude often require a minimum of three books, and preferably four or five in a series, before they will become interested.

So, two questions for you, dear readers:

(1) Do you prefer reading series in order? If so, must you start at the beginning, or do you read the most recent and then if you enjoyed it go back and start from the beginning?

(2) Do you have a required minimum number of published books before you’ll start reading a series, and if so how many?

For those of you who want at least three, my Seamus McCree novels are now eligible for your consideration. You can read them in order: Ant Farm, Bad Policy, and Cabin Fever. (Notice I subtly stole the alphabet idea from Sue Grafton?) For those who want at least four, I promise Doubtful Relations will be published in 2016.

~ Jim

The article originally appeared as a guest post on Debra H. Goldstein's Blog

Monday, September 14, 2015

Beta Testing Your Novel

Doubtful Relations Title Page
Last month I sent my current work in process, DOUBTFUL RELATIONS, out to five volunteer readers. I have received feedback from three of them; the other two sets of comments are scheduled back this coming week.

I call this process beta testing and the volunteers beta readers. These terms are used in different manners, so I need to define my terms.

For a few years during my career, I managed a group of software developers. In that environment, alpha testing encompassed the work the group did internally to make sure the software (new program or update) was doing what it was supposed to be doing. When we were reasonably comfortable it was, we performed the second level of testing, beta testing, named after the second letter in the Greek alphabet.

In beta testing, we put the software in the hands of real users in a real environment. We knew the program was not perfect. We were still fixing a list of known bugs (and periodically adding to the list). However, the product was sufficiently close and stable that we needed to expand testing past our myopic vision and turn it over to our users to point out flaws and issues we were too close to the program to see.

For example, let’s say you create a new way of cooking omelets that utilizes the energy from your morning workout. You develop a recipe that includes a list of ingredients and tasks, but fail to include the step in which you remove eggs from their shells. That step is obvious to you, because you always do that. The cook might not realize that your new technology still requires the separation of egg from shell and ruins the omelet. Your directions are not clear.

The combination of plot development and character motivations in a novel take the place of directions in a recipe. I ask beta readers to let me know of plot bumps or holes and of characters who do something that seemingly does not make sense. I will have already addressed any problems I discovered on my own as well as those indicated by my alpha reader, who has read the manuscript at a much earlier stage. [Some writers use critique groups as their alpha readers, others use a trusted writing partner or friend. I rely on my life partner, Jan Rubens.]

I know flaws remain in the manuscript. I have not yet polished the language, and because I have made changes to a draft of the manuscript immediately before releasing the beta version, I may have introduced new typos and included a sentence or two that might make one wonder if I had flunked English as a second language. Readers can ignore those kinds of problems, as long as they are not too frequent, and instead concentrate on the main issues of plot and character.

Unlike software beta testing where, as flaws are corrected, updated versions are periodically released to the users, I now typically have two discrete passes for beta readers. What I have so far described is the first pass. Once I have the manuscript in “final” form—perhaps ready to submit to agent or publisher or for self-publication—I will ask a different set of folks to read the manuscript looking for anything wrong. This is beta testing in the sense it is a real product placed in real users hands in order to receive feedback prior to publication. However, by that point I am in the final steps of my quality control process and readers should not be finding any major problems. I hope they will find the stubborn typo or homonym error, as well as any formatting issues. Perhaps because these tasks are so different from the first beta readers’, I should refer to this group as my gamma readers?

Today’s status:

Are you curious what feedback the first three beta readers have provided me? I have not read their detailed comments because I am waiting for all five readers to respond before returning to work on that manuscript. Based on the cover emails I received, two of the readers enjoyed the story, but had specific suggestions on how I can make it stronger. A couple of their cover comments confirmed concerns I had and other comments shed light on issues I had missed. Great stuff!

The third volunteer was so disgusted with a major character that she stopped reading the manuscript with 60% still to go. Since the character in question acts in ways real people act, I need to look beneath the reported problem. (Tom Wolfe writes best sellers about people I don’t like and don’t much care about, so the issue is probably more than that the character is not emotionally attractive.) When considering this one reaction along with other beta reader observations, I must determine if I have not sufficiently defined that character’s motivations so her actions make sense. Or perhaps I have insufficiently defined other characters’ motivations so their reactions to the unlikeable character are understandable. Either way, this reader’s reaction to stop reading will have done me a great service.

Unlike a piece of software, there are no absolutes in the writing business. What one person sees as a flaw, other readers may love. When I review all the comments, I must do so with the filter of my own understanding of the story and remain true to my writing style and voice. I know the comments include excellent suggestions that I look forward to implementing. I know they will contain hints of problems that will require me to ferret out their underlying causes to solve. There will be individual reader preferences that I will need to ignore to stay true to myself (and sane, since some comments invariably contradict other comments).

I can’t wait to dive into the next draft.

~ Jim