The Shatzkin Files

Subscription models seem to me to be for ebook niches, not a general offer

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Share on RedditShare on StumbleUponEmail this to someone

Another fledgling ebook retailing venture came through our office this month touting a subscription proposition. I told the entrepeneur “I’m skeptical of the subscription model for ebooks,” and he said, “I know”.

We had a great chat, but I’m still skeptical. When I say that, I mean I’m skeptical that a general offering subscription model can work.

There certainly is a logic to subscriptions, particularly for those who think the book business should learn from other content businesses. Cable TV really started with subscription and then only later moved to pay-per-view, which is more like the ebook sales model (but not exactly). We have Netflix for movies and TV, Audible for audiobooks, and a host of services for music, the most successful of which seems to be Spotify.

I have a Spotify subscription, even though I don’t use it much. Perhaps foolishly, I’m comfortable spending $119.88 a year (which is what $9.99 a month comes out to) to have access to just about any song I might ever want to hear instantly when the urge (or suggestion) to hear it arises. (Spotify very seldom disappoints me by not having the song.) And that’s even though most of my listening needs are satisfied with the 6,000 or so songs I have in my iTunes repository of which the best 1,000 are on my phone.

Spotify was cited by the entrepreneur I met as a motivation for him to start his ebook subscription business. As he correctly pointed out, “sharing a playlist” with a fellow Spotify subscriber enables them to immediately — with no additional cost or friction — “consume” that music. Sharing an iTunes playlist with somebody just leads them to having to make purchases which, quite aside from the money, put time and (a considerable) effort between receiving the playlist and enjoying it.

So, it is posited, this logic should apply to books. With a host of very explainable exceptions, I’m not sure it does, at least not anytime soon.

I’m fresh off a speech in Washington about what the DoJ doesn’t understand about publishing. The answer, if boiled down to a single word, would be “granularity”.

According to the MPAA, North American movie releases for 2007, 2008 and 2009, were 609, 633 and 558 respectively. There are foreign films and perhaps some below-the-radar indie films that must be added to that number to reckon what’s being made available, but it gives you an order of magnitude.

The Big Six publishers average more than 3,500 titles a year each. And there is far more production of titles beyond the Big Six in publishing than there is production of movies beyond the Hollywood studios. It would be very conservative to estimate that there are 100,000 new professionally-produced book titles a year intended for consumers. (Many more are published for professionals or as school or college texts and were you to add in self-published ebooks, which sometimes reach big audiences, they would multiply that number.)

Commercial releases of music would fall in between movies and books in number, but much closer to movies.

That’s the short answer as to why most people share music and movie experiences with far more friends and acquaintances than book experiences. It is also the short answer to why people outside the book business just can’t grasp it; each one of those books is a separate creative and commercial endeavor, down to having its own contract, its own development path and schedule, and its own marketing requirements.

(It also helps explain why many people who use libraries for some of their reading don’t use it for all. No library will have all the books a voracious patron would want to read.)

In the days before and digital books, there were two kinds of subscription services that worked for consumer books.

Book clubs offered price deals and curation (help with selection) but it was the price deals that really attracted members. Before ubiquitous bookstores (which arrived in the 1980s), Book-of-the-Month Club and The Literary Guild got the highest-profile books distributed to consumers who would have had a hard time getting to them (as well as those near bookstores who just wanted the convenience of mail delivery.) As bookstores spread, the Clubs found that “niche clubs” (around mysteries, science fiction, or subjects like gardening) were apparently more profitable than the big general interest clubs. (“Apparently” is a highly operative word, but the explanation of that will wait for another day.)

The other subscription concept that worked was the “continuity series”. The market leader there was Time-Life Books. These books were about a particular subject (World War II, say) and they were “packaged” specifically for the series and not available in stores. Continuity relied on the combination of intense subject interest and the “collection” mentality: somebody who started collecting the series didn’t want to have holes in their collection.

Both models were pretty much blown out of the water by online book purchasing which suddenly made every book available for home delivery to everybody everywhere.

In specific niches, subscription models can work very well. The granddaddy of them on the digital side is Safari Books Online, originally conceived and built by O’Reilly in partnership with Pearson. Safari serves a community of programmers and has a huge collection of instructional and reference books which they can use on the job. Most users of these books dip in and out of them, rather than reading them straight through. And they frequently like the idea of checking out what several books might say about a problem they’re tackling.

Safari pioneered the model of dividing the publishers’ share of the subscription fees by metering usage. The more your book is viewed, the more money you get from the pot. And since users of Safari will almost always find the answers they need within the service, leaving your book out means it won’t be found and used. Since at least some of the time Safari usage could lead to a sale of the book itself (even if not very often for most books), that discovery element is lost along with any Safari-generated revenue if the book isn’t included in the database. A publisher should feel pretty confident that they aren’t losing many sales being inside Safari.

(The model that looks like “all you want for a price” to the purchaser and like “pay per use” to the content owner in even purer form than Safari does it is the deal offered by Recorded Books for its digital downloading service for audiobooks to libraries. There are other subscription models in the library space; it is a distraction to the point of this post to get into them which is why they’re not covered here.)

O’Reilly saw at the beginning that their books alone wouldn’t be the strongest subscription offer so they were open to participation by others from the very beginning. Safari is exceptional in at least three ways: they are bigger than one publisher; they are built on a professional user base; and they deliver value primarily through chunks, not end-to-end reads.

But if a publisher is strong in a niche, a subscription service can work for them too: Baen Books (science fiction) and Harlequin (romance) are two niche publishers who have sold subscriptions successfully. (In fact, Harlequin recognizes sub-niches, further segmenting their audience for better subscription targeting.) The Osprey-owned sci-fi house, Angry Robot, offers subscriptions. eBooks by subscription are also part of the model for Dzanc, which does more literary books (fiction and non-fiction; they’re really less niche-y, except for “quality”) and it will be interesting if they can make the “quality” paradigm work the way “romance” and “science fiction” do.

Sourcebooks is a general trade publisher, but they have a robust romance list. They’re trying to establish a club and community called “Discover a New Love” which operates more like the old BOMC: subscribers can choose one of four featured titles each month in addition to getting other benefits from discounts on other books to early looks at some titles.

Subscriptions are offered in the children’s ebook area as well. Disney Digital Books has a monthly subscription service, as does Sesame Street eBooks. In both cases, the model is browser-based delivery rather than downloads.

F+W Media is a publisher that works across many verticals (niches). They had two big head starts. One is simply being vertical. They have audiences that are defined by their interest, which has been the key to making a subscription offer work in the book business. The other is that they were once publishers of magazines and operators of book clubs, so they have experience with direct customer contact and managing those relationships. They also had a lot of names. And F+W is managing subscription offerings for many things other than ebooks.

Most of F+W’s communities have been non-fiction (subject-specific) and they offer subscriptions for content in art, writing, and design. But they are also venturing into the romance market now and their Crimson Romance offer is an “all you can read” model. Baen introduces the wrinkle of releasing a novel in stages to subscribers, like a serial.

And we note in the recent reminder that the TED conferences started doing ebooks (sort of: only within an iOS app) that a subscription model is part of their thinking too. Once again: in a niche. The app that enables them to manage subscriptions is powered by The Atavist, which is another attempt to build a following for a publisher distinguishing itself by its content choices, like TED or Dzanc, rather than around already-established consumer clustering (romance, sci-fi, or a topic like writing or design.)

It is worth noting that there are “all you can eat” subscription offers and ones that are limited but which offer discounts on further purchases. That variation exists in other media too. Spotify is one price for everything; Audible and Netflix meter your use and you can pay more if you consume more.

There’s a pretty strong pattern here to the subscription offers we see.

They’re usually done by publishers. (Safari isn’t a publisher anymore, but it was started by publishers.) That means they’re working with the publishers’ margins (bigger than an aggregator’s margins). Controlling the product flow means they can make good use of intereaction with their audience, learning through data and conversation what they should be doing next. And, most important of all: from a product offer point-of-view, they’re focused.

They’re precisely the opposite of Spotify or Netflix or Audible who all want every single song, movie or TV show, or audiobook they can lay their hands on.

So, what about a more general model for ebooks?

It hasn’t happened yet and I don’t think it will anytime soon, despite the ambitions of my recent visitor. The challenges of putting together the title base for one are daunting and, as I hope this post makes clear, so is providing and demonstrating persuasive value.

I can see only one player that might be able to pull off a more general subscription offering in the near term. (You can guess who that is.) The “whys” of that will be the topic of a future post.

One thing that is pretty certain is that when there are many publishers offering subscriptions in their niches (and someday there will be), they’ll each be powered by a Cloud-based service of one kind or another. None will be asking the IT department to create the software to handle it.

I will admit that I haven’t programmed anything specific about “subscriptions” into the “Book Publishing in the Cloud” program we’re running on July 26, but if that’s what any attendee wants to find out, they’ll have a great opportunity at our “Conversations with an Expert” session to get the answers. Almost all of the speakers will be available during structured chat time, as well as representatives from the great companies that are sponsoring the event. 

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Share on RedditShare on StumbleUponEmail this to someone

  Back to blog

  • Hello, Mike. I’ve been starting to think that this framework we have of thinking about publishers as niche or general my not be less substantial and more of a mental barrier than anything else. Even the most consummate niche publisher publishes in more than one niche and even the most general publisher makes its publishing decisions in part based on what of its niches are performing well. The difference in the distinction has less to do with what gets published than how the marketing team is aligned in it’s relationship to readers and consumers. I’d argue that general publishers are a collection of niches that aren’t acting that way when it comes to relating to readers and potential consumers.

    • Absolutely right. The way I say it is that “general” publishers have to think of themselves as “multi-niche”. And my stock suggestion for how they should think about themselves is to go talk to their special sales department and find out how *they* see the lists. They’re the only audience-centric component of a publishing house. They have to curate and customize for special interest stores and lists all the time.


  • I agree. In Spain there is some smoke-vendors trying it, but only the niches you can develop some bussines with a subscription model.

    • Spain is home to 24Symbols, which is trying to launch a general subscription model.


  • Hi Mike, 
    Interesting piece you wrote here. I have to say that I (for the first time) do not agree with you. I follow your theory on how it’s easier to make a niche subscription model successful, but I believe that in the end, that is not (only) what the customer wants. Apart from all the difficulties in arranging rights and getting all publishers alined, there is one big trend that can’t be overlooked: people are moving from ownership to access (when they want, where they want, how they want and for a good price and with a big title catalogue). What I missed in you piece, is a consumer centric view. It is mostly written from the publishers point of view. And that is one of the big things we have to learn (rather today than tomorrow): it’s about what the consumer wants (if he is aware of it or not). If you offer a good subscription model for a niche (Safari is a good example), you are probably the best solution for that type of reader, of, better said: for that type of book titles. But what if this person who uses Safari a lot, plans a vacation and wants some fiction titles to read when he’s away? He has to go to another service. If there is one. And what if, because of Safari, he has decided to only use subscription models for reading, and not to spend money on buying digital books anymore? My guess is that he will pirate them (because the legal options for now would mean a fragmented book collection). What do you think of a Spotify for Books with the apps (like they have it for music) included? The apps allow currators and publishers to add filters to the huge amount of titles available. But it also makes it possible for anyone to browse other genres if they want, or when they are tipped, or for whatever reason. 

    I’m convinced we have to take a good look at how people are using, consuming and paying for digital and in the last place think about how to overcome the ‘troubles’ it brings us (rights, many publishers, filtering, etc.).

    • Timo, you’ve built your logic around the assumption that somebody using Safari will be so hooked on subs that they’ll steal their pleasure reading. That’s beyond a stretch; it’s a leap.

      I’ll do a future post on general. There are plenty of good reasons it will take a LONG time for anybody to aggregate what they’d need to make a general subscription model work. Except for one player, of course.


      • Not quite, I assume that the trend we see in general (Spotify for music, Netflix for video), will also apply for books. And both Spotify and Netflix are (for a large part) successful, because they offer ‘all’ the titles available. You can do that for a niche of course, which will be easier. But there will also be a demand for such a service in general trade publishing. And you can’t call that a niche. I fully agree that something like that would take a long time, and that rights agreements with authors should be possibly altered (which can be difficult enough in itself), but that doens’t mean that the public, the reader, is changing into someone who only wants access (and preferrable can find every title at one single service) to content, instead of owning it. How we (as publishers, etc.) deal with this, is something completely different.

      • Although it is heresy in some quarters to say this, the consumer can’t always get what the consumer wants.

        For example, I can assure you that there are *many* consumers who would like their local Borders to still be open and their local Barnes & Noble to stock as many print books as they did three years ago.

        Sometimes, commercial realities rule. This will be the case here.

        You will note that I said that I didn’t think a general offer that made any sense could be put together by any company except one any time soon. I’ll say that’s five years.

        Could we have a thriving “general” subscription model 20 years from now? Certainly nobody knows, but I doubt it.


      • That is absolutely true, and sometimes consumers don’t know what they want, until they’re presented with something new of which they (then) think they never could have lived without. 

        But there is one big problem for publishers, and that is that the digital books can be found everywhere. Legal or illegal. Which is a huge difference with for instance music and games back in the days when digital started to grow mature. Now there is a complete ecosystem (called Bittorrent, or whatever equivalent), which will be used if the legal offerings don’t meet their needs. 

        So, to come back to your reply, if we don’t do what the consumer wants, the consumer gets it elsewhere. And, to make the comparisson one more time with music, Spotify helps fighting piracy in Sweden by now, and is an example that if you offer a good service (what the consumer wants), for a good price with ‘all’ the titles, the rights owners can make (more) money with digital then without it. And I personally do believe this will also apply for books in the future. Not solely, but it has to be one of the tastes of digital that consumers can consume.

      • Timo, see the last comment. It’s an open question whether with *this particular *digital solution, the artists are really making more money. Or enough money.


  • Do you see an opportunity for a “social club” where groups of say 5 can share titles for a subscription fee?

    • The problem for any offering of that kind is aggregating enough titles of interest to the group to make the offer attractive. For reasons I’ll get into at another time, I think that will be very hard in the “general” space. But what you’re proposing could certainly work if it were offered by one of the many niche players I cite in this post.


  • Kassia Krozser

    Mike, I agree with your larger point about subscription-based business models. For general trade publishers, it would only work if they have very strong niches within their overall list. Most don’t (Tor, for example, might fit this bill, but their offerings may be too diverse for subscriptions…then again, maybe creating an “editor’s choice” or selection by Patrick Nielsen Hayden would be a great entre into this model).

    You are, however, missing a big part of the motion picture industry in your discussion about granularity: television. It is, arguably, a huge force in the subscription business. Netflix and iTunes see great business due to TV Product. And legacy subscription services like HBO aren’t getting subscribers based on their feature film offerings — it’s original television programming driving new customers.

    When you factor TV into your analysis (and you really have to) plus direct- to-video, you see the output of the motion picture industry in the US alone is pretty massive.

    (I do think quality can be a strong basis for a subscription model, especially if the publishing house has a clear, well-defined point of view. Am curious to see how Dzanc fares. Dan Wickett has spent a decade building his audience.)

    • You’re right that I skipped TV and shouldn’t have but I don’t think it changes the granularity argument very much. You’d know better than I how many SKUs are added by adding in the TV shows but I don’t think it would put the Hollywood offering level anywhere near books. It might get it closer to music, though.


  • Pingback: Spotify is Revolutionizing Music Consumption — Why Books will go the Same Route — McCarthy Digital()

  • Giulio Blasi

    Let me anticipate a couple of question for your next post 😉

    Why only one company and not, say, 3 or 4? Since Cloud platforms will be crucial don’t you think that after a first period of practical monopoly in this area, large media companies and national telcos could perfectly work together to compete with the few global incumbents (not sure if this word is used in english in antitrust discourse) that now dominate the market?

    It’s weird posing questions on a post that has not been written yet but, hey, this is true fidelization 😉

    • The greatest challenge to attempt a general model will be aggregating the content. Does that give you a hint?


      • Giulio Blasi

        Sure, and it suggests we have at least 4 possible players in this game (restricting our scope to ebooks offerings only, I mean excluding complex subscription mixes with audio + video + ebook + newspaper + whatever): Amazon, Google, Apple, B&N + MS. It also suggests (since we have strong national segmentation in the trade publishing sector) that national publishers should put all their efforts in creating national aggregators to keep control (and the related margins) of this game…

      • True, if assembling the content were just a matter of contact with the publishers, then you picked the right competitive set. But there’s another component: the willingness to lose a lot of money so that you can get content people don’t want to give you. I think that narrows it down. I’ll put up a post, probably on Monday, that gets into that in more detail.

  • Another Steve

    Last year, Sam Rosenthal, the founder of the indie label Projekt Records, said that Spotify and similar services were “not a viable way forward for the music industry.”  He withdrew all of the label’s  titles from Spotify and the other on-demand music streaming services.  Projekt was far from the only label to do so.

    Similar complaints are easy to find on the web.  The main complaint is that the services pay tiny amounts of money to the rights holders.  According to one story, a million plays of Lady Gaga’s big hit single “Poker Face” on Spotify earned the grand sum of $167.  

    From the standpoint of the musician/record label, would be worth doing if most subscribers used the services to explore new music, and bought what they liked from legitimate sources.  But if they listen to Spotify instead of buying music, it’s no good.  Many rights holders are making the determination that the latter is the more common and likely scenario.  

    • And Mr. Rosenthal is absolutely right. I think Audible is under similar pressure. Whether the subscription models are sustainable long-term is still an open question. The Internet is too young to draw hard-and-fast conclusions about viability from the evidence we have so far.


  • Excellent post, and I think that it frames well the distinctions between different media types that drive consumption and monetization dictates. 

    It was an easy evolution to unbundle the song from the album because the unit of value still worked (arguably better) at the level of the song. 

    I think that with books the challenge is to define the unit of value whereby the whole book can be ‘systemically un-bundled’ into smaller units, and then aggregating into a larger library so as to create the kind of deeper value that supports a subscription model.

    Reference books are logical places for these things, as the Safari example underscores, as are segments like education and business where there are new types of media/engagement units that can be cobbled together to create new value (e.g., cliff notes, best practices ‘recipes,’ quizzes, etc).

    The challenge for most publishers contemplating this path is that there is both a material cost to re-factor content AND an income risk that new ventures in this arena don’t cost-justify.

    It’s a bit of a Catch-22 for these folks, as the clear trend for these folks is getting increasingly squeezed by Amazon, especially if/when there is no more B&N.

    • Thanks for the comment. I agree.

      I addressed this question in a post a couple of years ago that I think is one of the more important ones in the history of this blog.
      I take note of the discouraging word at the end of your piece; I think we have some time before B&N actually “goes away.” The $600 million committed by Microsoft should cushion the commercial realities for a while.


      • I will check out that post. Thanks. 

        As to B&N, I think the real issue is viability of the big box footprint. I can tell you that they just are not doing the sales per square foot to support their current configuration. Their future lies more in a B. Dalton sized model, but that’s not easy path to get to from here.

        Just because you have cash in the bank doesn’t means you can burn it simply to keep warm. Also, one assumes that the MSFT investment (I think it was $300M) is more geared toward NOOK than bricks and mortar.



      • Microsoft put in $300 mil as an investment but also guaranteed $300 mil more for uncertain return and payback over the next couple of years. It is $600 mil in cash for B&N from MS.

        You’re right about the smaller footprint. I wrote about that too.

        Nov 8, 2009:

        Jun 7, 2011:


      • This is the part of the post where I say that I am not a paid spokesman for you, serving up alley-oops to posts you have already written. 🙂 Cheers, Mark

      • Well, just for that, I pulled your check out of the mail!

        Thanks for your comments. You’re welcome here any time.


  • Pingback: Explaining my skepticism about the likelihood of success for a general subscription model for ebooks - The Shatzkin Files()

  • Pingback: 4 Readings this week « Allegany County Library System Director's Notes()

  • Pingback: The OPLIN 4cast » Blog Archive » OPLIN 4Cast #305: Paying for it()

  • Steve

    Hi, Mike. Could a whole country be a niche itself? If you had all the publishers in one medium-size country working together in the same project, under the same umbrella, with the same strategy and pricing, media on their side and a strong brand supporting it, could it work?

    • The key for a publishing niche would be “common audience”. So my first reaction is that it would work for something like “Wines of Australia” but I think it would be hard to make it work across all topics for a country.