Two Bits at Six Months

Last June I announced that I had published my book, Two Bits: The Cultural Significance of Free Software. It was released both as a book by Duke University Press and as an open access publication via a website that I created and maintain. For scholars in my fields—anthropology, history, science studies, media studies—this is one of the first experiments, if not the first, of this kind. As such, I’ve been doing my best to keep some notes on the process, with a mind towards reporting on the results of going open access with a first book.

Herewith, therefore, are two reports generated by Google Analytics, which is hands-down the most un-evil thing Google has ever done (General Report | Traffic Source Report). These reports are chock full of information, beautifully organized and fascinating to explore. Unfortunately, they are also pretty hard to interpret. I’m posting them now, because I think they show a few things pretty clearly, such as the initial spike of interest, the fact that 4 times as many of my readers use Firefox as do Internet explorer, the role of small communities in creating attention (savage minds, hastac, and a handful of close friends account for a significant portion of the traffic to the site).

My book is in kind of a strange space. On the one hand, it is a conventional academic book, a first book by an assistant professor (now tenured, thank you very much Duke University Press); it is accessible, but not popular; it has a large potential audience beyond academics because of its subject matter; and it is beautifully designed and people tell me it is well written. So much for the pro column. In the con column: it is long, it contains complicated theory in the first chapter, including Habermas, which is fatal to any reader even in small amounts; it doesn’t have any sound-bitable arguments and people tell me it is poorly written.

In short, it’s a pretty standard academic book, and therefore a good candidate for this experiment. People always ask what I had to do to convince Duke to let me release the book. On the one hand the subject matter made it easy: I couldn’t respect myself, or Duke, if a book about Free Software were not freely available. On the other hand, I think they were really eager to experiment, to see what would happen. I created the website, so they didn’t have to; and they agreed to use a CC (By-NC-SA) license and to give me the pdf and a very clean HTML copy (thank you Achorn) for distribution. The designer, Cherie Westmorland, used an open source font and the Boston Public Library let me use the cover image. All told, things worked out swimmingly, and the whole process has been, well, entirely normal. Duke is making as little or as much money on the book as they do on others of its ilk, and yet I am getting much more from it being open access than I might otherwise.

So what have I learned so far? A few things:

1) The Internet is dead. Well maybe it’s not that bad, but the era when simply putting something online guaranteed orders of magnitude more readers/viewers/listeners than normal is long gone. To put a finer point on it, let’s say the ‘Age of Boing Boing’ is dead. Sorry University Presses, you missed it. The place is just so saturated with everything and everybody that it now feels more like normal life and less like some special place. This amounts to saying that things have returned to normal levels of hard work. To get a book to sell, one has to invest a lot of work in marketing it, promoting it and distributing it—but all these things now include new forms of marketing promoting and distribution online. Just putting a book online means nothing unless one is going to work hard to bring attention to it (a fact Rex has noted repeatedly as well). How do I know this: because the Google reports tell the story. All the spikes in traffic correlate precisely with mentions in major and minor media outlets, ranging from Savage Minds to the New Yorker. Placing links in widely read places (print or online) increases traffic. Full Stop. But more than that, I know this because the ratio of print sales to downloads downloads to print sales has been 3 to 1 (Thanks, Cathy for the correction) . Not 1000 to 1 or even 100 to 1, but 3 to 1. That’s kind of amazing. It means that neither my outsized expectations of hordes of geeks downloading the book, nor Duke’s fears of massive numbers of lost sales have come true.

2) I have tenure. Putting my book online did not ruin my career. Having Duke publish it, as opposed to, say, some online vanity press, contributed to my tenure case, but simply having it available for free is not career suicide. Quite the opposite, I would say. I have more requests now for talks, reviews, contributed papers, conferences, interviews and projects than I can accept, and probably more than half of them come from people I don’t know from Adam, which means people who have found the book in public rather than through connections with my peers and friends. Lots of people are assigning the book in class, or bits of it, which I can only assume is facilitated by the ease of access. Duke, of course, might not like to hear this since it means people are assigning the book without ordering copies for class, but I’m ambivalent. On the one hand, I would like those people to assign the whole book and for Duke to be remunerated as a result; on the other hand, I know what creating a syllabus is like, and how great it is when something can be added just by inserting a link, as opposed to dealing with bookstores and administrative systems for ordering the book–a book students may or may not buy anyways.

3) I’ve had a pretty excellent amount of media attention. There are books it might be compared to that have done better: Jonathan Zittrain’s book came out at the same time, and he was on the Colbert Report, as was Clay Shirky. But as much as I love Colbert, that’s exactly the opposite of the kind of attention I would want. I have no “message” which I want a hundred million people to hear; I have a scholarly book which I wish Zittrain and Shirky would read, not Colbert and his audience. Nonetheless, I have had mentions in The New Yorker Blog, The Times Higher Education Supplement, Technology Review, Inside Higher Ed, and others. I’ve had conversations with people from Korea, Argentina, Brazil, and India about the book. I’ve had excellent response from European scholars interested in the book. In short, I can’t complain. According to Duke, the amount of marketing that went into my book was more intensive than most, and this may no doubt accounts for some of that attention. Frankly, it’s more than enough. I’m not quite sure what I would do with more, but I do know that with a bit more marketing, the dynamics of attention might conceivably change much more dramatically than just ten years ago. For some books that university presses publish, this fact is worth mulling over.


Christopher M. Kelty is a professor at the University of California, Los Angeles. He has a joint appointment in the Institute for Society and Genetics, the department of Information Studies and the Department of Anthropology. His research focuses on the cultural significance of information technology, especially in science and engineering. He is the author most recently of Two Bits: The Cultural Significance of Free Software (Duke University Press, 2008), as well as numerous articles on open source and free software, including its impact on education, nanotechnology, the life sciences, and issues of peer review and research process in the sciences and in the humanities.

19 thoughts on “Two Bits at Six Months

  1. Thanks for the update. I wonder if the real value of the internet won’t show itself for a while? First, there is the long tail phenomenon. At the point when the typical academic book might be out of print, your book will still be there online. Second, it might be that the real effects are felt when your book has been digested by those currently reading it (or planning on reading it) and they begin to make use of it in blog posts, other books, etc.

    I know that our experience putting our film online for free is still paying off for us. I can’t say about sales, simply because I’m not sure of the statistics, but certainly in terms of creating new connections and opportunities. Then again, a 15 min film lends itself better to online viewing than an academic book.

  2. Hi, Chris, This is really interesting–and I’ll reblog it on the HASTAC site. One correction: I think you mean to say that the ratio of downloads to print was 3:1, not the other way around, right?

    We are finding more and more that “the Internet is dead.” Authors now (ie print authors) are all wanting “a website” as if that means anything without a ton of legwork and even more luck. I don’t have to tell you fab Savage Minds folk that you only get attention if you have a product that inspires attention and a network that needs to attend. Good luck with all you do!

  3. Great post, Chris! Having your thoughts on this ongoing experiment has absolutely been one of the biggest payoffs as far as the Press is concerned. The balance between “the internet is dead” and the ways it makes a difference to have the book available there seems right. I wonder how much this is the new floor, and how much the attention is still about the novelty of what you did for this kind of book. Does it become the next version of those forlorn and unvisited web-sites to go with a book; or a necessary version, that people will expect to be there?

    We would prefer students bought the book — even if they didn’t read it in class as assigned. Kerim, I don’t know about the “typical academic book” but except for an art book or two, everything I’ve worked on at the Press since 1991 is still in print, so it’s easy to get a fifteen year old book. We’ll see how easy it turns out to be to get a fifteen year old link to work!

  4. Chris thanks for this great report on the life of the project. It seems to reinforce what a lot of people already knew, or at least suspected: the website and electronic text promote the physical text. They also facilitate quotation and searching of the text. However, the people who care about the book buy the book and are do not opt for the ‘cheaper’ PDF download (clearly, there is a utility in the paper book that people who fear ‘free PDFs’ don’t get).

    One thing that you didn’t mention was whether there was any uptake of the books advanced web x.0 remix features. I’m guessing that people will ‘remix’ Two Bits in the way that scholars always have remixed each other’s works: citation, influence, reviews, and so forth in established academic genres. The right mix of motivations and arenas necessary to produce the Beyonce/Kelty Youtube mashup of the book probably don’t exist. Which is not to be dismissive of Kelty or to argue against making the book available for remix.

    But again, thanks — I think this experiment is a great demonstration of how unfounded so many fears about open access truly are.

  5. Chris, I was just wondering, given your experiment in making the hold-in-your-hand-turn-the-pages-put-on-your-shelf book two bits also available online, whether you’d considered the inverse experiment – making the online-only document Hau to do things with words also an hold-in-your-hand-turn-the-pages-put-on-your-shelf thing, perhaps something like a pamphlet?

    This is mainly a theoretical question (although I probably would buy a copy), because I’m wondering whether part of the issue here is that some people hold to an evolutionist idea that print slowly gives way to electronic media, leaving print extinct and everything computery triumphant. But of course the story, as you are showing, need not be an evolutionist story – and one side of this is that technology also provides new opportunities for making available in print things that might not have been as easily available in past generations.

  6. Richard… that’s an absolutely fascinating idea. I guess because that paper was impossible to publish given its length, style and format, I had just consigned it to being something like a sketch of later work… but I also know that some people still read it. Creating a pamphlet would give it a different kind of life… much like the scholar who publishes 10 articles and then re-publishes them as an edited volume of writings. Clearly they are both “published” but the two objects have a different value.

    Another re-mix suggestion I’ve had (from Biella Coleman) is simply to re-order the chapters of Two Bits, with perhaps minor changes to the intro, to create a book that is more geek-friendly–one that doesn’t start off with Habermas, but with UNIX. Since I spend all my time writing comments on my own posts at SM, I of course have no time to do this 🙂

  7. I found the traffic source report especially interesting because it implies a key mix of source types that are crucial in order to get over the “internet is dead” dynamic that Chris is describing. The mix of types seems to have a hierarchy of importance, but the breadth appears important. There is high profile media coverage, there are formal promotional mechanisms (the Duke blog, for instance), there are bona fide blog discussions, and then, and this is what interests me most, there are the shout-outs of friends and well-wishers who maintain a modest web presence and have some kind of small public or at least enough of a presence to attract keyword searchers who can then be captured and passed on. I show up as one of several such shout-outers on Chris’ report and my scan of the data suggest that, in aggregate, such well-wishers make a difference.

    Real web-people (I am not one) I guess know all about this stuff (its called link-love, right?), but for scholarly projects its seems to be both crucial and a mutation in time-honored citational practices.

  8. Shortly after Chris announced that _Two Bits_ was available on line, I downloaded and began to read it. Then, I became distracted as other matters intruded. _Two Bits_ became yet another PDF somewhere on my computer; but with nothing to remind me that it is there and no on-going conversation about it popping up in my email, it wasn’t until this thread began that it caught my attention again. I could, of course, say something similar about the dead tree books on the shelves beside and behind me. They, however, have the advantage of eye-catching covers when I need a break from the computer screen and scan the shelves to rest my eyes. I guess what I’m saying here is that, without a particular need or reminder, online material tends to slip away and lacks the visible permanence that the books on the shelves enjoy.

    The marketer in me reflects that I might have been a good idea to offer popular bloggers and/or vocal discussion group participants a preview of the book before making it public, to build and sustain interest that would mitigate the distraction that affects anything we see on line. This is the sort of thing that Apple, for example, does very well, stirring up interest in advance so that when the new product is launched, there are already on-going conversations about it that newcomers can join.

    Just brainstorming.

  9. Chris,

    There seems no way to count those, like me, who both downloaded the book and bought the print copy. Each version has its own utility. One day I hope publishers will bundle print and digital together rather than hording their digital riches or, worse, giving them only to the Google-zon, which is not ultimately in their interest or mine.


  10. You’ll notice that I (pointedly) said of the website that I “created and maintain it”… I would have loved for Duke UP to be the subject of these predicates, but they do not yet have these capacities, and as far as I can tell in my recent experience, neither do Cornell University Press, NYU Press, Palgrave or Paradigm, and probably most other university presses. This fact is, at minimum, troubling. It’s obviously not sustainable to expect each academic to create and maintain the digital version of their own publications. We are already our own secretaries and any number of other things. And yet, will this ever become something university presses routinely do? I certainly hope so, but I don’t see much movement it that direction yet. I know that within a year or two, I will have reached the maximum number of websites I can maintain for these purposes, and I will either have to stop publishing or find another way to make my digital me available…

  11. I really like John McCreery’s observation of actual bookshelves vs. digital archives of pdfs. I know the library community has had lots to say on the benefit of serendipity in browsing open stacks. For my own experience of the library this has been tremendously important (and frustrating when faced with closed-stack libraries in Russia). Amazon and other on-line booksellers try to replicate this through things like “what other people who bought this book bought”. Such problems could probably be solved quite easily through some clever software that serves our pdf collections more dynamically.

    Chris’ experiment is also interesting to me because we are on the cusp of an actualization of portable/readable electronic books. It seems this year or the next we’ll see a host of new e-paper/style devices that will offer both affordable and powerful alternatives to traditional books. When that happens we’ll really see the circulation and sale of digital books ramping up. For now, I still prefer reading from a conventional book itself than from a screen.

  12. …so what are those 3:1 figures? Exactly how dead is the internet, how dead UPs, and how do your numbers stack up against UP books that aren’t marketed as heavily by Duke to the New Yorker Blog, The Times Higher Education Supplement, Technology Review, Inside Higher Ed, and others?? Has “normal” changed, in your remark that we have returned to normal?

  13. 3000 to 1000, so far. The first number is the number of hits on the pdf, so it doesn’t count any of the other formats online. The book sales includes, according to duke, a relatively high number of hard-bound copies as well. The interesting thing to watch will be to see whether the ration stays at this level or steadiily increases… but that depends a lot on whether my book is forgettable or simply forgotten 🙂

  14. A thousand dead-tree books is not bad at all. It has been some years ago, but I vividly remember a piece in the _Chronicle of Higher Education_ in which the author, an editor at Harvard University Press, let drop the remark that the average sale for a monograph in the social sciences or humanities was 700. My own book, _Japanese Consumer Behavior: From Worker Bees to Wary Shoppers_, U. of Hawaii Press, 2000, has sold a bit over 3,000 copies. That is, of course, in part because it was written deliberately to appeal to people who do business in Japan as well as academics in a variety of disciplines associated with Japan studies. The most positive reviews have been from a Japanese sociologist and an Italian political scientist. It certainly didn’t hurt either that the Japanese ad agency that created the research institute whose researchers provided the eyes through which I was looking at social and cultural change in Japan bought 250 copies and distributed them to its clients.

    I note with interest Craig’s remark that,

    bq. Such problems could probably be solved quite easily through some clever software that serves our pdf collections more dynamically.

    We are already very close. I am currently using a combination of Evernote (for random stuff picked up off the Net) and Zotero (for more systematic note-taking and bibliography generation. Now, if somebody could find a way to do for PDFs what Zotero does for library collections, i.e., grab and catalog the references, ideally with filtering to sync just the new ones, it would then be possible to quickly assemble the network of references to which a particular PDF belongs….It could then become the mark of serious scholarship to have looked at least two steps out from the primary inspirations of one’s own research…

    Again, just brainstorming.

  15. Adobe’s Digital Editions is, I suppose, meant to serve your PDFs in a more useful (or at least visual, bookshelf like) way – an iTunes for eBooks. But last time I used it, it wasn’t quite where I needed it to be to actually find useful.

    The thing I find most profoundly irritating about PDF versions of print media (journal articles) is that the content providers do not fill out the Metadata fields for the files they serve – I download something from JSTOR, look at the file properties and find the Title, Author, Subject and Keywords fields blank, blank, blank and blank. If they were filled out and I had a filter for my Bib software I could scan all of those files on my HD and autoimport the lot. Does anyone know if something does this?

  16. Terrific post and comments. Kudos to Duke UP (hi, Ken!) for this. Very hard, as you all say, to measure these results precisely but good to have more evidence.

    I work with NYU Press and NYU Libraries in a new (just one year old) position to develop digital publishing services. Very early days, but we’re envisioning projects that to publish scholarly materials in *both* print and online, with their distinct affordances. Such press/library collaborations are beginning to emerge. The challenges are many but the opportunities enormous. Time will tell…

    @ John McCreery and Craig: Check out Mendeley ( a tool that lets scholars organize and share PDFs, including a way to grab the citation. Still in beta (and funded by venture capital, so who knows where it will go) but looks promising. Also see Papers (now in version 1.8.6) which is Mac-only and winner of an Apple design award.

    Both built by European grad students looking for a solution to manage their research material and connect with people working on similar stuff. Copyright issues may emerge if publishers object to the public posting and peer-to-peer sharing these tools seem to enable. But obviously a need waiting for a solution.

  17. Wow just playing w/Papers its really fun and useful — just about the right weight of app for me. Thanks for the recommendation Monica!

  18. Papers looks lovely for those who have access to repositories via their universities. This independent scholar is still looking for someone who will package access to them at a reasonable flat flee.

  19. @ckelty:
    bq. It’s obviously not sustainable to expect each academic to create and maintain the digital version of their own publications. We are already our own secretaries and any number of other things. And yet, will this ever become something university presses routinely do? I certainly hope so, but I don’t see much movement it that direction yet. I know that within a year or two, I will have reached the maximum number of websites I can maintain for these purposes, and I will either have to stop publishing or find another way to make my digital me available…

    This is a role for university libraries – probably the home institution of the author, but I suppose it could also be the library of the university affiliated with the press.

    Lots of librarians have been working terribly hard to start institutional repositories to make their faculty’s work openly available, but faculty haven’t had a lot of knowledge or incentive to contribute to them. (See pretty much anything written by Dorothea Salo.)

    Your experience points out why repositories are useful – open access is good for scholarship, but creating the access is a lot of work. Maybe your next book can be online at

Comments are closed.