Image via Wikipedia
Publishers who boast about the number of publication downloads may be inadvertently admitting that the technical implementation of their service is limited — downloads, especially of PDFs, are often a rough proxy for offline viewing.
The metrics game for academic communications is a complicated and often a dirty one. What to measure? Citations? Impact factor? Page views? Downloads? Or perhaps the best measure is simple relatedness, where the keywords of an abstract are input into Google, and the search results yield a number for Web pages that may be relevant.
The notion that the quality of published work is quantifiable should give us pause; we have not learned that when Elizabeth Barrett Browning said, “Let me count the ways,” she was being satirical. Numbers are poor proxies for anything but other numbers, and the gap between a cultural phenomenon and the number that ostensibly represents it can lead to misunderstanding and even manipulation.
While the world of digital media is moving to ubiquity and integration, the shadow world of offline viewing persists. There are good, but short-term, reasons for this:
- Bandwidth is still not omnipresent, as anyone who tries to get a high-speed wireless phone connection in my hometown can attest.
- Some people prefer to download and print out articles, tomorrow’s atavism, since handy digital reading devices (iPod, Kindle, etc.) are only now getting a toehold in the market. What happens offline remains offline.
Offline viewing removes content from the evolving, integrated digital world. It works against the interests of authors, who want their work to be read, referred to, and commented on where it will have the most impact, and that is online. Authors increasingly will want their content to be embedded in a networked world that, from the point of view of technical implementation, more closely resembles Facebook than library stacks.
When you say that your PDF downloads are up, you are acknowledging that your Web service facilitates a high degree of offline viewing. This in turn means that your service is not fully, even relentlessly, integrated into the ongoing communications medium that is the Internet today. Readers have to make an effort to comment online about what they have read offline. Far better to have the tools for annotation (public and private) and citation right there, on the Web page, where the immediacy of digital media can increase the engagement with an author’s work.
We read over and over again about all the low-cost ways to publish scholarly material. We don’t (thank god) read much about the low-cost ways to deliver a baby or the great bargain we got on open-heart surgery. Some things cost a lot because we value what the expenditures bring us.
Scholarly communications will increasingly involve more and more sophisticated technical implementations because technology is one means for an author to reach wider audiences. No doubt these new tools will generate new metrics, which will themselves be subject to distortion. Authors, however, will increasingly evaluate publishers by asking, “How many people are in your IT department? What is your overall IT budget? What is your roadmap for technological development? May I speak to your CTO as well as your editor-in-chief?”
When a publisher replies by citing PDF download statistics, the author should smile politely and say, “No, thank you.”
Discussion
6 Thoughts on "Downloads as Failure"
Joe,
– Perfectly true from what I experienced. I talked about author service as an important aspect of online publishing end of 2007 on an ALPSP seminar – as a theory at that time. Just a few months later it became evidence.
During the development of the de Gruyter eBooks platform editorial colleagues started occasionally dropping by my office with authors (humanities and linguistics!). Asking for a three minute presentation of how books would be presented in future on our platform.
Authors were absolutely keen on having their books published with the kind of interactive functionality that initially had been developed for electronic journals. And colleagues told me later that demonstrating our eBooks functionality in a beta version helped a lot winning those publication projects.
best – Joachim
A very thoughtful piece, although I wouldn’t go as far to say that PDF downloads are indicators of failure, they just reflect a different kind of reading.
Most of my online reading is more like skimming. Before I go to a graduate seminar, however, I print off articles, scribble on them, underline, circle, write notes (and sometimes expletives) in the margins.
Not just old graduate students like me do this — 20-something grads do it as well.
Instead of starting into our laptops in class, we look and engage with each other. While this type of interaction with the text may not be captured in statistics, this human-to-human type of interaction is a much more valuable form of pedagogy than a human-to-machine interaction.
I agree with Phil (please don’t stop reading my comment on this basis alone!) 35% of our online users tell us that they go to our Web site to print our PDFs. This may be because they want to take notes on the paper, this may be because they want to read in the bathroom which is a wireless dead zone or this may even be because they just like the feel of paper. I don’t think this means our Web sites have failed. If anything, I think it means that we have succeeded in getting a print user to an online medium…even if only for a few minutes.
I envy future generations who won’t know what “a PDF” is, or may look on it with the same mixture of amazement and disdain we’d reserve these days for Etch-a-Sketch.
As a student on an online Master’s course, I’m resigned to wrestling with “surrogate A4” – including the worst type of all, text-less TIFFs embedded in PDF. I rejoice when I do find scholarly resources that use the web intelligently – blog-wise, wiki-wise. The InnovateOnline journal is one of my current benchmarks for doing this kind of thing well: I can even easily read the HTML-formatted articles on my mobile.
As developers of repository-class systems, though, PDFs are still our bread-and-butter, from the users’ POV: but if we stand our ground and put our case for better ways the tide might eventually turn.
I think this is just wrong. Joe’s argument assumes that downloading equals printing, since he equates it with “offline viewing.” But anyone reading an article on screen has “downloaded” it; this has nothing to do with printing or offline viewing and is, in fact, a legitimate measure of the use of content — and therefore a valid metric for publishers to use. All articles have to be downloaded to be read; how users read them, online or offline, is a completely separate issue.
I always enjoy your posts and articles, Joe. You write with style and wit, and that’s a treat in this business. Though I enjoyed reading this post, too, I must take exception with some of your points.
Elsewhere (liblicense-l), you have said that “every time I see a PDF I reach for my revolver” (I’m pretty sure you did not intend for this particular allusion to be analyzed very closely). Clearly for you the PDF file has become synecdochic for a laggard and risk-averse scholarly publishing community not fully embracing the latest online technology.
This is really your argument here and not the validity of measuring usage with PDF downloads, which is a perfectly reasonable and legitimate practice in most cases, as Doug LaFrenier rightly points out. Although you imply it, I don’t know anyone who argues that downloads measure quality. COUNTER certainly doesn’t.
I will grant you the premise that a lot of STM publishers, having made the transition online, are still bound to the print paradigm in many ways. The PDF file can, but doesn’t necessarily, represent this state of affairs–and may very well, as you say, become “tomorrow’s atavism.” (No demise, however, has ever been more exaggerated than that of print in academic publishing.) But the PDF file is still a good representation of the vast majority of articles and still has value for those who do not want to read articles online (there are legions of these) or spend $350 for a Kindle (there are legions of these, too).
And in a great many cases, as you well know, PDFs are a complement to the interactive, full-text XML/HTML versions. So even if we’re talking about printing PDFs, it’s not always a case of either/or: offline viewing or “integrated digital world.” The two things can and do co-exist very nicely, thank you. More important, they serve users in different ways, as others here have noted.
Finally, you are probably overestimating the desire and the willingness of authors (outside the peer review system where they are credited) to annotate or comment on what they read online (this can be done in a PDF, of course) for the benefit of their fellow readers and researchers. The principles that drive Web 2.0 continue to be misapplied to scholarly publishing. But that’s another blog.”