PeerJ launches
The indefatigable Peter Binfield, formerly of PLoS, has launched a new venture: PeerJ. There’s a very good interview up with Binfield and co-founder Jason Hoyt at John Dupuis’s blog that helps the curious gain a sense of the venture.
There has been some chaff here and there about it from the usual suspects. The Loon is not impressed. She’s learned to take attacks from certain quarters as fear that the venture under discussion just might work. If it weren’t viable, they wouldn’t bother expending pixels on it. Enough of that, then.
The revenue model here is quite novel. Rather than charging per paper, PeerJ charges per author: a one-time fee (presumably adjusted for inflation over time) gains perpetual right to submit papers for review (though not to have papers accepted and published; PeerJ’s one-page elevator pitch would do well to make this clearer).
The Loon must assume that Binfield and Hoyt have done some analysis of PLoS’s author field (especially articles-per-author and authors-per-article), such that they believe this viable. Supplemented by the usual smorgasbord of institutional memberships, grants, and (perhaps) other freemium services or advertising, it might well work, if the author fee has been correctly set. It’s a model not unlike the profitable service Pinboard, whose founder had sharp, smart things to say about ventures that don’t charge service beneficiaries for service.
It could come a cropper, financially. The biggest unknown in the Loon’s mind is the cost of production. PeerJ is promising the entire format suite—XML, HTML, PDF—which means a complex and expensive typesetting system for articles that make it through the gauntlet. This must be outsourced (the Loon can’t imagine Hoyt and Binfield doing it all by themselves!), which puts the costs somewhat out of PeerJ’s control. While typesetting to PDF and image management are at rock-bottom commodity prices, the Loon believes markup-based workflows aren’t.
Cleverly, though, PeerJ has mitigated one potentially serious business risk: a rush of poor-quality submissions from developing nations seeing an alternative cheaper even than the likes of predators like InTech. Because PeerJ is honest, it will accept such authors’ membership fees and then reject their submissions as many times as needed, avoiding the expensive editorial work needed to bring such submissions to adequacy.
This does, of course, mean that PeerJ won’t put the likes of InTech out of business. This is a sincere pity; someone should. Nor will PeerJ solve the problem of developing nations’ cargo-cult high-impact-factor worship.
If it sticks, however—and as the Loon said previously, she wouldn’t bet against Binfield, who’s the smartest in the business—it could do yeoman’s work cutting author-side fees at other publishers, particularly the cynical big-pigs’ “hybrid” programs. The Loon would very much like to see that. She wishes PeerJ all the best; the BAE is likely to buy a membership just to support the venture.
In some fields, like computer science, researchers are already used to doing their own typesetting and layout. The publisher provides the templates (a LaTeX style file, or a Word template) and the researchers put in their content and send the publisher the result.
If they’re planning on relying on that, and then auto-generating the PDF, HTML, etc., automatically, their costs can stay quite low. (The results won’t look as pretty as professionally typeset journals, but computer scientists have found them quite serviceable in their field.)
They’re attacking biomedicine first, though. That means both complex typesetting/markup problems (math, tables, charts, graphs) and a non-LaTeX culture.
The whole LaTeX thing gets brought up in too many contexts where it’s not relevant, if you ask the Loon (which you didn’t). Author typesetting works only in an extremely limited range of disciplinary contexts. Reusing author keystrokes works in more disciplines, but is also somewhat more labor-intensive for publishers (the horrible messes authors make with Microsoft Word must be seen to be believed).
Even in computer science, Word is often only reluctantly accepted; as you point out, you can make all kinds of structural mess in Word even with a template. You tend not to do that in LaTeX because structurally messy markup typically won’t compile in the first place, or is trickier to write than structurally clean markup.
The Drake, though, has seen non-LaTeX layout done on the cheap in biomedicine, though. For instamce, the NCBI Bookshelf formats its biomedical articles essentially as streams of lightly marked up text, with everything else (including math, tables, figures, etc.) simply dropped in as interpolated inlined images. (There’s also some special handling of citations, but nothing particularly complex.) It’s not not nearly as pretty to look at as a well-laid out professional journal article, but it’s readable and comprehensible. Not only is it less complex than LaTeX, it’s also the sort of thing that could easily be handled with a modern-day blogging or wiki authoring tool.
The Drake wouldn’t be surprised if that’s what PeerJ does for author-driven layout: Point authors to an blog-like authoring website and say “here, put in your text and drop in pictures of your figures and equations and the like”, and have the design of the authoring tool enforce a basic, parsable structure. Most computer-savvy folks can be expected to know or learn how to use such tools. And if they need some assistance, either with the basic layout functionality, or in making a nicer-looking, more complex or customized design, that’s something one could potentially charge extra for.