Saturday, April 05, 2014

Interview with Jean-Gabriel Bankier, President & CEO of bepress

Founded in 1999 by three Berkeley professors, bepress (formerly Berkeley Electronic Press) spent the first decade of its existence building up a portfolio of peer-reviewed journals — much like any scholarly publisher. In 2011, however, it  took what might seem like a surprising decision: it decided to sell all its journals to De Gruyter and reinvent itself as a technology company.
Jean-Gabriel Bankier

Instead of publishing journals, bepress is now focussed on developing and licensing the publishing technology it created for its earlier publishing activities, and its flagship product is a cloud-based institutional repository/publishing platform called Digital Commons.

Digital Commons is currently licensed to more than 320 academic institutions, who use the software to publish over 700 journals, 94% of which are open access. This publishing activity is invariably managed by the institution’s library, and often includes the publishing of books, conference proceedings, data sets, audio-visual collections, and other digital content types too.

Is this a sign of things to come: Publishers becoming technology companies and librarians becoming publishers? President and CEO of bepress Jean-Gabriel Bankier believes it is. As he puts it in the Q&A below, “Library-led publishing is an integral strategy in the university taking back ownership of scholarly communication.” As such, he adds, the future of scholarly publishing now “lies in the hands of libraries and scholars.”

To support his argument Bankier cites a US study in which 55% of the universities and colleges surveyed said that they are offering or considering offering library publishing services.

Moreover, bepress is not the only game in town for libraries looking for a publishing platform. In 2001 the Public Knowledge Project released the first version of the open-source publishing software Open Journals Systems (OJS), and today OJS estimates that over 6,000 journals are being published using its software. Many of these journals are undoubtedly being published (or soon will be published) by university libraries — e.g. the library at University College London and Stellenbosch University library

We could also note that in 2012 US-based Amherst College announced that it was launching its own press. This will publish peer-reviewed books in the liberal arts, and will be managed by the library. 

What all this means, says Bankier, is that if publishers “want to continue to play a significant role in supporting the changing needs of the research community” they will need to consider following the example of bepress, and morph from content provider to technology company.

Doubtless other publishers would challenge this assertion. But whatever the future holds, I think anyone interested in open access, or scholarly communication more generally, will find what Bankier has to say below of great interest.

The Q&A begins

RP: As I understand it, bepress was founded as a scholarly publisher in 1999. Can you say briefly who founded it and what the initial goal was? Is it for profit or non-profit?

J-G B: In 1999, UC Berkeley professors Robert Cooter, Aaron Edlin, and Ben Hermalin banded together to launch Berkeley Electronic Press, now simply called bepress.

The heart of bepress has always been about listening to faculty and responding with simple technology-based solutions that support scholars in the rapidly changing world of scholarly communications.

Initially, for us, that meant exploring alternatives to commercial scholarly journal publishing which were plagued by slow turnaround times, limited access, and unreasonable prices. Later, that meant providing authors and universities themselves with the means to publish their research openly and widely. We are a for-profit company.

RP: You say bepress is a for-profit company. I assume the shareholders are the three founders? Can you tell me what the company’s revenues and profits were for the last financial year?

J-G B: Yes, the founders are shareholders. The company was born with just a little seed money from the founders, parents, and incredibly supportive friends and neighbours, most of whom continue to own part of the company. Bepress has never had venture capital or private equity. Berkeley isn't far from Silicon Valley, but we weren't that kind of start-up. Our first office, after we moved out of one of the founder's kitchen, had no natural lighting and ceilings so low that it necessitated skidding around mismatching, three-wheeled chairs to avoid banging one's head on the ceiling.

I'm happy to report that around 15 years later, we've now got offices with actual windows and chairs that don't wobble. Our business doesn't wobble any more either. Our revenues are around $10 million a year with an unbelievably low cancelation rate for subscribers (below 1% in 2013). We are very stable and run at a modest profit. It is a great feeling to finally be able to send small dividend checks to those friends and family members who put their faith in us back in the beginning.

RP: How many journals did bepress come to publish, in what fields, and were they subscription journals or OA journals?

J-G B: We started with a few titles in economics, and eventually grew to publish roughly 65 journals altogether. In addition to economics, our emphasis was in education, health and medicine, law, policy, science and technology, and statistics.

We simultaneously sold subscriptions to libraries and made the articles freely available online. Many people suggested that we were crazy to use such a model, but it was really the best of all worlds. The authors and editors got to maximize readership and share their ideas, readers got access to high quality scholarship, and libraries got a sustainable journal subscription model. We raised prices on our journals by a grand total of only three percent over 10 years.

RP: In 2011 bepress sold all its journals to the German publisher De Gruyter in order to focus on developing software tools to assist the research community publish its own journals. Why?

J-G B: The library community had little appetite for new subscription journals, but a big appetite to support publishing on their campuses. According to one study produced by the libraries of Purdue University, Georgia Institute of Technology, and University of Utah in 2012, “55% of universities and colleges are offering or considering library publishing services.”

We saw this change in appetite first-hand. In the one-year period predating the sale, we launched only four new journals under the bepress moniker. Over the same period, we helped libraries launch over 100 new open-access journal titles using our software. Starting new commercial journals had become a challenge. By our count Springer, Taylor & Francis, and Sage and Wiley together introduced fewer new journal titles to market in 2012 than the publishing programs of the Digital Commons’ community of libraries. 

Given this change, it became clear that the best way for us to advance scholarly communications was to focus our energy on supporting library-led publishing, so we exited the commercial subscription-based journal business.

It is worth noting that some of the editors who started their journals in Digital Commons had originally approached commercial publishers. One acquisitions editor from a major publisher looked at the list of journals that publish on Digital Commons and wistfully listed those that had approached her but were impossible to take on because of the economics of subscription sales.

Behind the scenes something very interesting was going on in scholarly communications. Libraries had made it more risky for publishers to start new journals. In turn, some publishers tried passing that risk onto editors, societies, and universities in the form of minimum number of guaranteed subscriptions or minimum annual guaranteed payments. Luckily, many of those journals had another option: to come to their libraries for publishing support.

Today Digital Commons is used by over 700 journals. We believe the future of scholarly publishing lies in the hands of libraries and scholars.

RP: I am particularly interested in the products Digital Commons and SelectedWorks. Can you say something about these and give me some indication of pricing?

J-G B: Digital Commons is a hosted platform designed for the full spectrum of scholarly publishing. We provide libraries the tools, training, and expertise to build essential publishing programs on their campuses.

Digital Commons supports journals as well as books, conference proceedings, data sets, audio-visual collections, and all other digital content types that showcase the whole of an institution’s publications and scholarly output. The content is all the institution’s own; we provide the platform, the support and the expertise.   

University of Wollongong Research Online and Purdue University’s ePubs are two great examples of the Digital Commons publishing platform in action.

SelectedWorks is the author equivalent to Digital Commons. It allows faculty to organize and share the breadth of their research, works, and achievements. A SelectedWorks site looks like this. These faculty profiles can be integrated with Digital Commons to allow authors to have their works and profile discoverable alongside the institution's broader scholarly corpus. Integration with the Digital Commons Network takes that connection one step further by tying together the works of authors with the works of others in their discipline.

As for cost, Digital Commons and SelectedWorks together cost a fraction of an FTE; far less than the cost of having local staff to support users and open source publishing systems in a comparable manner.

Inspiring to see

RP: You say that Digital Commons costs a fraction of an FTE. I wonder if you could expand on that. Since it is a hosted service I assume Digital Commons is billed by means of an annual subscription, rather than a one-off payment. Could you give me some specific examples of what different sorts of libraries could expect to pay per annum to use the service?

J-G B: You are correct. Both Digital Commons and SelectedWorks are sold as annual subscriptions.  Size, as determined by FTE, plays an important role in pricing. As you might guess, larger institutions pay more than smaller ones. The average annual subscription fee for Digital Commons and SelectedWorks in 2013 for an academic library was around $20,000 and $8,000 a year respectively. 

RP: As I understand it, Digital Commons is designed to allow universities to publish their own journals on top of their institutional repositories. Presumably it does not matter which repository software they use? Can you say how many universities are currently publishing journals in this way, and what has motivated them to do so? Are they all US-based institutions?

J-G B: Digital Commons is an integrated platform to support the full spectrum of scholarly publishing and institutional repository needs. Imagine ContentDM, DSpace, Dryad, Figshare, ePrints, Open Journal Systems, Open Monograph Press, and Open Conference Systems all rolled into one solution, and you have the Digital Commons platform.

For some institutions, Digital Commons on its own meets all of their publishing and repository needs. Others use Digital Commons in combination with a myriad of local digital preservation and digital asset management solutions. There is, nonetheless, a clear trend toward gradual consolidation as institutions migrate collections stored on DSpace, ContentDM, Omeka, and OJS onto Digital Commons.

There are more than 320 academic institutions that currently use Digital Commons for the full spectrum of scholarly publishing. Three hundred of those are U.S. based institutions and the remaining 20 institutions are in Canada, Australia, Asia, and Europe. These publishing programs are being led by campus libraries that see the opportunity to meet the publishing needs of their faculty and students in a scalable way. It’s been inspiring to see how much faculty and student editors appreciate the library in this new role, and how it has improved their perception of the library’s role on campus.

RP: How does Digital Commons differ from OJS? For instance, Digital Commons is not open source I think. If not, presumably Digital Commons is a more expensive option? And if so, what extra value does bepress offer?

J-G B: Digital Commons is not open source like OJS. I’m a fan of open source software, but I’m an even bigger fan of software in the cloud. Cloud-based ejournal publishing software is the future; it is easier to use, scale, build upon, and support. Digital Commons is delivered via the cloud, while OJS is mostly locally installed.  

As I mentioned earlier, Digital Commons costs a fraction of an FTE, which makes it cheaper than installing and maintaining local software whether open source or otherwise. That is all the more true when one considers the many systems people wish to integrate with that continuously change and therefore require code changes. On top of that, as soon as you take advantage of the open source software by modifying the code, you will typically find yourself stranded in a given version or investing serious money to port your modifications to a newer version.

Let me give you a concrete example: I’m sure you won’t be surprised to hear that readership on tablets and mobile devices is booming. Faculty want the fruits of their labor — their journal — to look great on their iPads and iPhones. In response to this growing demand, last year we rolled out responsive designs for all of the journals, repositories, and other collections on Digital Commons and SelectedWorks.

This means that the web pages respond to the visitor's screen size and change the layout accordingly, so it means the journals look great on an iPhone. Check out this Digital Commons journal example on your mobile phone and then go look at your favourite OJS journal. It was a significant amount of work for us to produce a beautiful mobile interface for journals, but the beauty of a service in the cloud is that once we got it right it was a snap to roll it out for everyone in the community.

But, while we are low cost, we compete mainly by being high value. Moving beyond where the code is stored, Digital Commons also differs from OJS because Digital Commons is a service, not just software. By that I mean it includes people with publishing and technology knowledge who support editors with site design, workflow design, support, training, and publishing expertise. And, for those journals ready for it, Digital Commons includes additional professional publishing services such as abstracting and indexing, DOIs, and preservation in CLOCKSS and Portico.

Finally, we provide service without limits: unlimited storage, unlimited users, unlimited training, unlimited support, unlimited journals, unlimited bandwidth, unlimited get the point.

Our approach frees libraries and journal editors from the technical hassles of running software so they can focus on running a journal. We all know it is a lot of hard work to launch a new journal—e.g., recruiting an editorial board, recruiting authors, and preparing a first issue.

We see our role as helping with everything else. We’ve seen that when given such support that library-publishing programs flourish. Brigham Young University and Kansas State University’s New Prairie Press are both recent examples of converts from OJS to Digital Commons.

Philosophical and practical questions

RP: You said that institutions are migrating to Digital Commons from platforms like DSpace, Omeka and OJS. The latter are open source platforms, whereas Digital Commons is a proprietary platform. I suspect that some OA advocates might argue a) that open source solutions are more appropriate for library-led publishing (since, for instance, open source could be said to better fit with a university’s mission) and, b) proprietary platforms introduce the danger of vendor lock in. I wonder if you could respond to these two points.

J-G B: I know your question reflects the viewpoint of some in the community. I believe it’s something felt stronger on your side of the Atlantic than on my side these days, but I also know that the libraries that use our services sometimes face these types of questions as well.

I think the question of “are you using the ‘right’ kind of software for your open access initiative?” has the unfortunate effect of dividing the community. The bepress community of scholarly communications libraries are no less committed to open access than libraries using open source software, but taken to the extreme, this question devalues the accomplishments of the librarians who build successful OA publishing programs on their campuses using options like Digital Commons.

You were spot on in your Q & A about the state of open access when you identified the lack of cohesion in the OA movement as its biggest failure to date. For those who missed the piece, you wrote: “I wonder if the movement has not been its own worst enemy. After all, it is far from unified, and it has spent a great deal of valuable time arguing with itself rather than winning hearts and minds or taking the practical action needed to achieve its objectives.” I would argue that Digital Commons is one such practical option for libraries to pursue open access library publishing.  

Now I’ll hop down from my soapbox and respond to the two parts of your question. The first part, (a), is philosophical and principled while the second part, (b), is practical. I’m going to tackle the practical argument about vendor lock-in first. I would argue that established open standards are pretty good protections against vendor lock-in. It is really not complicated to migrate from one platform to another platform because all platforms in the space, including Digital Commons, support an open standard, called Open Archives Initiative (OAI). We’ve supported the move of more than 50 collections from different locally-hosted, open source platforms to Digital Commons (and a handful the other way as well).  At the end of the day, the content and metadata are easily portable.

To your first point (a), the connection between open source and open access is strong in the library community, no doubt about it. But here’s the rub: library-led publishing is not a service intended for librarians. It is a service for faculty and students provided to them by their library.

Faculty and students are a demanding bunch; they are increasingly comfortable evaluating all sorts of digital tools, and they expect the best in the tools they choose for themselves. They have a certain goal — to publish a successful journal, for example — and they will be grateful when the library provides them with an excellent publishing experience. If they are poorly served, they will either abandon their library as a publishing partner, or perhaps worse yet, accept the service and complain about it. To meet its mission, the library should ask, “How can I provide the best publishing services to my campus, given the available resources?”

Librarian Karen G. Schneider summed it up better than I ever could when she wrote: “I hate the idea that for some librarians if a particular software is open source, hands down, it’s the right choice. The right choice is the software that meets the mission. While the principles behind open source are admirable, when an open-source product doesn’t meet your library’s needs, your first obligation is to your users.” Karen G. Schneider on July 26, 2006.

RP: Realistically, how simple is it to start publishing a journal using Digital Commons? Is much training required?

J-G B: From our perspective, it is possible to launch a journal and train the editors in a day. If your question is "how easy is it for editors to learn to use the platform?" then the answer is: very easy.

Editors tell us that ours is the easiest to use of all the manuscript submission systems out there. We do all the work of customizing workflows for them. They learn the system in an hour (or an hour and a half if there are lots of editors learning simultaneously), but we train and re-train them as often as necessary.

This model of unlimited support is especially valued by libraries that would like to support student publishing efforts. Students graduate, and the editorship changes hands; it’s a big relief to the library to always have bepress available to train students and answer all of their questions, and to serve as institutional memory when the editorship changes hands.

RP: Are universities (presumably the librarians) able to do all the work themselves, or do they need to buy support services? If so, what kind of support services, and how much are these likely to cost?

J-G B: Librarians certainly could do all of this work themselves. They have to do all the work with OJS already, after all. With Digital Commons, however, they get teams of publishing experts and platform experts at the ready to help them support their faculty and students. There are no extra costs, so they don’t have to second-guess themselves if they ever want more help or services. They can say “yes” to all the requests and questions from their editors and authors. Unlimited support and service is incredibly important to us for this very reason.

Trend is towards openness

RP: I am told that most of the journals using Digital Commons are OA journals. Are you able to give me any stats on this, and current trends?

J-G B: 94% of the journals using Digital Commons are open access, and of the 6% that are subscription-controlled, the majority make their back issues openly available.

The current trend is towards openness as more journal editors realize that the more open they make their journals, the greater their success. It’s been several years since we’ve seen editors of an open access journal using Digital Commons jump ship to sign on with a publisher and move their content behind a pay wall.

In fact, we’ve seen just the opposite — editors of subscription journals making their back issues open access, in turn increasing readership and subscriptions revenues.

We’ve also seen editors of subscription journals with open back archives on Digital Commons take the next step and abandon the subscription model altogether for open access.

We’ve even had a few journals leave publishers for the library in order to make their journals open access. If you look at the Law Review Commons alone, you will find over 170 journals that are open access, almost all of which used to be exclusively subscription access.

RP:  How do universities fund the OA journals they publish using Digital Commons, and can you say something about the average total annual cost of running an OA journal on Digital Commons? Do many of the journals operate article-processing charges?

J-G B: With one exception, none of the journals operate an article-processing charge. Authors who publish in these journals can be sure that their scholarship has been accepted by the journal on its merits alone. The one exception is the “Open Journal of Occupational Therapy” at Western Michigan University. It charges an article processing fee of $200, and $100 for student authors.

Because the Digital Commons publishing model is unlimited, once the university makes the initial investment, they can support as many journals and other publications as they would like.

Most libraries consider supporting a publishing program on campus to fall into their core service offerings, and the fact that publishing capabilities are simply part of our integrated platform makes it easy for them to support the full spectrum of needs, such as conferences, data sets, multimedia publications, textbooks, and so on. There are so many publishing needs on campus when you understand it as a full spectrum, and librarians are eager to meet those needs.

RP: You say that the publishing program is usually treated as part of a library’s core service offering. Do I understand from this that we cannot know what it costs to run a journal on Digital Commons as libraries don’t break out the costs?

J-G B: I see that I didn’t answer your last question very well. Thanks for giving me a second shot. This is actually a really tricky question because the incremental cost of running a journal on Digital Commons is zero. There is no charge per journal. Digital Commons is kind of like those passes you used to get to travel all around Europe by train. I think they were called Eurail in the US and InterRail in Europe. Your question is like asking me “how much did the trip from London to Paris cost?” Since I paid one flat fee for the pass, dividing it up to figure out exactly how much one specific trip cost would depend on how many other trips I took.  

One thing is for sure; when I travelled by train in my youth I was certainly incentivized to take full advantage of my pass. It allowed me the opportunity to explore and experiment. We find that libraries that use Digital Commons are equally incentivized to get the maximum out of their investment. We think it is clearly a model that works.  

However, with Digital Commons the benefits of the unlimited service are actually much better than an unlimited train pass because the benefits are cumulative — so journals launched in any given year are always covered in all future years, with no related change in cost.

RP: Can you share with me any other stats that would give readers a sense of what is happening in the scholarly communication space currently, especially with regard to OA?

J-G B: We recently presented a poster at the Library Publishing Forum in Kansas City that offered a number of interesting statistics. I think one of the most interesting ones is what we’re seeing in terms of the growth and success of library-led, open access publishing.

We’re seeing an increasing number of institutions supporting not just two or three journals, but ten, fifteen, or even twenty journals.  These journals are often serving the needs of underserved disciplines, such as arts and humanities and education. 

RP: There is a lot of talk today about the need for the research community to “take back ownership” of scholarly communication. What are you views on this, and where do you see Digital Commons fitting into the debate? Are the journals being published using Digital Commons merely a fringe activity in your view, or are they the first buds of a revolution that will see today’s scholarly publishers disintermediated?

J-G B: Library-led publishing is an integral strategy in the university taking back ownership of scholarly communication. For the library to offer publishing as a campus-based service means that faculty and students have access to professional publishing tools from their own university, without needing to deal with commercial platforms or predatory publishers.

I’ve been attending the SPARC conference (the major OA event in the US) since 2008 and this year was the first time I had an overwhelming feeling that there is simply no slowing down the open access movement. It is raining green and gold. The benefits to authors, editors, readers and academic institutions are too great. Our momentum is unstoppable.

And let me add one trend that I think speaks volumes about the fact that this is far beyond the first buds of a revolution. 143 of the journals hosted on Digital Commons are edited not by faculty, but by graduate students and undergraduate students.

In an article about the award-winning Journal of Critical Thought and Praxis, founder Cameron Beatty explained the value of their new journal succinctly when he said “We were doing social justice work and some of the journals we wanted to get published in weren’t necessarily interested in the work that we were doing. So we were like, ‘What does this mean for us as graduate students? We need to get a job; we need to get published. Where is there a space for us that want to do critical social justice work? Why don’t we create our own?’”

Let’s turn now to the undergraduate journals. As a research apprenticeship, students are learning far more than the principles of peer review or how to deal with an editorial board (though that’s all really useful). They’re learning that they can have direct access to scholarly publishing, and that they can take a journal from idea to fully realized publication just by talking to their library. It’s going to be awfully hard to explain to them why they can’t just keep doing that when they become professional academics themselves.

If you are interested to learn why faculty and students are excited about this trend in their own words, I highly recommend this Webinar. I also would suggest following Scholarly Communications Librarian and Associate Professor Stephanie Davis-Kahl, from Illinois Wesleyan University. In March she was named the 2014 ACRL/EBSS Distinguished Librarian.  

Transition from publisher to technology company

RP: There is also a view that scholarly publishers will need to morph from being content providers to technology companies. Would you say that this is the path that bepress has taken? If so, do you expect all or most scholarly publishers to have to take the same path in the coming years?

J-G B: For us, the move from publisher to technology company was organic. We only became a technology company when the tools we created and honed to support our own journals became popular. Editors started asking us if they could use our tools when working with other publishers. We began licensing the software as a service and realized that there was a great need for both the software and the support bepress provided, including technical updates, publishing expertise, optimized web discovery, and customer support. We saw that, in line with our mission, we could serve the unmet needs of scholars by offering this software service more widely.

All in all, our complete transition from publisher to technology company took ten years. We were early, but I do think other scholarly publishers are going to have to follow a similar path from content provider to technology company if they want to continue to play a significant role in supporting the changing needs of the research community.

RP: What other changes do you see for scholarly communication going forward: the phasing out of pre-publication peer review? The death of the traditional journal and paper? Other likely developments?

J-G B: The consideration of new forms of review is extremely valuable, but I do not ultimately see the phasing out of pre-publication peer-review. If anything, I see a phasing in of pre-publication peer review, especially among undergraduates — we’ve seen it become a valuable teaching tool that improves the quality of the student work at many institutions. It is a just a matter of time before peer-review tools also find their way into primary education. My oldest son is nine, by the time he is in high-school, peer-review publishing will be part of his curriculum. There are already a few such journals.

Though print is slowly becoming less prevalent, I do not see the passing of the traditional journal model. These will remain as the cornerstones of scholarly communications for some time. This doesn’t mean that there will be no change — there already has been.

The biggest that we’ve seen is an increasing acceptance that there is a continuum of scholarship so that other works, in addition to the peer-reviewed articles, will count toward tenure and promotion. So in addition to providing tools to publish traditional journals, we are constantly seeking ways to support the publication of these other types of works.

Ultimately, the biggest change I see happening is increased awareness on the part of academic libraries that library-led publishing should be a core service that they offer to their faculty and to their students, and that this is a natural role for them that fits into their existing skills, their mission, and the future of their profession.

RP: Thank you for taking the time to answer my questions.


Mike Taylor said...

Very interesting, and encouraging!

"bepress is not the only game in town for libraries looking for a publishing platform. In 2001 the Public Knowledge Project released the first version of the open-source publishing software Open Journals Systems (OJS), and today OJS estimates that over 6,000 journals are being published using its software."

Just to mention a few other alternatives.

Annotum is an open-source publishing platform based on WordPress, which is used by PLOS Currents.

The widely-admired PeerJ system is freely available on GitHub (though there's no single installable package to my knowledge).

I know I've read about others, but can't call them to mind right now: anyone?

John Baez said...

Sounds like great news... but the fact that bepress sold all its journals to De Gruyter makes it clear that like any commercial entity, we have to be careful to remember they're in this for the money. This is okay as long as we don't forget and make the mistake of assuming they're "benevolent".