It’s easy to overlook the fact that print and digital are not simply formats but entire ecosystems as well. For those of us who labor in the mines of STM journals, this oversight is understandable, as we can be lulled (asphyxiated?) into believing that all the world is digital, artifacts of “print consciousness” notwithstanding (e.g., the PDF, the version of record). (Yes, print consciousness may be the topic of a forthcoming post.) If print and digital did not in fact represent entire ecosystems, when digital products came along they would have just been slotted into the place that print once occupied. But we know that is not what happened. Digital media, for example, gave rise to comprehensive full-text search, which helped to restructure the environment. Searching on a library’s catalogue, once a very powerful tool, now has yielded pride of place to various discovery services (e.g., Summon) and, of course, Google Scholar. Among other things digital media facilitates the development of Big Deal marketing of journals, demand-driven acquisitions for books, and the emerging world of data analytics. This is not to say that the format issue is not important in itself. It is: the switch to digital formats reduces inventory costs, for example, and increases speed to market. The cruel irony of innovation, however, is that none of the many good reasons to switch to digital formats anticipated the broader structural changes in the ecosystem that we know of as scholarly communications. It is ever thus: we innovate tactically, but wake up to strategic transformations.
Among the structural changes–that is, changes in the ecosystem–that the introduction of digital media has brought to scholarly communications is the reduced emphasis on the individual product and the new and expanding emphasis on the family of products. In print a journal was a stand-alone item; authors submitted papers to it, libraries and individuals purchased it. But now a journal is joined by many siblings, in a Big Deal literally of a thousand or more sibling, and cousins, nephews, and everybody’s favorite aunt. The portfolio of even a small publisher may include journals, books, databases, conferences, newsletters, Webinars, and much more, all of which increasingly are asked to work together. The various properties share many things — administrative overhead, for example, or a common customer database — and they also reinforce one another. Think of each property as a node on the network, gaining strength through the linkage to the other nodes. This has important implications, not least of which is Metcalfe’s Law: the value of a network is equal to the square of the number of nodes attached to it. Every successful publisher is going to become the manager of a network publishing system. Success hinges on who gets there first.
The network publishing model is sometimes, perhaps usually, confused with consolidation. Publisher A has 500 journals and sees that Publishers B, C, D, and E each have close to 2,000. Publisher A embarks on a strategy to get bigger and become more competitive. Publisher A begins to acquire more journals, often risking overpayment in pursuit of size. Meanwhile, B, C, D, and E are also acquiring other journals like crazy. One day everybody wakes up to the fact that about half of all journal revenue has been hoovered up by six publishers. Speculation begins about how six can become five, how five can become four. This is the world we are living in right now, and it still has several years to run. But this strategy may be pursued simply for the sake of market clout, as librarians like nothing so much as acquiring all their content from a very small number of sources. Finding deeper and often subtle connections between properties may not be part of this. Cost reductions, of course, are very much a part of the consolidation game: when Publisher B eventually acquires Publisher A, look for an RIF — Reduction in Force — around the world.
A better network model, while continuing to pursue consolidation where market forces invite it, looks to assemble a collection of properties — all different nodes on the network — that engage the various members of the community the publisher serves in all aspects of the members’ professional activity. Yes, we have a traditional journal or likely a collection of journals in (say) the veterinary sciences, but we also have a book program, a Gold OA journal, probably a cascading journal program, a series of conferences, Webinars, and an alerting service for new developments in the field. We may also have a hosting service for preprints and another (most likely a front-end to AWS) for hosting and managing researchers’ experimental data. There is no end to the number of nodes that can be connected to the network. The business questions for attaching a new node are:
- Is the new node likely to generate cash itself?
- Is the whole greater than the sum of the parts? That is, does the new node add value in some way to other nodes on the network, for instance, by providing greater granularity in end-user data?
- If the new node cannot be a positive cash generator in itself, can we quantify the cash it generates for the other nodes it supports, making a cash subsidy a prudent investment?
That last bullet point is an important one. It cannot be said loudly or often enough that not everything in a business needs to make money. I observed this recently with a client that had a costly and cumbersome arrangement for participation in certain policy boards. Why bother? is the first question to ask. But then it became apparent that that participation allowed the organization an inside track to acquire the rights to important publishing assets. One node — policy participation — was “subsidized,” but it was in fact a critical piece of the financial performance as a whole. The “everything must carry its own weight” model of finance can be too narrowly construed. CFOs often need English majors to get at the true economic value of a complex enterprise.
We should keep in mind an obvious point: not all nodes in a network are created equal. It is almost certain that successful network publishers will have a flagship journal or reference book around which the rest of the network will organize itself. The important management challenge — with an eye on the shareholders of ten years from now — is not to allow the network to serve the flagship alone but to position the flagship to serve the network. This is because the primacy of any one property in a networked age must compete for attention with all the rising stars among the flagship’s rivals. It is not possible to acquire all the rivals, but it may be possible to coopt or preempt them.
And some nodes are simply harder to create. It is not hard to create a journal (though it can be hard to market it successfully), but it is terribly hard to create a social media venue. This is because social media services tend to be of the winner-takes-all variety, and the winners typically are anchored in the huge numbers of the consumer market. We have yet to see the blockbuster play in social media for professional information, though Mendeley (now a node on the Elsevier network) is a candidate. Publishers may find that, at least in their subject areas, a social media node will elude them or be under the control of a clever rival or a Web-scale consumer service. That would mean that all of the information about the end-use of the sidelined publisher will be gathered by entities that don’t have that publisher’s interests in mind. The case for being bold and early is clear.
A network publishing model has greater nuance than a simple brand extension model. In brand extension the flagship brand is used to enter the marketplace with a family of products. In a network model the brand may not be used for certain nodes, as it could discourage some elements of the broader ecosystem from participating. How enthusiastic would the American Chemical Society be about referring its members to a service branded by the Royal Society of Chemistry? Suppressing the brand in some instances may be shrewd, as is taking a minority interest in certain properties. Not all nodes of a network must be wholly owned by the network publisher provided that each node is playing its role for the benefit of the overall network.
If a publishing network is not owned outright by a single publisher, if it occasionally declines to use its primary brand, if it has unprofitable as well as profitable nodes — at what point do we ask, What the heck is holding this thing together? Aside from the obvious synergies in cost containment (e.g., more clout with vendors) and marketing (cross-referrals), the unifying element is the data that can be derived from all the usage across the network. We know now that Professor Jones wrote this article, contributed a chapter in that book, attended this Webinar, delivered a paper at that conference; we know how her experimental data is being used and by whom, what materials were assigned in her undergraduate class and what her graduate students are working on — we know, in other words, all the things that make Edward Snowden, Facebook, and Google so creepy. But we also know how the research community as a whole is moving across the network, what topics are of interest to them, and what is declining in importance. This user data, whether personal, anonymized, or organized in clusters, provides the platform for new economic activity. Inasmuch as the rule for data is that more is better, the bigger the network, the more powerful the position of the organizing publisher. Perhaps we are very, very far from seeing the end of the age of consolidation.