Third Time is a Charm?

Third Time is a Charm?

1

S2-Whoa

2

Since its inception, the Subscribe-to-Open (S2O) business model for journals carried some clear risks, particularly the free rider problem. It’s long been clear that models where financial support for journals was voluntary and the library saw no loss in access if it stopped paying were going to be a difficult sale into some markets. As Gwen Evans (then Executive Director of OhioLINK) put it, “…under the pressure of decreasing budgets in higher education generally, but especially in public institutions, administrations will not continue to pay for something that they will receive for free.”
 
Relying on the goodwill of librarians who want to support new, more equitable models for open access (OA) will only get one so far. At least, that seems to be the lesson learned by the International Water Association (IWA), which in recent weeks has had to “adapt” its S2O model. When IWA’s move to S2O was first proposed, the challenges were very clear to the organization’s leadership. Not only would current subscription levels need to be maintained, but to succeed IWA needed to significantly increase subscription revenues to make up for the gap in funds lost from licensing, rights, APCs, and other revenue streams that would disappear with the shift to OA. After gaining enough library investment to flip the IWA portfolio of ten hybrid journals to OA in 2021, support has waned to the point where S2O no longer appears sustainable. “Sadly, while many previous subscribers have maintained their subscriptions under IWA Publishing’s version of S2O, a significant number of institutions across the world have not. Since the launch by IWA Publishing of S2O in 2021 our subscription revenue has declined considerably.”
 
Although IWA continues to refer to their program as being under S2O, they have essentially moved it to a “pure publish” model, where the journals will remain free to readers worldwide, and authors at subscribing institutions won’t have to pay to publish. Those not at subscribing institutions will pay APCs on their papers. This is a difficult position for a portfolio that published around 2,000 articles in 2023. Institutions that don’t publish significant amounts of articles each year with IWA may be better off paying the occasional APC than continuing to subscribe at what may be a higher cost. Failing to flip back to subscription-access makes the IWA’s S2O model entirely toothless, and at this point it is unclear how they expect to claw back all that lost revenue. If you’re going to threaten to flip the journals back away from OA, it’s important to follow through on that threat.
 
In contrast, EDP Sciences states in their 2024 Open Access and Transparency Report that subscriptions for their S2O journals are on the rise. Even so, that’s not enough to cover the costs of operating the journals, with subscriptions still only totaling 51% of publication costs in 2023. Additional grant funding and other sources brought the EDP S2O portfolio up to covering 72% of its costs. Four years into their S2O experiment, EDP’s journals still remain far from independently sustainable.
 
These sustainability shortfalls should raise red flags for publishers considering the S2O model. S2O was always, in our opinion, somewhat geography-, participation level-, and time-limited. For journals with a global reach that rely on readership outside of wealthier nations with relatively robust library budgets, S2O sustainability appears increasingly unlikely. Further, as the model has picked up steam and once enough publishers moved to it, S2O would appear to no longer be able to easily “fly under the radar” of administrative cost-cutters. And clearly, as libraries’ budgets become more and more constrained over time, their willingness to pay for something they receive for free remains limited. Perhaps the only thing surprising here is how short the timeline for S2O to hit this point turned out to be.

The Great CDP Race

3

Atypon announced the upcoming launch of their CONNECT Customer Data Platform at their recent Community Day. CDPs (customer data platforms) are in a rapidly growing marketing technology category, with over 180 vendors and growing. According to the CDP Institute (which, we realize, is not disinterested), the global CDP market is projected to grow from $5.1 billion in 2023 and $7.4 billion in 2024 to $28.2 billion by 2028. Even discounting those numbers considerably, the revenue from CDP systems (in aggregate) is poised to eclipse the entire professional and scholarly journal market and likely by a wide margin. 
 
CDPs gained traction in consumer publishing and B2B (business-to-business) media about six or seven years ago, though they’ve existed for over a decade. They’ve just started to be adopted in our industry over the last two years. (Note: For more on this trend and its implications, along with use cases and other insights, listen to C&E’s Colleen Scollans on the DelCor podcast Reboot IT, episode “Customer Data Platforms: What? Why? and When?“)
 
Atypon CONNECT, like Hum (another CDP platform provider), is likely banking on the value of a CDP that understands scholarly communication and association use cases. Hum’s AI-first and content-centric orientation has put a fresh spin on the CDP category, and they have seen success as a result. Atypon, with its connections across Wiley Journals and Wiley Partner Solutions, hints interesting possibilities around audience identification and customer data enrichment. We are eager to learn more. 
 
But don’t count out the vertically broader CDPs. CDPs with track records in consumer publishing, B2B media, and retail are also making inroads into our sector (e.g., platforms such as BlueConic, Lytics, Tealium, Bloomreach, and Treasure Data). These vendors really understand modern marketing use cases – and have proven they can successfully meet the needs of publishers and associations.
 
Regardless of the increasing number of CDP vendors, the core principles remain constant: strategy leads, clear business cases drive decisions, and organization preparedness (especially the marketing team) is essential. Ultimately, CDPs and other marketing technologies should align with and support strategic objectives; if you are unsure where to start, start there. 
 
We welcome more CDP choices, as selection in MarTech (marketing technology) is beneficial. Different organizations have different use cases and business case drivers. Some organizations’ journals are self-published, while others partner with commercial publishers. Some organizations integrate audience and customer data across the association, while others keep their publishing unit and data separate. Marketing teams come in different sizes and are in different states of maturity. Organizations also have vastly different marketing (and data) technology stacks (and roadmaps), and as a result they rely on their CDP for different capabilities. Moreover, CDPs offer varying capability sets, service models, user interfaces, native integrations, and pricing structures. This is why technology evaluations are very common when selecting Marketing Technology tools, including CDPs. 
 
This is definitely a space to watch. Anyone have popcorn?

Collusion

4

Much of this month’s publishing community chatter revolves around the announcement of an antitrust lawsuit against six publishers (Elsevier, Springer Nature, Taylor and Francis, Sage, Wiley, and Wolters Kluwer) along with 50 unnamed “John Does” to be identified during discovery. 
 
The suit seems flimsy to us at The Brief but we emphasize that we are not lawyers and cannot offer any cogent legal analysis on the claims. Regardless of whether the suit is successful or not, this is another example of publishers being cast in the role of enemies of research progress. 
 
For the last two decades, publishers have been called out for their (one time) opposition to open access, their high profit margins, their consolidation of research output, and their general commercial nature (and that is on a good day!).
 
In this lawsuit, publishers are simultaneously pilloried for both their high prices and for not adopting practices that would vastly increase prices (contradictions noted by C&E’s David Crotty in The Scholarly Kitchen). This no-win situation is typical for the industry, as it lurches from one public relations disaster to the next. This particular black eye is perhaps unique in that there are no ”good guys” or “bad guys” to be delineated from the claims. Nearly all publishers, whether large, small, commercial, or non-profit, practice the activities described in this complaint and could end up in the court’s crosshairs. 
 
Two factors we couldn’t help notice in the lawsuit that are also worth mentioning: First, the law firm involved states that they had been “investigating the case” before coming upon UCLA professor Lucina Uddin as their named plaintiff. It strikes us as somewhat odd that the plaintiff is someone who has benefited so much from publishing in the journals from the publishers named in this suit (isn’t being hired as a professor at one of the most prestigious universities is the world the entire point of publishing in these journals?). We would love to know more about the origins of the lawsuit, and who is financially backing it. Second, as we wrote about in July, there remains lingering misunderstanding about which publishers actually dominate the market, and we can’t help but feel bad for Sage being caught up here while larger organizations (as measured by article market share) like MDPI, Frontiers, IEEE, and ACS were all deemed unworthy of similar attention.

Briefly Noted

5

Has AI reached a point where it can replace scientists? Investors in Sakanai AI seem to think so, with Nvidia among others providing $100 million-plus in funding. While the “AI Scientist” is claimed to “automate scientific research and discovery,” those claims come with major caveats and limitations. Jimmy Koppel offers a really helpful overview on just what it is that Sakanai AI is doing and whether it’s of value. He describes it as a “masterwork of PR” and doesn’t “think there’s much there at all.” We also credit this series of tweets with one of our new favorite phrases, “Artificial intelligence is really natural stupidity”, citing Sakanai AI’s ability to reproduce the worst behaviors of the human reviewers it was trained on.
 
The Internet Archive (IA) has lost its appeal in the case over Controlled Digital Lending (CDL) and its pandemic-era National Emergency Library (NEL). The appeals court rejected the IA’s argument that its activities were fair use, calling it “unpersuasive”. We at The Brief still remain a bit baffled at the IA’s overreach here, and the staking out of an extreme position with the NEL that went far beyond the typical standards of CDL programs.
 
Apple is making significant changes to how their Mail program displays messages. Rather than showing the first few lines of an email, users will now see an AI-generated summary of the message. Marketers should take heed and consider how this will impact how they craft email messages.
 
No one has learned the painful lesson value of performing extensive due diligence before purchasing a portfolio of journals more than Wiley, but Sage has likely joined that club as well, as it has had to retract more than 450 papers from a journal that came with the purchase of IOS Press last year.
 
Researcher identity is increasingly seen as the next important integrity evolution for scholarly publishing. Just as banks make sure to verify an applicant’s financial history before offering a loan, so too might journals hope to better understand the research history of the authors on a paper before granting acceptance. A preprint released earlier in 2024 offers one route, identifying characteristics and connections of authors within authorship-for-sale networks that can be screened for in submissions. Hylke Koers (CIO for STM Solutions) suggested in a recent interview that identity verification may go even further, and that, “…we will see an increased need for publishers to better understand, and validate, the identities of individuals that they interact with throughout the editorial and peer-review process – not just authors, but also reviewers and (guest) editors.”
 
Is anyone still using Twitter? If so, then a recent study may be of interest. Posts about scientific publications got more engagement when they included personal commenting (rather than just sharing the title of the article) and images.
 
A study testing the prescribed vocabulary used in eLife’s research assessment statements found it to be ambiguous, with readers’ interpretations of the implied ranking not matching the intended ranking of the report.
 
The Center for Open Science’s TOP (Transparency and Openness Promotion) Guidelines have been updated. The TOP Guidelines have been invaluable in C&E’s efforts to help publishers put together open science policies and we look forward to digging into the new frameworks offered by the update.
 
A French court has ruled against the University of Nantes’ mandate that required its researchers to publish open access. The French publishers’ association, Syndicat national de l’edition, states that the university’s policy “infringed on the copyright and academic freedom of teacher-researchers.”
 
Sage is the latest publisher to announce efforts to license content for AI training, and like others involved, makes clear it will pay royalties on any licensing income in accord with its contracts with authors, editors, and societies. Meanwhile it appears that AI is wreaking havoc in polluting the use of language online. Wordfreq, an open source program that tracks the changing popularity of different words, has been shuttered because AI spam has overwhelmed human-generated language online, making the data no longer useful.
 
EBSCO has released its 2025 Serials Price Projection Report, which suggests publisher prices for eJournal packages will rise 3.5% – 4.5% in the coming year. Reasons given for the increases include inflation, growth in the literature, and investment in AI and other technologies.
 
Given the increasingly well-known concerns about research integrity in guest-edited special issues, could publishers be retroactively removing special issue labeling from articles to avoid scrutiny? In a concerning blog post, Dorothy Bishop suggests that MDPI is indeed doing that. 
 
Science covers a study that found 263 suspicious reviews prepared for 37 journals that appear to be AI-generated. The motivation for faking a peer review for someone else’s paper remains murky, given how little credit most academics receive for doing this voluntary work. If you’re not going to actually review the paper, then why not just decline the invitation? The article suggests that some (but not all) of the reviews ask authors to add citations of the reviewer’s papers to the articles. The publisher where nearly all of these suspicious reviews were found? MDPI.
 
Akhilesh Ayer has been named the new CEO at Cactus Communications.
 
One of the more confounding problems that continues to plague open access and open science progress is the lack of clear definitions of the terminology used. Rick Anderson wrote about this back in 2017. Things don’t seem to have changed all that much in the last seven years, with the possible exception of that terminology confusion expanding further. In a recent thought piece, Danny Kingsley observes that “OPEN research infrastructure” (infrastructure for research that is non-commercial, community-owned, and open source) is a different thing from “OPEN RESEARCH infrastructure” (which encompasses all infrastructure that can be used for open research, regardless of its ownership/openness). As Kingsley notes, “Words matter. It is very hard to have meaningful conversations and make informed decisions when we cannot agree what we are talking about.”
 
Quis custodiet ipsos custodes? It turns out that if you’re going to publish a study about science reform and offering claims about the importance of preregistration of research plans, it helps to actually eat your own dogfood and preregister your experiments. A Nature Human Behavior study is being retracted for just that, claiming to have preregistered experimental plans without actually doing so and then HARKing (Hypothesizing After the Research is Known). “It seems like a lot of the aspects of this paper and its process sort of mirror the original concerns that motivated reforms in the first place.”
 
Are one in seven scientific papers really “fake”? That’s the claim made in a new meta-analysis, which critics suggest is either “slightly misleading” or “not a rigorous way to get an estimate of anything.”
 
An interesting new preprint suggests that science journalists have recognized the concerns in reporting unvetted research to the public and have curtailed their use of preprints.
 
In late breaking news as The Brief went to press this month, Clarivate’s Web of Science has paused indexing of the megajournals Heliyon and Cureus. Readers might recall from last month’s Brief that these were two of the journals experiencing rapid growth as authors shifted papers away from MDPI and Frontiers titles. Whether Clarivate’s concerns here amount to anything is yet to be determined.
 
Say what you will about the utility of “glam” journals like Science and Nature, one thing they are both apparently very good at is causing paper cuts. A study by Sif Fink Arnbjerg-Nielsen and colleagues found that the paper used for Science and Nature is among those most likely to inflict a cut. 
 

***
“Metascience research is sometimes not metascientific.” – Daniele Fanelli, Heriot-Watt University