In 2021, I was asked by a nonprofit scholarly publisher of scientific journals to respond to a prompt inspired by an Inside Higher Ed article (Tepper, 2021). The prompt asked for a comparison of the disruptions to scholarly communication from the Sci-Hub (n.d.) and Open Access (OA) initiatives (SPARC, 2022) to the disruptions of Napster to the music industry at the turn of the century. To refresh our collective memories, Napster was a free, peer-to-peer music file sharing service enabled by the proliferation of high-speed (for the time) Internet service. Internet-savvy music fans, many of whom felt financially exploited by record companies, concert ticketing agencies, and music merchandisers, felt little compunction about downloading music for free. Napster’s rise, fall, and aftermath is a story unto itself (Winter, 2013). For the purposes of this article, I will highlight two key outcomes of this disruptive technology.
1) À la carte
Since the mid-1960s, record companies utilized a business model based on album sales. Single songs — singles — were sold, but at a higher per-song price and with the understanding that, if someone later bought the album on the strength of a single, they would effectively be paying again for a song they already owned. This model enabled record companies to promote one or two singles to radio stations and record stores in order to sell consumers entire albums. Occasionally, albums were full of hits. Sometimes artists capitalized on the album medium by writing and arranging songs that worked together to form a whole greater than the sum of its parts (e.g. Pink Floyd’s The Wall). Most often, an album consisted of hits and misses, songs good, bad, and mediocre, but no matter what, consumers were forced to buy them all. Napster divided albums into their constituent parts — songs — rendering each song into a unique digital MP3 file. This enabled Napster users to download only the songs they wanted — the hits — and arrange them in whatever order, or playlist, they preferred.
Due to a new technology, the record companies’ power had been suddenly and greatly diminished. Music fans could consume only the songs they wanted without buying the album. And the devastating impact to record companies extended beyond simple piracy; even after Napster was shut down in 2002, the launch of iTunes in 2003 cemented the individual song as the new standard commercial unit of music. The impact of Napster on the music industry cannot be overstated. In the years following Napster, total U.S. music sales declined from $21.5 billion to only $6.9 billion, adjusted for inflation (Tepper, 2021).
Historically, music companies wielded at least as much power over artists as they did over consumers. Without a record deal, artists had limited to no access to radio airplay, concert ticketing agencies, venue promoters, and the music press. Record companies’ artists and repertoire (A&R) representatives played kingmakers by determining which bands to sign, despite their opinions being proven wrong as often as they were right (Barker, 2015). Napster disrupted this oligopoly by providing a platform where music fans could listen to artists outside the mainstream without any financial barrier, eroding the music companies’ tastemaker role and enabling music fans to decide for themselves what they liked. Napster was succeeded by platforms like MySpace, Bandcamp and Soundcloud, which allowed unsigned, independent artists to share their music directly with the public, bypassing record label gatekeepers. If an artist could cultivate a strong following, they could parlay that support into a record deal from a position of power. Launching a music career via social media platforms like YouTube is now commonplace. Both the emergence of the à la carte and the direct-to-consumer models in music closely mirrors the emergence of these models in scholarly communication.
Sci-Hub is the scholarly equivalent to Napster, a platform where content can be downloaded in its constituent parts (articles), irrespective of copyright protections. Unlike Napster, which was shuttered after just three years, Sci-Hub has been available for 11 years and currently contains over 87 million articles (Sci-Hub, n.d.). Like the music industry, scholarly publishers have discovered that many consumers have no qualms about downloading copyright-protected content for free, and researchers may have a stronger justification. They are not merely seeking a new jam; theirs is the pursuit of knowledge. This easy access to content has a real impact on research. A 2021 study found the “articles downloaded from Sci-Hub were cited 1.72 times more than papers not downloaded from Sci-Hub and that the number of downloads from Sci-Hub was a robust predictor of future citations” (Correa, 2022). This demonstrates that paywall access barriers matter. Rightly or wrongly, in the pursuit of knowledge, many researchers who encounter significant difficulty accessing a scholarly article will simply move on and look elsewhere for an alternative, more easily accessible article that supports their work.
Open Access initiatives are the scholarly equivalent of direct-to-consumer music and social media platforms (SPARC, 2022). Though still refereed via the peer review process, OA provides unrestricted access to research without any financial barrier to readers. However, just as in the music industry, someone must foot the bill. OA initiatives shift publication costs from reader to author, most often in the form of upfront Article Processing Charges (APCs).
While APCs are often viewed as a proven means of transitioning to OA, it is worth noting several drawbacks. First, APCs are rooted in the old business model. “The APC model represents a lateral move in terms of access, greatly improving access for readers, but shifting the inequity in the system onto authors” (Crotty, 2019). APCs represent a new revenue stream for publishers, which is either good or bad depending on the publisher. Nature, for example, charges more than $11,000 USD for an APC, which is completely beyond the budget of researchers in low income countries (Else, 2020). APCs are unpopular with many researchers because they view their research budget as a zero-sum game. For every article published via an APC, it becomes that much more difficult to fund a graduate student’s project or pay a postdoc’s salary (Clarke & Esposito, 2021). APCs exacerbate inequities. APCs inevitably discriminate against authors who have less institutional support and against fields of scholarship for which such support is scarcer (Michael, 2019). And finally, strictly speaking, APCs are not Plan S compliant unless the author can deposit their article in a recognized repository at time of publication without any embargo (Russell, 2019).
More is less
At the same time, the rate of scholarly output continues to increase. About 8,000 new scientific articles are published every day (Board, 2021). While once scholars waited for good research to be published, now they cannot consume it fast enough. Historically speaking, there are many benefits in moving from times of information scarcity to times of information abundance, but taken to the extreme, a new problem emerges that calls for a novel solution: in an age of information abundance, how can researchers keep up?
Big data, big promise
Artificial Intelligence (AI) promises to extract meaning from large datasets. AI can reveal connections, trends, and insights that would be difficult or impossible for researchers to see on their own (Scholastica, 2020). AI can suggest potential peer reviewers based on similar research output, rather than popularity, accessibility, or geography (UNSILO, 2019). Studies have demonstrated the effectiveness of automatic technical checkers in the editorial process (Upshall, 2021). Large-scale data analysis can identify areas of interconnectedness in seemingly disparate disciplines (Luther, 2021). AI scans of troves of preprints for keywords will alert researchers to emerging areas of research before they are formally published (Fraser, 2021). And, aided by AI, someone in every scholarly field will be able to curate the RapCaviar of new scholarly works (Spotify, n.d.). Technology has created a problem of abundance which only technology can solve through algorithmic curation.
Startups as bellwethers
When an industry is being disrupted, startups will appear, whiffing opportunity. Evaluating the startups can provide hints at where the industry’s future may be heading. UNSILO (2020) is working with the largest for-profit publishers to build tools that utilize AI, such as manuscript technical checks. Companies like scite (2022) are building what amounts to a case law database for scholarly publications. Just as attorneys and judges can easily search case law to find judgments in support of their argument and judgments that raise questions about it, researchers will be able to find articles that support, or raise questions about, their hypotheses. FREYA (n.d.) aims “to extend the infrastructure for persistent identifiers (PIDs) as a core component of open research.” Meta, once a Toronto-based, AI-powered research search engine startup acquired in 2017 by The Chan Zuckerberg Initiative, “helps scientists find the latest papers related to their own projects, while assisting funding organizations to collaborate with researchers and identify high-potential areas for investment or impact” (Constine, 2017). Although not a startup, it’s worth noting that RELX, Elsevier’s parent company, has repositioned themselves in anticipation of the changing landscape of scholarship. They now refer to themselves in shareholder reports as an Analytics company, and increasingly pursue deals not with librarians or academic departments, but directly with Higher Ed administrations (RELX, 2020).
The common denominator across all of these companies is that they promise to derive value from large datasets. Much like the recommender algorithms we see at work in large online content databases like YouTube, TikTok, Netflix, and Spotify, it stands to reason that as more and more scholarly content becomes available via OA, researchers who cannot possibly metabolize all of it will find value in tools that curate, collate, and meta-analyze it.
The conspicuous gap between discovery and publication
The ease and speed of sharing information online has shined a light on the length of the scholarly publication process. Just as Napster enabled the leaking of completed songs before the record company could design and execute a coordinated album marketing and release campaign, researchers can share white papers, datasets, and preprints online in minutes, often 9-18 months before a journal can accept, peer-review, edit, and publish it (Scholastica, 2019) (Wicks, 2020). Viewed through the lens of urgent events like wildfires or a pandemic, the lag time between scientific discovery and scientific publication can be a matter of life and death. The value of the peer review process is unassailable, but the status quo has real-world costs. “The rapid dissemination of knowledge is critical; closed, siloed knowledge slows progress to a degree society cannot afford” (Gentemann et al., 2022). Many journals have adopted issue-in-progress workflow, issue preprints, and utilize automatic technical checkers to expedite the knowledge dissemination process, but a more radical solution — such as the widespread adoption of Open Science — is needed (Gentemann et al., 2022).
Prepping for an uncertain future
Just as the music industry eventually adapted to singles as the new commercial unit of content, the rise of streaming and the decline of physical media, and artists establishing themselves independently on social media platforms, nonprofit scholarly publishers will be forced to adapt to a new paradigm still taking shape.
How, then, should a nonprofit scholarly publisher prepare? Though the future will be messy and unpredictable, there are already many bridges to help transition from the traditional publishing model to OA. In a recent report, The Society Publishers Accelerating Open access and Plan S (SPA-OPS) project identified 27 business models and strategies that can be deployed by publishers to transition to OA (Wise & Estelle, 2019). There are no panaceas here, but some combination of these models could help many publishers transition to OA.
Reading the signs available to us, one can imagine two futures. In the first, dystopian future, independent societies and small publishers, unable to transition to new business models, are shuttered; less well-funded researchers can't afford to publish because all costs have been frontloaded in the form of APCs; an oligopoly of for-profit tech companies sign exclusive contracts with Higher Ed administrations; top research is dependent on proprietary AI and analytics tools; and a commercial aggregator creates a metric — call it a scholar’s “eScore” — that, if high enough, would earn them tenure. In such a scenario, the commercial aggregator could press their thumb on the scale of their own journals and artificially lighten the weight of third-party journals.
More optimistically, one could imagine a second, utopian future, where research is federally funded and/or independent societies and small publishers adopt new OA business models to fund research; tenure is no longer dependent on publication in a narrow set of journals; and powerful open source AI and analytics tools level the playing field for all researchers.
Taking this all into account, I believe that nonprofit scholarly publishers have a bright future. I base this belief on a fundamental of economics: if something has value, people will pay for it. Societies & publishers have always and will always add value to research (Wise & Estelle, 2019). Even though OA is upending the status quo, the work nonprofit scholarly publishers do and the work societies do will always have value.
Just as Apple co-opted the idea of file downloading from Napster in the form of the iTunes store, pairing it with a new technology (the iPod) to dominate the marketplace, nonprofit scholarly publishers will be well-served to join forces with the right technology companies to combine the value each party provides for the benefit of researchers.
What to do now
In assessing recent proprietary survey data from a nonprofit scholarly publisher, it is evident that societies and independent publications have concerns and questions about OA business models. In the short-term, nonprofit scholarly publishers can play a consultative role, educating societies on their 27 options and helping them choose the model, or combination of models, that best suits them. This could be done at scale using tools available from the Society Publishers’ Coalition (2022). My overall advice to nonprofit scholarly publishers, or any organization in uncertain times, is to foster an organizational culture that is nimble; engage with stakeholders as much as possible, even if it seems like you are over-engaging; make multiple small bets because no one is certain which approach will be successful; partner where it makes sense; reward experimentation; be transparent with stakeholders; and leverage your unique value.
Let your mission be your North Star
Mission-driven organizations that exist to serve the needs of their stakeholders should lean on engagement and inclusivity to ensure that their strategy remains aligned with their purpose. In an environment of rapid, messy, uncertain transition, always look to your stakeholders as your North Star. You can never get too far off course if you listen to those you serve.
Board, N. S. (2021). Science and Engineering Indicators 2022. National Science Foundation. Alexandria, VA: National Science Foundation. https://ncses.nsf.gov/pubs/nsb20214
Clarke & Esposito. (2021, June). Citation Advantage? The Brief, 35. https://www.ce-strategy.com/the-brief/citation-advantage/
Constine, J. (2017, January 23). Chan Zuckerberg Initiative acquires and will free up science search engine Meta. Tech Crunch. https://techcrunch.com/2017/01/23/chan-zuckerberg-initiative-meta/?gucc…
Correa, J. L.-R. (2022). The Sci-Hub effect on papers’ citations. Scientometrics, 127, 99-126.
Crotty, D. (2019, October 9). Retrieved from Scholarly Kitchen: https://scholarlykitchen.sspnet.org/2019/10/09/roadblocks-to-better-open-access-models/
Else, H. (2020). Nature journals reveal terms of landmark open-access option. Nature, 588 (7836), 19-20. doi: 10.1038/d41586-020-03324-y
Fraser, N. B. (2021). The evolving role of preprints in the dissemination of COVID-19 research and their impact on the science communication landscape. PLoS biology, 19(4), 1-28.
FREYA. (n.d.). The FREYA Project. https://www.project-freya.eu/en/about/mission
Gentemann, C. C., Erdmann, C., & Kroeger, C. (2022). Opening Up to Open Science. Issues in Science and Technology, 38(3), 57-59.
Luther, J. (2021, May 20). Content at Scale – The Third Wave. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2021/05/20/content-at-scale-the-thi…
Michael, A. (2019, October 24). Ask the chefs: Beyond the APC. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2019/10/24/ask-the-chefs-oa-busines…
Barker, E. (2015, June 17). 14 Record lables who rejected bands who went on to be massive. NME. https://www.nme.com/photos/14-record-labels-who-rejected-bands-who-went-on-to-be-massive-1427448
RELX. (2020). Annual Report and Financial Statements 2020. https://www.relx.com/~/media/Files/R/RELX-Group/documents/reports/annua…
Russell, B. (2019, October 23). Guest Post — The Future of Open Access Business Models: APCs Are Not the Only Way. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2019/10/23/guest-post-the-future-of…
Scholastica. (2019, October 25). How journals are using overlay publishing models to facilitate equitable OA. https://blog.scholasticahq.com/post/journals-using-overlay-publishing-models-equitable-oa/
Scholastica. (2020, January 10). 5 Scholarly Publishing Trends to Watch in 2020. https://blog.scholasticahq.com/post/scholarly-publishing-trends-to-watc…
Sci-Hub. (n.d.). Sci-Hub: …To remove all barriers in the way of science. https://sci-hub.hkvisa.net/
scite. (2022). https://scite.ai/
Society Publishers’ Coalition. (2022). Our mission. https://www.socpc.org/
SPARC. (2022). Open Access. https://sparcopen.org/open-access/
Spotify. (n.d.). Playlist. https://open.spotify.com/playlist/37i9dQZF1DX0XUsuxWHRQd
Tepper, S. (2021, July 13). Article title. Inside Higher Ed. https://www.insidehighered.com/views/2021/07/13/higher-ed-facing-reckoning-much-music-industry-already-has-opinion
UNSILO. (2019, November 1). UNSILO AI in Academic Publishing Survey 2019. https://unsilo.ai/wp-content/uploads/2019/11/UNSILO-Survey-on-AI-in-Aca…
UNSILO. (2020). Rethinking publishing with AI. https://unsilo.ai/
Upshall, M. (2021, April 16). How effective are machine checks for manuscript submissions? UPSILO. https://unsilo.ai/2021/04/16/how-effective-are-machine-checks-for-manus…
Wicks, P. (2020, July 15). How long does it take for a peer-reviewed journal article to be published? 5 tips to speed up the process for digital health companies. Medium https://medium.com/@PaulLikeMe/how-long-does-it-take-for-a-peer-reviewe…
Winter, A. (Director). (2013). Downloaded [Motion Picture]. https://alexwinter.com/projects/downloaded/
Wise, A., & Estelle, L. (2019). Society Publishers Accelerating Open Access and Plan S - Final Project Report. London, UK: Wellcome Trust.