New business models for the open research agenda

Share this on social media:

The rise of preprints and the move towards universal open access are potential threats to traditional business models in scholarly publishing, writes Phil Gooch

Publishers have started responding to the latter with transformative agreements[1], but if authors can simply upload their research to a preprint server for immediate dissemination, comment and review, why submit to a traditional journal at all? Some journals are addressing this by offering authors frictionless submission direct from the preprint server. This tackles two problems at once: easing authors' frustrations with existing journal submission systems[2], and providing a more direct route from the raw preprint to the richly linked, multiformat version of record that readers demand and accessibility standards require.

How do we preserve trust in the scientific record in this open research world, and who will mediate it? Both preprints and open access have been criticised for enabling the rapid dissemination of unsound, fake or harmful information[3]. Proponents of this view have cited the furore over a coronavirus bioRxiv preprint, which suggested, with much hyperbole, that 2019-nCoV contains genetic material from HIV. The preprint was rapidly withdrawn following strong evidence that the results were spurious and false. Yet, isn’t this precisely how science should work? In this case, the community acted within days to disprove and remove bad science, whereas journal retraction can take years[4].

These arguments will continue, but it seems that both preprints and open access are part of the future scholarly communications. So how can both publishers and vendors thrive in this new environment? If scholarly publishing is in crisis, the public understanding of science is more so. Growing concerns about fake scientific news, poor reporting of research in some quarters of the media, and lack of trust, are having profound consequences – declining vaccination rates being just one example. 

I argue that these crises present an opportunity to reinvigorate both the supply-side and demand-side of research dissemination. 

Supply-side business models in an open, preprint-first world

New tools and services harnessing AI are emerging in the industry. Their aim is to reduce the time needed for desk-editorial screening of submitted manuscripts (which typically takes around 30 minutes) to one or two minutes. They achieve this by flagging problematic areas of the paper both from a structural and scientific perspective; verifying affiliations; and checking cited works for retraction, confirmation, refutation and overall quality[5]. These tools need to be integrated into the submission workflow, and platforms such as ScholarOne and Editorial Manager have begun trialling a number of such integrations. As it would be uneconomic for a journal to subscribe to all these services, and a number of them have overlapping functionality, we are likely to see some consolidation in this space.

A natural extension would be to apply this screening technology to preprints, both to ease the burden on volunteer screeners used by a number of preprint servers, and provide basic screening for those preprint servers that currently perform none. Improved screening, with the help of automation, has the potential to build trust in this content. With developments in technology enabling format- and domain- agnostic document parsing, this is an area we have recently been exploring at Scholarcy.

Demand-side business models

Dissemination of early-stage research as mobile-unfriendly PDF is arguably a technological step backwards. If preprints are here to stay, the reading experience needs to be improved. A number of vendors have developed native XML or LaTeX authoring environments which enable dissemination in richer formats. BioRxiv has been funded by the Chan Zuckerberg Initiative to convert PDF submissions to machine-readable content for web and mobile[6]. Regardless of the open research agenda, the need to process and reformat manuscripts into richer formats, using automated tools or otherwise, will continue to provide a business model for specialist vendors.

Making research available to a wider readership, both inside and outside academia, is important for visibility, impact and accountability[7]. And clearly explained, validated scholarship is vital for tackling pseudoscience and fake news. The popularity of events such as Pint of Science and The Infinite Monkey Cage, coupled with the rock-star status of some prominent scientists, suggest there is growing demand for mediated access to the latest research - if it can be communicated in the right way.

Business models providing a platform for the dissemination of research summaries to a lay readership have already been adopted by companies such as Sparrho and Kudos, and there are further opportunities to scale this up with the help of technology. A recent entrant is Science2Innovation, which connects industry to academics with in-demand skills via plain-language 'blitzcards' that describe industrial applications of the latest research.

A number of publishers request that authors supplement their work with a lay summary or set of key findings at submission. Some authors find this challenging, and given the growth in research output, writing summaries post-publication is hardly scaleable. As a result, Elsevier, Scholarcy and others are applying machine-learning to automatically create this information directly from the manuscript combined with background knowledge from verified sources. The technology is still somew way off from being able to generate an accurate, complete, plain-language summary that provides both the 'what' and the 'why' for a given research paper - let alone a collection of papers on a single topic. 

However, there is scope to apply current technology to generate a draft as a starting point for the author, or subject expert, to craft into a compelling story. The team at University of Manchester library have been using this technology as part of their OpenAccess+ service[8]. When a new paper is published, the UoM OpenAccess+ creates a Twitter thread which comprises a AI-assisted summary of the paper's findings, Altmetric mentions, and tags the funders and other interested researchers in the field. Steve Carlton, research services librarian, says: 'Our service helps researchers reach a wider audience, and removes some of the barriers to that audience understanding and using their work'. 


  1. Hinchliffe LJ (2020) Revisiting - Transformative Agreements: A Primer. The Scholarly Kitchen, 6 February. Available at:

  2. Birkhead T, Montgomerie R (2017) The Frustrating Process of Manuscript Submission. The Scientist, 10 May. Available at:

  3. Poynder R (2019) Open access: Could defeat be snatched from the jaws of victory? Available at:

  4. Oransky I, Marcus M (2020) Quick retraction of a faulty coronavirus paper was a good moment for science. StatNews, 3 February. Available at:

  5. Hinchliffe LJ, Clarke M (2019) Fighting Citation Pollution — The Challenge of Detecting Fraudulent Journals in Works Cited. The Scholarly Kitchen, 25 September. Available at:

  6. Callaway E (2017) BioRxiv preprint server gets cash boost from Chan Zuckerberg Initiative. Nature News, 26 April. Available at:

  7. Louët S (2019) Content marketing boosts open access adoption. Research Information, 25 November. Available at:

  8. Carlton S (2019) #OAWeek2019: Open Access+. Library Research Plus Blog, 21 October. Available at: