Extra Views or Fake News: Understanding Open Access Ebook Usage

2018 University Press Forum

[Editor’s Note: This essay also appears as part of the annual University Press Forum in the May issue of Choice and on ChoiceReviews.org.]

Charles Watkinson, Director,
University of Michigan Press

Whatever their basis in fact, numbers matter in public rhetoric. Was the women’s march in Washington really three times the size of the audience at President Trump’s inauguration? Was the Trump inauguration crowd bigger or smaller than the Obama inauguration crowd? And were these crowds equally motivated, more or less polite, demographically more or less diverse?

While the political stakes may be lower, quantitative and qualitative metrics of impact and indicators of engagement are becoming major challenges for open access book advocates and detractors alike.

Journals and books are not equal in their migration into the digital environment, and nowhere is this difference clearer than in the area of usage tracking and reporting. For journals, a range of bibliometrics and altmetrics exist, powered by almost ubiquitous stable identifiers (especially DOIs), sophisticated abstracting and indexing services, and agreed standards of reporting. For books, no index can claim any degree of comprehensiveness, and standards for usage data such as COUNTER are inconsistently applied. A particular challenge for book publishers is that most ebooks are not sold directly to customers from publisher platforms, but rather pass through a supply chain of intermediaries and aggregators such as Amazon, Ingram CoreSource, EBSCO, OverDrive, ProQuest, JSTOR, and Project MUSE. To a greater or lesser degree, these organizations view information about use as proprietary and share information inconsistently and in formats that cannot easily be compared. Project Muse and JSTOR are particularly good at sharing usage information with their publishers, while EBSCO and ProQuest are not so good. No organization, of course, is as secretive as Amazon.

At the same time, information about use and engagement is the currency of open access publishing. The promise of open access is that a publication will be downloaded, used, and cited more often. Open access publishers need to show such impact to continue to receive support. Funders look for this information to demonstrate return on their investments. Authors are eager to show evidence of their reach.

The existing challenges of tracking ebook usage are magnified by a proliferation of other hosting platforms, both those that provide services to publishers (such as OAPEN) and others that act independently by leveraging the open licensing of true open access books (such as the Internet Archive). There is also the emerging challenge of new forms of scholarship that explore formats beyond the traditional container of the book and make tracking even more complicated. While the publication of digital scholarship objects and open access are not mutually dependent on each other, innovation in the form of a publication and the business model that supports it often coalesce.

Several groups have been working on issues surrounding open access book usage, but their work has been happening in a siloed fashion. In the U.S., Project Meerkat was initiated as a 2015 Triangle Scholarly Communication Institute symposium that investigated the issues around “collecting and analyzing usage data for online scholarly publications” (all ebooks, not just open access) and proposed to develop a “publishing analytics data alliance” for all scholarly monographs. In Europe, the HIRMEOS project is developing shared infrastructure to support the effective integration of monographs into the European open science infrastructure, including the development of tools and services for the collection, aggregation, and visualization of usage data from multiple platforms. Meerkat and HIRMEOS have foregrounded nonprofit publishers, leaving to one side the important work that some commercial publishers are doing in this space. Multinational aggregators and platforms have published several important public reports that contribute to our understanding of the problem as well as conduct internal analyses. For example, in 2017 JSTOR made available its ebook data to Knowledge Unlatched Research to explore open access book usage, and Springer Nature released an analysis of open access book usage on its platform. Both reported increased usage of open access versus equivalent closed access books, while the JSTOR report also suggested encouraging patterns of disproportionate use in the developing world and at smaller institutions such as high schools and community colleges.

Support for open access book publishing will not grow unless a compelling case is made for investment. Little of its promise can be demonstrated without the availability of aggregated ebook usage data, agreement on what is important to analyze, and tools for reporting to authors and funders. What may be needed is a crossindustry gathering of stakeholders that can align around the challenges, agree on a framework for engaging with them, and propose a series of next steps. Some of the challenges are technical (how do you compare usage stats expressed in terms of chapter downloads versus those reported as whole book downloads?), but the larger ones revolve around governance. What type of structure could reassure competing companies enough to deliver usage information to a trusted third party? Meerkat has suggested a “data trust” arrangement inspired by practices in other industries—those who pay data in can take data out. What provisions can be made to protect the privacy of end users while still revealing useful and actionable intelligence about how open access books are used?

Until we can view usage across platforms, our best option for telling the story of the open access advantage may be altmetrics. Because any mention on the open web that uses a standard identifier (ideally a DOI) should be trackable, it doesn’t matter which platform the book is delivered through. Altmetrics providers should be able to pick up a signal of engagement that can contribute to telling a story about use. But how convincing that story can be when altmetrics is still in its infancy is open to question. Until we have agreed standards for aggregating, analyzing, and communicating open access usage data, we won’t know much about whether OA books are making the difference we hope. It’s just too easy to say “that’s fake news, folks.”

Back to the main University Press Forum 2018 page.

About the author:

Charles Watkinson is Director, University of Michigan Press.