A report into how readers discover journal and book content has been published by Simon Inger Consulting. The report is drawn from the results of a 2015 survey of over 40,000 respondents, and combines this with data from earlier surveys conducted in 2012, 2008 and 2005 to create a 10-year comparative study.
The full 66-page report can be downloaded from the Simon Inger Consulting website here.
Some of the most striking findings (to me) in the report are:
“increased role for social media in discovery”
As you’d expect and hope for in these modern times.
“no significant appreciation of social media sharing or article-level metrics”
But perhaps these functions are of high importance to those that do like them? Because I do really like them. It is also interesting that the respondents show an increase in popularity of discovering articles through social media, yet not be supportive of functions that facilitate this, or be interested in seeing what happens to an article once it reaches the social media sphere.
“Table of Contents alerts have reduced in popularity”
This is perhaps symptomatic of article-level searching overtaking ‘push-notifications’ of mass content. Readers are no longer interested in a full journal issue, but specific articles within that issue.
“Library discovery has shown some significant gains” / “Academic researchers now rate library discovery as highly as A&I”
Goes to show that everyone loves a good librarian!
Academics use Google Scholar more frequently that Google for searching (paraphrased)
The results around search engine preferences are interesting and cover several pages of the report.
There are big differences across subject areas – “People in Social Sciences, Psychology, Education, Law, and Business use Google Scholar more to find journal articles. However, people working in Humanities, Physics, Mathematics, and Religion & Theology prefer to use Google”;
Differences in regions - “Google Scholar is the dominant search engine used for journal discovery in China” and “[people] leaving Google seem to have shifted predominantly to Baidu”;
and effects of region-related usage on the access-types – Google Scholar is more adept at returning results that point to articles in free-access repositories. The study found that “The use of Google Scholar seems to be reduced in many African and Asian countries.” and that “For approximately 60% of the time, readers in high income countries in the academic sector are accessing articles from a free resource.” Therefore, if not for publishers then for science in general, there may be great benefits in advocating the use of Google Scholar across less wealthy nations.
“Publisher web sites are becoming more popular as a search resource”
This may be due to the more sophisticated functionality in publisher sites, such as reference linking or more nuanced marketing.
Relatedly, the study also investigated users’ opinions on several aspects of publishers’ websites and found changes in the features readers appreciated.
Figure 49: Journal web site features, 2005-2015 by Simon Inger Consulting, licensed under CC BY-NC 4.0
Features that increased in usefulness rating from 2012 to 2015:
Related articles/content suggestions. Increased from ~46% to 63% (17%)
This is now the highest rated of all features of publisher websites, and speaks to a preference for more intelligently personalised experience.
Reference linking. Increased from ~45% to 53% (8%)
This is related to the previous related-articles function, and is representative of the increased technology and functionality behind publisher sites.
Images for use in presentations. Increased from ~36% to 44% (8%)
Again, this may speak to more sophisticated technology in sites, and the adoption of software such as Figshare, which allows greater flexibility and greater scope for authors to supply detailed supplementary data. Or it could be author/user services providing useful resources.
Saved Search and Search alerts. Increased from ~21% to 25% (4%)
Again, speaking to the appeal of personalisation, and the more nuanced search options available.
Features that decreased in usefulness rating from 2012 to 2015:
Selected articles / Editor’s choice. Decreased from ~49% to 26% (23%)
Again, this may be symptomatic of most readers’ preference for personalised content and uniquely configured searching at the article level (strong theme!), as opposed to being directed towards content.
I find these results interesting in that, one purposeful role of journals in the community is to curate content for their target audience – often a group of field-specific readers. One would expect that if an Editor of a journal selects a particular paper, it would come as a strong recommendation, or imply some significance, enough that the idea of an ‘Editor’s choice’ article would be appealing. But it appears this may not be the case.
Table of contents alerting. Decreased from ~62% to 46% (16%)
Another illustration of reader’s preference to discover content at the article-level, rather than the journal-level. The usefulness of receiving a full contents list has decreased in favour of more targeted, user-sensitive notifications and functions.
News. Decreased from ~38% to 24% (14%)
The decline in popularity of this feature may be due to the more interesting news-type content being placed on blogs or in other areas of publisher sites, such as author services. Moving it from journal pages directly may mean it is less discoverable to casual users, and again, as with the social media sharing and article-level metrics, could perhaps be of high importance to people that use them, but of less importance to more casual or less heavily ‘involved’ readers.
Manuscript submissions. Decreased from ~43% to 33% (10%)
A strange response, which is not addressed in the report. Perhaps people who submit to journals do not need to use the publisher’s website to find them, despite them being related to the journal hosted on the publisher site. Or perhaps there were more respondents who did not submit articles in 2015 vs. 2012, so did not need this function at all?
Searching. Decreased from ~56% to 51% (5%)
Blank searching has fallen significantly, in favour of the more accurately directed linked-reference and intelligent suggestion functions. But perhaps contradictory to the 4% increase in popularity of ‘saved searches’, since you cannot save a search without having configured it first.
There are a great deal more very interesting observations to be found in the full report. Visit the Simon Inger Consulting website to download it here.
Simon Inger is on Twitter as @siox14. Discussion of the report can be followed on Twitter with #howreadersdiscover.