Working with change

The e-content environment has changed rapidly in recent years. Just a few years ago, most publishers did not offer access to digital versions of content published prior to the genesis of e-journals in the mid-1990s. Now, many publishers have begun initiatives to digitise the entire backfile archives of their publications, some published as long ago as the 19th century. Such initiatives, as well as new titles being launched, have the potential to increase the amount of scholarly content available to users by orders of magnitude.

Digitisation of both new editions and historical material in books and reference works also expands the universe of existing scholarly material available in digital form. And the addition of completely new types of collections, such as media collections of still imagery, video, and audio, further increases the amount of online content available to the end-user.

There are also new avenues emerging that help users to discover and access all this additional content. The most important trend in this area is the inclusion of indexing information for scholarly content in popular search engines. Google, for example, now has an active programme for working with publishers to index content that formerly was hidden behind authentication mechanisms. With this indexing available via Google Scholar and with Google and libraries working together to include institution-specific authentication via Google’s Library Links programme, users now have an alternative path for finding and accessing scholarly content.

It is important to understand what a dramatic change this new environment represents for users. Formerly, they would usually begin their search for scholarly articles in an abstracting and indexing database to find interesting article citations and then follow links to full text. This was the primary mechanism for performing topic- or subject-based searches. If the user was looking for a known item, such as a specific journal or article, they would often begin in the library’s OPAC (Online Public Access Catalog) to locate a journal-level link to the publisher’s site and then drill down through the available issues to articles of interest. Google Scholar circumvents both of these content discovery avenues by allowing users to search for articles directly using the familiar syntax, style, and intuitive relevance ranking of Google. Having this access to article-level indexing information, including robust citation analysis, without having to use a specialised database or library catalogue as a starting point, allows a larger audience to access information.

Figure 1. MetaPress 2.0's dimensional navigation (right side of page)

User behaviour

The challenge for publishers is to build brand identity and facilitate discovery of their content in this new environment. Most users will never see the publisher’s home page or a list of publications offered by the publisher. Instead, the vast majority of users now arrive from search engines such as Google Scholar directly at the article level, where they have one-click access to the full text of the specific article that piqued their interest in the search results list. Studies of user-behaviour and web analytics data at MetaPress bear out this trend. In 2003, over 60 per cent of all users following links to MetaPress publisher sites arrived at the publication level. In 2006, over 60 per cent of users arrived at the article level. The total number of users arriving at the publication level continues to grow every year but the number of users arriving at the article level is growing at a far greater rate and will continue to do so.

One of the most common user-behaviours seen on publishers’ sites in 2006 can be described as follows: The user arrives on the site at the article level, scans the metadata (article title, authors, abstract, etc.) to verify that they have linked to the desired article, clicks the ‘Download full text’ link, and then immediately afterwards uses the browser’s ‘Back’ button to return to their search results list on a search engine such as Google Scholar.

User studies and session length statistics show the validity of the growing trends described above. Statistics from early 2006 show that close to 30 per cent of all linking traffic to MetaPress publisher sites originated from Google and Google Scholar. Those statistics also show that 58 per cent of all sessions on MetaPress publisher sites were less than one minute in duration.

A web-wide trend

This type of information-seeking behaviour, characterised by highly-precise linking to specific content and very short sessions at any given website, is a web-wide trend that affects all sites, not just publisher websites. In their 2006 book Prioritizing Web Usability (New Riders), Nielsen and Loranger noted that the amount of time spent on any given website is becoming shorter and shorter. Their research showed that most pages are viewed for less than two minutes. Sessions are becoming shorter but the total number of sessions is increasing. From an ‘information foraging’ perspective (see Peter Pirolli’s book Information Foraging: A Theory of Adaptive Interaction with Information, Oxford University Press, 2006), users are more likely to ‘snack’ on information today rather than consume the large ‘meals’ of years past.

An important goal for publishers in this new environment must be to find ways to meet more of the user’s information needs after their arrival on that specific article landing page, beyond just clicking on a full text link and going away. By offering links to related content and a clustering search engine that allows content discovery via link-following rather than typing, publishers hope to increase session length and keep more of the user’s attention for scholarly information focused on their own websites.

MetaPress 2.0, which was launched in October 2006, offers many new features that have grown from our understanding of user behaviour. Because users spend so little time at any given website, it is imperative that publisher sites eliminate obstacles that hinder users’ discovery of and access to information – every click is important. To this end, MetaPress 2.0 offers a clustering search engine with dimensional navigation that allows users to not only find relevant results quickly but to explore and refine those results by simply following links. Search results can be categorised along many dimensions (such as by publication type [journal, book, reference work], date of publication, subject, etc.). Those categories and associated refinement statistics (e.g. how many results of each publication type appear in the results list) are displayed alongside the results and users can refine results by simple link-following.

Our research shows that most users do not type in more than one search query during a session on a publisher’s website. So if a user does not find an article of interest in the search results produced by their initial query, they will often leave the site rather than attempting to refine the search by adding keywords. We have found that users are much more likely to follow refinement links than to modify queries via typing and are thus more likely to see and consume additional content on the publisher’s site when using dimensional navigation.

In addition, MetaPress 2.0 offers features on article landing pages that give the user more reasons to stay on the publisher’s site and discover and explore additional content. For example, using standards such as OpenURL, Digital Object Identifiers (DOIs), and CrossRef, MetaPress publisher websites ‘light up’ links to articles cited by the article that the user is viewing. Our platform can also go the other way and provide links to other articles that cite the article being viewed. Whereas links to references always point to older articles, ‘forward linking’ to articles that cite the current article are useful for finding more recent research on the specific topic that the current article covers. This type of citation indexing and linking can also give the user an indication of how important the current article was to later research.

A page-level article preview feature allows the user to preview an article’s individual pages directly on the article page without having to download a PDF file or open Adobe Reader. This feature can help the user make more informed decisions about whether to download the entire PDF article.

In addition, links to related content in e-books and reference works can be displayed directly on an article page, giving the user more relevant content to explore with minimal additional navigation.

The proliferation of digital content has changed the way that researchers search for information. Publishers must embrace new features to give their users the best knowledge of and access to their e-content.