Your OPAC sucks, part n: users make end runs around your resource portal

If you go back far enough in time you start to hear a chorus of complaints about library web sites.  I won't repeat them all now, but I will bring to your attention the latest salvo against library web design, this one from Inside Higher Ed entitled The Library Web Site of the Future from Steven J. Bell.

According to the Ithaka report, academic librarians rated the function of the library as a gateway for locating scholarly information as “very important.” Asked to assess the performance of libraries as their portals to scholarly information, however, faculty in all disciplines rated them considerably differently. Compared to earlier years of this Ithaca study, faculty no longer perceived the library as an important portal to scholarly information. While the library Web site is not specifically mentioned in the report, for the 21st century library, the Web site is the de facto gateway to electronic research content. The report makes clear that faculty increasingly access what they need elsewhere or simply find alternate routes around the library Web site to get to their desired library e-resources.

Bell calls out this effect (where the alternate referral path is almost always Google related) in this example:

• An articled titled “Measuring the ‘Google Effect’ at JSTOR” by Bruce Heterick appeared in the June 2008 issue of Against the Grain, and it documented the increased access of JSTOR content via Google Scholar. JSTOR usage has increased dramatically since its inception in 1997. But more recently a new growth wave is propelled by referrals from non-traditional sites. Heterick writes “another order of magnitude change in scale is introduced when we begin to look at the number of links coming to JSTOR directly from Google and Google Scholar.” The number of links to JSTOR articles from Google-referring URLs increased by 159 percent from 2006 to 2007. It’s just one more reason to avoid the library Web site as a research starting point.  

More on the JSTOR Google Effect:

What factors may have contributed to this growth? In addition to users accessing JSTOR directly through our search and browse interface, we have begun collaborating with search engines to provide additional pathways to the archive. Specifically, in early 2006, in response to requests from students, faculty, and researchers using the JSTOR archive, JSTOR finalized an agreement with Google to allow the popular search engine to begin "crawling" and indexing the book reviews and full-length articles archived in JSTOR for discovery purposes. A "crawl site" was created especially for this purpose, and the vast majority of archived journals available in JSTOR have now been indexed in Google Scholar, as well as Google's main search.

Navigating a big collection of materials is hard; it's always been hard, even when the material is in physical form (viz. the Harlan Hatcher Graduate Labyrinth).  If you look at sites outside of Google that generate a lot of referral traffic, it's because they have a lot of outbound links - not just links to the top level of collections, but also links deep within databases to individual items that are directly relevant to the question at hand.  If you had a 1, 2, 5, or 10 million deep research library, how would the top page of your web site be relevant to more than a handful of people?

Comments

[John Simmons]

It would be great if their search parsing worked at all. I can go in knowing the title of the book I want and yet the Catalog Search can never find it. I end up using Amazon.com to look up the book and get the ISBN and paste it into the catalog search. They have more videos than books. The "librarians" are only there to update aadl rss feeds. The signs are in Arabic but not English, when there are signs. No signs on the floors of the Main Library downtown. Super Smash Brothers tournaments. Abstract art built by a chimp friend of the Library. E-mail notifications for pick-ups that aren't there for two more days. The A2 ignorant smugness that we all love. Yeah, I am with you man. I have better results with GOOG and searching for the book title plus aadl. How broken is that when an outside engine works better than your own with system level access to the database?