25 Jun 2015

Bear V Shark (...or Discovery platforms V Google Scholar)

This post has been prompted by many sources: my own experiences as both a teacher and a researcher; a keynote at the recent Lir Annual Seminar on Thinking the Unthinkable - Utrecht University Library's decision to drop their discovery service and rely on Google Scholar; articles like this which highlight that whilst Google Scholar may be lacking in authority it offers an attractively comprehensive index; and this one too, which finds "no significant performance difference between the two discovery services [Summon & EDS] and Google Scholar for known-item searches. However, Google Scholar outperformed [emphasis added] both discovery services for topical searches" (Ciccone & Vickery, 2015*).

I'm perhaps a little unusual as librarians go, in that I firmly consider myself a pragmatist. A lot of the time, I subscribe to the 'good enough' approach when it comes to searching for information - enough being the key word here (caveat: obviously this is in the context of the specific user's needs i.e. most people I see are not doing systematic reviews or PhD level research but studying to get a degree and a job in the real world). Most students don't need every article ever published; they need a reasonable selection and more importantly in my opinion, the ability and skills to filter,evaluate and use them. For the most part, our graduates will not go on to be academic researchers; they will become teachers, entrepreneurs, office workers, policy analysts, managers. They will leave our libraries behind, and with it, their access to subscription databases and content. Where can they look for information now?

Most of us would agree that discovery platforms are not ideal for detailed, systematic or comprehensive searching, but rather work best when used as a scoping or initial searching tool to inform further searching on specialist academic databases. Indeed this is where they excel - and the same goes for Google Scholar. Yes, the volume of results is overwhelming. You would never be able to look through them all, but that's really not the point. Moreover, most library discovery tools are exactly the same in this respect. GS is intuitive, it's simple, and it's associated with one of the biggest brand names in the world.  When linked to a library's subscription resources, it provides easy access to paid-for content, as well as a lot of openly available content that may not show up in a discovery service (repositories and hybrid journals for example, for more on the latter see here). Sure, library visibility may be lower on GS than with an in-house or rebranded discovery platform, but if we are relying solely on generic, widely available publisher content to demonstrate our value to our users, we probably won't last long anyway.

Discovery services do offer one big advantage over GS, that is, the inclusion and integration of the library's catalogue. However, our catalogues are still very useful tools in their own right, and sometimes people just want to see what books we have on the shelves without the noise and inconsistent metadata that discovery services can sometimes open up. There is also the obvious risk that libraries might potentially become very dependent on Google, who could simply remove their Scholar service overnight if they so wished. This possibility is extremely unlikely in the short term in my view, and in the long run, well...

For some libraries, especially smaller ones, using GS instead of a discovery product may free up valuable resources (both financial and human) to concentrate on other areas where we can create much more value for our users, such as developing our unique and distinctive collections, and focusing on the expertise that our staff can offer our users. We are always claiming our users want a "Google search experience". If this is true, why not give it to them?

*Ciccone, K., & Vickery, J. (2015). Summon, EBSCO Discovery Service, and Google Scholar: A Comparison of Search Performance Using User Queries. Evidence Based Library And Information Practice, 10(1), 34-49. Retrieved fromhttp://ejournals.library.ualberta.ca/index.php/EBLIP/article/view/23845/17954


  1. Great post Michelle

    I have long been a fan of Google Scholar and would agree that it is a missed opportunity for libraries not to link their subscription resources to GS.

    I share your pragmatic approach - sometimes "good enough is good enough"!

    Kind Regards - NIamh

    1. Thanks for the comment Niamh. I have also received similar feedback via Twitter, so it seems we are not alone! :) I don't think there is a 'right' answer to this one, and different things will work for different people, but the old adage "if you can't beat them, join them" certainly makes me stop, think and ask questions about how academic libraries are looking at discovery in general.