Current user behaviour — analyzing link-server data


I’ve been analyzing some of the data for 2007 put out by our link server, Ex Libris’ SFX. The data I’ve been looking at are “clickthroughs” — in plain English, what the users click to be referred to the resources owned by the library. I won’t go into any great detail because this isn’t the interesting stuff, it’s the trending. Here’s what I found:

  • There were 189,037 clickthroughs in total
  • 94.5% of all clickthroughs came from just 16 sites
  • 80.6% from just six sites:
    • 32% from SFX’s A-Z list of eJournals
    • 19.5% from ISI
    • 9.6% from PubMed
    • 7.1% from Google
    • 8.5% from BIBSYS (our OPAC)
    • 3.9% from SFX’s browsing list

What conclusions can you draw from this data?

  • That’s a lot of traffic through this system! On average 517.9 clickthroughs per day
  • A lot of people (35.9%) know exactly which journal they are looking for (evidenced by their using the A–Z list and journal browse function in SFX)
  • ISI and PubMed important resources across discipline


Not every source links via SFX as we have other systems in place, so these figures represent a fraction of actual traffic.


Tags: , , ,

%d bloggers like this: