ER&L 2015 – Link Resolvers and Analytics: Using Analytics Tools to Identify Usage Trends and Access Problems

Google Analytics (3rd ed)

Speaker: Amelia Mowry, Wayne State University

Setting up Google Analytics on a link resolver:

  1. Create a new account in Analytics and put the core URL in for your link resolver, which will give you the tracking ID.
  2. Add the tracking code to the header or footer in the branding portion of the link resolver.

Google Analytics was designed for business. If someone spends a lot of time on a business site it’s good, but not necessarily for library sites. Brief interactions are considered to be bounces, which is bad for business, but longer times spent on a link resolver page could be a sign of confusion or frustration rather than success.

The base URL refers to several different pages the user interacts with. Google Analytics, by default, doesn’t distinguish them. This can hide some important usage and trends.

Using custom reports, you can tease out some specific pieces of information. This is where you can filter down to specific kinds of pages within the link resolver tool.

You can create views that will allow you to see what a set of IP ranges are using, which she used to filter to the use by computers in the library and computers not in the library. IP data is not collected by default, so if you want to do this, set it up at the beginning.

To learn where users were coming from to the link resolver, she created another custom report with parameters that would include the referring URLs. She also created a custom view that included the error parameter “SS_Error”. Some were from LibGuides pages, some were from the catalog, and some were from databases.

Ask specific and relevant questions of your data. Apply filters carefully and logically. Your data is a starting point to improving your service.

Google Analytics (3rd edition) by Ledford, Tyler, and Teixeira (Wiley) is a good resource, though it is business focused.

ER&L 2015 – Did We Forget Something? The Need to Improve Linking at the Core of the Library’s Discovery Strategy

Linked
“Linked” by arbyreed

Speaker: Eddie Neuwirth, ProQuest

Linking is one of the top complaints of library users, and we’re relying on old tools to do it (OpenURL). The link resolver menu is not a familiar thing for our users, and many of them don’t know what to do with it. 30% of users fail to click the appropriate link in the menu (study from 2011).

ProQuest tried to improve the 360 Link resolver. They focused on improving the reliability and the usability. They used something called index enhanced direct linking in Summon (basically publisher data) that bypasses the OpenURL link resolvers from 370 providers. These links are more intuitive and stable than OpenURL. This is great for Summon, with about 60% of links being IEDL, but discovery happens everywhere.

They also created a new sidebar helper frame to replace the old menu. The OpenURL will take them to the best option, but then the frame offers a clean view of other options and can be collapsed if not needed by the user. It also has the library branding, so that the user is able to connect that their access to the content is from the library, rather than just that Google is awesome.

 

Speaker: Jesse Koennecke, Cornell University

They are focusing on the delivery of content as well as the discovery. Brief demo of their side-by-side catalog and discovery search due to nifty API calls (bento box). Another demo of the sidebar helper frame from before, including the built-in problem report form.

 

Speaker: Jacquie Samples, Duke University

Does the website design for the Duke Libraries. They’ve done a lot of usability testing. The new website went out in summer of 2014, and after that, they decided to look at their other services like the link resolver. They came up with some custom designs for those screens, and ended up beta testing the new sidebar instead. They have a bento box results page, too.

The FRBR user tasks matter and should be applied to discovery and access, too: find, identify, select, and obtain. We’re talking about obtaining here.

ER&L 2015 – Monday Short Talks: ERM topics

Link
“Link” by Andrew Becraft

[I missed the first talk due to a slightly longer lunch than had been planned.]

Better Linking by Our Bootstraps
Speaker: Aron Wolf, ProQuest

He is a librarian trained as a cataloger.

Error reports are important, because for each one, there were probably ten instances not reported. Report early and report often.

Include the original query for the OpenURL in order to reproduce it. If you have the time, play around with the string data and see if you can “fix” it yourself and report that.

There are a lot of factors into how long it will take to fix whatever is causing the OpenURL error. They don’t want to raise false expectations by giving a date and time.

Once an error has been reported, it enters a triage system. If it has a broader impact, it will be prioritized higher. Then it’s assigned to someone to fix.

 

Trouble Ticket Systems: Help or Hindrance?
Speaker: Margaret Hogarth, The Claremont Colleges Library

We should be polite and helpful. Human.

Detail the issue as specifically as possible, with steps, equipment, screen shots, etc. Include account number or other identifier.

Vendors need to identify themselves in responses. They also need to include the issue in responses, particularly when the message trail gets long. Customers need to keep track of the trouble tickets they have submitted.

Respond promptly, even if it will take longer to resolve. Mine the trouble ticket data to create FAQ, known issues, etc. and add meaningful metadata.

Email is good for tracking the history. Online forms should have an email sent with the ticket detail and number. Some vendors hide their support email address, which is annoying.

If vendors require authentication to submit a ticket, provide examples of what information they are looking for.

Vendors should ask their most frequent support users for feedback on what would make their sites more useful.

Multiple tech supports make it challenging for reporting issues to large companies.

Jing screen casting is helpful for showing how to reproduce the problem, particularly when you can’t attach a screenshot or cast, since it provides a URL.

All of this is useful for your internal support ticketing systems, too.

ER&L 2015 – Understanding Your Users: Using Google Analytics and Forms

Google Analytics v2.0
“Google Analytics v2.0” by Panayotis Vryonis

Speakers: Jaclyn Bedoya & Michael DeMars, CSU Fullerton

There are some challenges to surveying students, including privacy, IRB requirements, and survey fatigue. Don’t collect data for the sake of collecting data. Make sure it is asking what you think it is asking to get results that are worth measuring.

Google Analytics is free, relatively easy to use, and easy to install. And it’s free. We’re being asked to assess, but not being given a budget to do so.

It’s really good about measuring the when and where, but not the why. Is it that you don’t see Chrome users because nobody is using Chrome, or is it that your website is broken for Chrome users?

If people are hanging out on your library pages for too long, then maybe you need to redesign them. We want them heading out quickly to the resources we’re linking to.

They’ve made decisions about whether to spend time on making sites compatible with browser versions based on how much traffic is coming from them. They’ve determined that mobile use is increasing, so they are desigining for that now.

They were able to use click data to eliminate low-used webpages and tools in the redesign. They were able to use traffic data to determine how much server support was needed on the weekends.

Google Forms are free and they can be used to find out things about users that Analytics can’t tell you. They can be embedded into things like LibGuides. There’s a “view summary responses” option that creates pie charts and fancy things for your boss.

They asked who they are (discipline), how often they use the library, where they use it, and what they thought of the library services. There were incentives with gift cards (including ones for In-N-Out Burger). The free-text section had a lot of great content.

The speakers spent some time on the survey data, but the sum total is that it matched their expectations, but now they had data to prove it.

#libday8 day 5 — queasy

funny pictures - can we fix it? yes we can!After a late start due to some unexpected things-that-must-be-done-now, I arrived and began to dig into the action items delayed from yesterday. This included responding to OCLC with information about a billing error, filing my notes from the discovery service presentation, and following-up on a related query from a colleague.

Added a new ejournal to our knowledgebase, but the default URL is different from what the publisher gave me. Added the custom URL and a note in Outlook to check on Monday to make sure the OpenURL linking works with the custom URL. Our KB provider does nightly refreshes of profile changes, so things we do behind the scenes aren’t live until the next day.

Unlike most of the reference librarians, there is one in particular who refuses to provide me with descriptions and coverage details of resources that go with their links on our website and LibGuides. I end up searching for descriptions on other library sites, and usually find something that will work. I added two resources for this librarian today, and rather than a simple copy/paste from the email generated by the form that every other librarian is able to send me, I spent about 20 min digging around for the information I needed. If it’s wrong, the only people who will care are the users, since I doubt this librarian even checks this stuff. This was evident because the librarian asked me to add three other resources that are already listed on the website and in LibGuides.

I tried to keep on slogging through, but the waves of nausea I’d been ignoring all morning were becoming harder to ignore. I decided it would be better to ride them at home than trying to work and maybe staying too long. Good thing I did, because the next 12 hours were very unpleasant, and the 12 after that less so. I’m posting this now a day late, and I’m finally starting to feel human again.

So, for library day in the life round eight, I’m signing out with a whimper.