by Kristin J. Henrich and Jeremy Kenyon
In the Fall of 2008, the University of Idaho Library Reference Department began exploring opportunities for expanding reference services, which at the time consisted of in-person, phone, and e-mail reference. Librarians began researching emerging technologies and their application for reference, and the department decided to pursue adoption of instant messaging (IM) reference. We had several questions to answer before selecting a software and implementing the service.
Who is our audience? Our primary audience included traditional college students, used to online multi-tasking, but also non-traditional students and faculty and staff. In addition, more UI students were taking distance education classes in remote locations, and the department felt that IM reference would provide additional outreach to these students.
How are peer institutions implementing this service, or are they? A survey of peer institutions (land-grant universities) found that 66% offered chat or IM reference, either in-house or as part of a larger consortium service.
What are our objectives in adopting this service? Our objective was to provide synchronous reference assistance to users in a relatively easy, effective, and low-barrier way.
Finally, and perhaps most importantly, what potential barriers are there to beginning this service? Our department was lucky to have administrative support for the project, and librarians were generally positive and enthusiastic about implementing this service. Although cultural barriers were low, technical barriers still existed: software would need to be selected and installed, librarians would need to be trained on the new software, and the service would need to be marketed to patrons.
What features do we need in a software? The library wanted software we could install and maintain ourselves, with widget functionality to embed in our library website. The software also needed to be open-source; like most libraries, the department did not have funding to adopt a proprietary IM client. We also wanted our users to be able to add UILibrary to their “buddy list” or simply to ask a question anonymously through the library website. The open-access social networking company Meebo fit the bill with their product Meebo Messenger.
To implement the software, the department created an account within Meebo Messenger, and also created accounts on four popular IM clients: AOL, MSN, Yahoo, and Gmail. Syncing these clients with Meebo Messenger allows patrons to use their preferred platform to ask questions, which then feed into a central interface for librarians. Instruction on the software was also conducted for librarians, and included a group demonstration, one-on-one practice sessions, a “cheat sheet” at the reference desk, and a “soft launch” in the Summer of 2009 to prepare for the formal rollout in Fall 2009.
The Reference Department officially began offering IM reference in Fall 2009. That semester, IM questions accounted for 2.9% of all reference questions, while IM questions accounted for 5.5% of reference questions asked in Fall 2010. Informal assessment has indicated that the service is popular with users and librarians, and the department may need to investigate a more sophisticated IM client to accommodate the increase in use by patrons.
The launch of IM reference was so successful that the Reference Department wondered what other emerging technologies would lend themselves to innovative reference services, and agreed to explore the implementation of SMS (text) reference. A recent study by the Pew Research Center found that 18-24-year-olds send, on average, 109 texts per day: that demographic comprises the majority of our users, so why not leverage a technology students already use? Other motivations for exploring the service were the increasing presence of mobile technologies by users (and librarians), the need to keep up with technology trends in other academic institutions, and technological advances which opened the possibility for open-source SMS reference.
While exploring available technologies, the authors discovered a blog post from Ohio University librarian Chad Boeninger, in which several freely available technologies were used in conjunction to implement a text reference service. After some evaluation, the authors decided to try it out at the reference desk, limiting the service to their personal mobile devices. The three technologies ingeniously mimic a telephone–there is a receiver, a ringer and a cord to connect the two.
The “receiver” is Google Voice. Launched in 2009, Google Voice is a voice-over internet protocol (VOIP) tool which enables a user to maintain a telephone presence without purchasing a phone line from a traditional phone company. It also enables us to perform text reference from a web browser, rather than a dedicated mobile device. On signup for the service, Google registers a phone number that the user selects given the user’s location and the list of available numbers. For the reference desk, we selected a number with a 208 area code largely for symbolic purposes, since long distance is less of an issue than in the past. Before using it however, there are two very important settings to implement–call forwarding to the primary reference telephone number, and text forwarding to Google Mail (Gmail). Call forwarding permits the library to maintain its traditional phone number for reference services over phone, meaning the Google Voice service primarily is used for its text capacity. Gmail is used because of its natural integration with Google Voice.
Gmail functions as the “cord” to connect the receiver, Google Voice, to the ringer, Trillian. Because Google also has an instant message service, called Google Talk, which requires a Gmail account, we are able to connect Trillian–an instant messaging client–via Google Talk/Gmail to the texts received by Google Voice.
Trillian’s only purpose in this process is to notify the reference staff of a new text message. That is how it plays the role of “the ringer.” It produces an audio signal when it receives the message from Gmail. Without it, at our busy reference desk, notifications are often missed if they are not accompanied by an audio sound. And in general, we have found that one does not want to use commonly heard sounds for one’s reference environment.
Setting up the service required a fair amount of attention to detail, but once established, using the service functioned much the same way as IM. An audio sound notifies the reference desk, a librarian views the incoming message, and types a response—albeit one of 140 characters!
Text reference was implemented in Spring 2011, with a soft launch in Fall 2010. Training followed the same pattern as that of IM reference, featuring a demonstration, one-on-one training, and guidelines posted at the reference desk. Early assessment indicates that text reference is not being adopted by users at the same rate that IM reference was adopted (1% of reference questions), although text reference has only been offered for a semester and a half. There are several possible reasons behind the slow adoption: students may not be using their cell phones as anticipated; increasing adoption of smart phones and corollary internet access may have negated the need for text reference; patrons may not remember the text number, which is by necessity different than the pre-existing university-assigned phone number. The reference department plans to continue offering the service and reevaluate its effectiveness in Spring 2012.
Text reference served as an initial foray into providing reference services for a mobile environment, and so a natural next step for the library was to develop a mobile-specific interface for our website. The mobile website was designed to support the increasing number of library members using mobile devices for much of their communication, studying and learning. Many public services librarians at UI have remarked on the range of patrons who use their phones to record call numbers and other bibliographic information before asking for help at the desk. We felt it was appropriate to provide these and other users with more mobile-friendly interactivity.
Our chief task in design was making priorities of our website content. Not all content can be displayed well in the mobile environment, and frankly, not all content needs to be. We settled on three categories: contact information, library hours, and the library catalog. This philosophy of simplicity over comprehensiveness may be unusual to most librarians, but it was felt by the designers to be essential for creating an efficient but usable mobile interface.
Simplicity also allowed us to retain one of our primary goals: keeping the page “shallow.” This means that the website has a user move a maximum of three pages into the website before either reaching a page with no further sub-pages, or reaching a page which only contains links off of our website, e.g. to bibliographic databases. What this means is that users are never more than two clicks down from the homepage and this restriction forces the designer to keep the entire website simple and clear.
The first iteration of the mobile website was developed very quickly by a member of the library staff. This design was never intended to be permanent, simply an exercise in creating a mobile alternative to the library’s website as our existing website did not display well on mobile devices. The content was discussed briefly by the library’s Web Committee–which approves all major web-related changes–and approved to become publicly accessible. Over the next year, however, there was a sense that the mobile website neither exploited our resources fully nor provided a consistent experience together with the traditional library website. Specifically, there were three major problems: 1) user interface, 2) branding and 3) functionality.
First, our links were not spaced properly. For touch-screens, such as on the iPhone or Android devices, links need to have sufficient space around them to prevent a user from accidentally clicking on the wrong link using their fingers. Also, each page was exactly that, a separate page. This meant that to move from one page to another required a separate request to the server; a new page had to load. If network connectivity problems occurred for the mobile user, the request had the chance to fail and the new page would not load. Thus, fewer requests mean there is less chance of interruption.
Second, our website was only partially branded. For a university library, brand affiliation with the larger university is essential. It reminds users “where” they are on the Internet, who they are affiliated with, and provides a visual consistency that corresponds to their own identity, either as members of the University of Idaho community or as visitors from outside. The University of Idaho, like many other higher education institutions, maintains a brand and style guide governing the visual output of the institution’s various departments and units. While we did not have specific mobile guidelines, there were color and image specifications that were not consistent with our main website or the university’s.
Third, the mobile website failed to exploit all possible mobile-friendly resources or communication protocols of modern mobile devices. For example, a number of our databases had developed mobile versions, but had yet to be activated by us, or added to the website. Further, our contact page failed to access the text messaging functionality of a phone when a user clicked on the text reference number, or the telephone functionality when a user clicked on the phone number.
We also enhanced the visual style of the page using the jQuery Mobile design. Aside from widening the space around the links, providing cells around the links and “Back” and “Home” buttons, it also provided a simple opportunity to brand the colors according to the University of Idaho requirements. We added the gold and silver colors, as well as improving the logo, and the page took on a dramatically improved, appropriate and consistent tone.
Lastly, a thorough inventory of our databases was conducted and all mobile-friendly versions were added. EBSCO-based databases required activation by our local EBSCO administrator, and all database links were prefixed with the URL for the proxy server, thus requiring authentication by all users. This is especially important given that many users will be accessing the site from non-University networks, such as wireless carriers Verizon or AT&T, even if they are on-campus. Therefore, the database servers will not recognize their IP addresses. Finally, we gathered the appropriate URIs for different types of communication protocols and encoded the contact information with the correct ones, e.g. tel: for telephone numbers, mailto: for email addresses, and sms: for text message numbers. This important addition allows users to simply touch a phone number and automatically activate the calling function on their device, for example.
In all, these changes made the website much more clean, effective and useful for users. We are still updating the website as necessary, but are very happy with the most recent version.
The initiatives described above are presented as models for other libraries seeking to implement similar services on a budget. Each project was implemented using open-source software, which means no bills for either software or server space. Each initiative is relatively low-risk and takes a minimal amount of time to install or implement (budget several hours for each), although time spent planning, testing, and training librarians can often, and should often, take longer. In particular, the IM and text reference implementation was simple enough that any library, regardless of technical expertise, could replicate our experience, although creation of a mobile library website may require assistance from library IT personnel. The University of Idaho Library has been very pleased with these services, which have allowed us to expand and enrich our services to reach previously underserved users and take advantage of new and exciting technologies.
For more information about the University of Idaho’s mobile initiatives, see http://www.idaholibraries.org/files/Mobile_Initiatives_UI.pdf.
Kristin J. Henrich and Jeremy Kenyon are Reference & Instruction Librarians at the University of Idaho.