Improving Library Digital Catalog Search Experience

Newman Library at Baruch College
March - July 2016

The Newman Library has been working to address usability concerns throughout their library's website. This specific project focused on the search bar and the incorrect mental models most users had associated with it. With the ubiquitousness of Google, many people view search bars as powerfully all-inclusive, but library resources are silo'd and require more nuanced search behaviors to be successful. 

Project included: user testing, wireframes, mockups

PrOJECT Goals

Primary: Craft a search bar experience that gives the user a better chance of arriving at the content desired
Secondary: Educate the user base about the complex structure of resource management within the library

The starting point is pictured to the right. The main problem with the current search bar is that it defaults to "OneSearch" as the scope. Not only is the name misleading (it does not, as it suggest, search the entire set of library resources) but most users skip over the options entirely and do not even realize they are searching the scope that has been pre-selected.

Redesign

The Design Pattern
After considering a few different set up options, we settled on a dropdown pattern. Additionally, we consciously decided against pre-selecting a default scope. Previous testing revealed that users were almost entirely unaware scoping was even an option, and we theorized this was due to both the visual layout and "Google"-esque search expectations (trust the system to understand the query without further filtering). We decided to require users to make a conscious choice about the vertical in which to search in an effort to produce more successful search results as well as to educate the user base about the library's resources. 

Scope Choices
We also updated the search scope options. Notably, we removed the "OneSearch" option as its proprietary name was misleading. Many reference librarians had noticed that students were using the scope under the mental model that it would access all library records (which was technically untrue). By splitting out the options into non-jargon terms, we hoped the students using the search would have a clearer idea of what their query would include. 

User Testing

We debated on the usefulness of the "helper text." On one hand, including it readily aided our goal of educating users about where to find what they were looking for. But the length of the dropdown brought up some concerns. Through moderated, task-based user testing (with 12 participants over two days) we determined that the helper text made searches significantly more successful and we moved forward accordingly. A full report of the user testing including the protocol used can be viewed here: http://bit.ly/2cCxCSw

Mockups 

Based on the user testing results, we developed a set of mock-ups for implementation. We were under the limitations of modifying only the search bar, none of the surrounding content or the location of the bar. The UX improvements were coupled with a few visual upgrades for cleanliness. Due to the nature of the school year and staffing availability, the implementation will not occur at least until December 2016 between semesters.