We are used to the way things are. Why not continue with our current methods?
A library is more than just 'stuff.' We provide services, data, tools, and perhaps most importantly, expertise. The current multi-tabbed search box is disjointed, and even with its multitude of options, still doesn't fully expose everything OSU Libraries offers. Further, website analytics indicate that users are often confused about how to find resources. For example, the WorldCat Local search box has been used to search for library hours; the website search box has been used to search for a database. We are making our library community work too hard and as a result they are finding too little. We must provide them with a better tool to discover those rich resources, databases, research guides, and people with expertise in their subject areas.
Read more abou OSUL's Working Principles for Discovery.
What is the plan for improving discovery?
The Libraries IT department is developing a hybrid approach, one that mixes local development with vended content to allow OSU Libraries to create their own discovery experience. It will be working with a vendor to provide discovery to our vended content via an API (Application Programming Interface), but for local content (all content controlled by the Libraries), developing local indexing and a custom Platform. (Read this overview of Library as Platform, if you like.) Our hope is to utilize a process of frequent refinement throughout summer 2017, so that we can arrive at a tool that will simplify discovery for all users, but primarily lower barriers for our undergraduate populations. We would like to begin deprecating WorldCat Local by 2018.
Concurrent Work Streams
- Discovery Layer - the Discovery Working Group will focus on prioritizIng and selecting of initial resources, and alpha test the initial bento-box proof of concept design [draft mockup (pdf)]
- Library Platform API - the Libraries Information Technology department will investigate technology options/available APIs, create endpoints for various content sources and develop API data models
- Resource Selection - focusing on serving our primary user audience, evaluate data availability of key resources
- Information Propogation - Complete Library Platform API, and work with Web redesign effort to embed rich data (schema.org) into content
Functional Experts (for both website and discovery redesign)
- Consultations with libary staff who have expertise in various areas to better inform the redesign projects [(blog post | full notes summary(docx)]
UX Cohorts
- User Experience Cohorts will assist with ongoing usability testing, such as card sorting, focus groups, survey and data analysis.
Beta testing
- As the proof of concept is iteratively developed, advancing from prototype to pilot to live production, the platform will be continually tested. It will be initially shared with a few targeted focus groups, then released to more audiences througout the year.
Wouldn't it be easier to just switch discovery to EBSCO EDS or other vended service?
A big benefit of the hybrid approach is that on the back end, APIs can be tweaked or even swapped out with minimal interruption for the user. Changes can be made without extensive downtime or change in staff workflow.
Can I see some examples of what our new discovery platform will be like?
Absolutely. But keep in mind that we are developing a Buckeye-centric experience based on evaluating user feedback and web analytics gathered from the OSU community. Therefore, these examples can only hint at how our discovery platform will function.
Type in any search term to see how they function:
- University of North Texas Libraries Search it All | Example result
- North Carolina State Libraries Search All | Example result
- Stanford Libraries Search All | Example result
- Other examples
When will the project be completed?
A good portion of the indexing work should be completed by Fall 2017, at which time user testing and analysis will begin in earnest. The user-centered design will progress from proof of concept to pilot, to fully-functioning platform service throughout the academic year. The intention is to begin deprecating older services some time in 2018.
How will we know if we have accomplished our goal to improve discovery?
The most effective way to measure success is through user-centered design. We will track improvement by making ongoing comparisons of usability data that we began analyzing in Spring 2016 for the Discovery Inquiry Proposal.As User Experience Leads for both the website and discovery redesign projects, Robyn Ness and Meris Mandernach began performing usability testing in April 2017 that provides even more data to guide us during this process. These presentations summarizing their findings:
OSU Web/Discovery User Testing (pptx)
Discovery Project User Testing (pptx)
Analysis of OSUL web site visiits (including the library catalog)
Analysis of User Survey Feedback (including discovery behavior)
When the discovery pilot is ready to go 'live,' various groups will participate in beta testing, and their feedback will inform any updates to the service.
How can I help?
We value your participation and feedback!
- Participate as a UX Cohort
- Subscribe to updates to the Information Technology blog and NewsNotes
- Provide feedback through this form