User Research To Completed Product: Spotlight

Spotlight - A Self-Service Tool for Showcasing Digital Collections

Project Motivation

Work on Spotlight began in 2014 with the recognition that Stanford University Libraries' (SUL) many impressive digital collections were a) difficult to discover, and b) limited in terms of how collection curators could publicize and provide rich access to them. For example, the screenshots at left (click image to view all) show how curators can create (using Drupal) a simple "mini-site" within the SUL website to highlight a collection, but this option doesn't enable users to explore items in the collection or view item metadata, provides limited flexibility in constructing the mini-site pages, and offers minimal visual appeal.

My group had previously produced custom-built sites that provide a compelling user experience for individual digital collections (e.g., Revs Digital Library, Bassi-Veratti Collection), but these were multi-month development efforts. We needed a way for curators, instructors, and other non-technical people to create their own attractive, full-featured digital exhibits, without developer involvement.

Environmental Scan

I began my research by gathering information and impressions from three main sources. First, I needed to better understand and be able to communicate the gap in SUL products that Spotlight might fill. Our digital collections are stored in the Stanford Digital Repository and can be referenced through a variety of means: blog posts, announcement pages, the mini-sites mentioned above, our online library search catalog, etc. The second and third images at left are diagrams I made to document and share our digital collection ecosystem and what it lacked that an exhibits solution might address.

Next, even though I knew Omeka, a popular digital exhibit platform, wasn't a good fit for us, it could serve as a good model. I analyzed existing Omeka sites and installed Omeka locally so I could build my own exhibits with it. Finally, I surveyed many other museum and exhibit sites, analyzing them to better understand how they handled searching and browsing, what special "exhibit" features they offered, and the qualities that the best exhibit sites had in common.

Stakeholder Goals

Although giving curators a better, self-service option for creating digital exhibits was a primary motivation for beginning the Spotlight project, there were other factors I needed to consider in my research. Many of Stanford's collections are donated and we wanted donors and their contributions to be more visible, both to show appreciation and to encourage other donations. Stanford Universities Libraries wanted to better expose the increasing number of digital collections it possesses. My group, Digital Library Systems and Services, is a leader in the digital library community and we wanted to initiate an open source exhibits solution that others also needed, but lacked the resources to develop on their own. Part of my research, therefore, involved talking to these various stakeholders to determine and document what their unique goals were.

User Interviews

My next major phase of work was conducting interviews with potential users of Spotlight. I developed a series of primarily open-ended questions and interviewed people at Stanford who would likely be involved with digital exhibits in one capacity or another. These interviews took place at the interviewee's workspace so I could better understand the environment and context in which they did their work. (The image at left, for example, is a display case from a physical exhibit curated by one of the interviewees.) I audio-recorded the interviews and had them transcribed so the results were easier to work with.

In addition to providing insight into how curators wanted to showcase their collections, how they wanted end-users to experience them, and what their technical skills were, interviewees brought up many questions that required me to explore feasibility options with our developers and management. The second and third images at left document some of these aspects of the user interviews.

Stakeholder Approval

Drawing on the user research and discovery work I outline above, I prepared a conceptual design document that summarized the work I had done and made a case for why a self-service product for creating digital exhibits made sense for SUL and the broader community. The document included concept diagrams to illustrate the gap in our current products. Personas conveyed the types of users who would benefit from Spotlight and scenarios for how they would use it. I included prioritized feature requirements to convey the scope of the effort, and a conceptual site map to help illustrate the main components of the Spotlight application (examples of these are shown at left).

I presented the conceptual design document at a meeting that included the various stakeholders. I answered questions, obtained useful feedback, and the project was approved and scheduled to begin development within a couple of months.

Information Architecture

I next developed the information architecture for the product. This included site maps that illustrated the application structure at high-level, for both the end-user part of the application and the creator, or exhibit builder, part of the application. With the major elements of the application mapped out, I then created wireframes that detailed the components and layout of each page of the application (a few examples are shown at left). There were about 50 pages of wireframes for the initial version of the product.

Before beginning development, I shared the completed conceptual design and information architecture documents with the digital library community, including in a conference presentation and poster, to ensure other institutions were aware of our plans and to obtain broader feedback on the designs.

Development

A small team (three talented developers and myself) began development on Spotlight in early 2014. Working from GitHub issues derived from the wireframes I describe above, we completed a first version of Spotlight in Spring, 2014. My role on the development team was to create work tickets, answer questions and explore options for design and interaction details as they arose, handle the visual design, and on occasion contribute front-end related code commits.

Because our group develops and manages many products, after a few months we had to move on to other projects. As the GitHub activity graph at lefts shows (in green), however, we've since returned to Spotlight for two other ~3 month development cycles.

Iterative User Testing

A critical part of Spotlight's product development cycle was user testing. As soon as most of the main features were in place, I began formative evaluation to find out how well the product was working for its intended users. I scheduled two-hour sessions with several Stanford curators who had previously created "mini-sites" for their collections using our library's Drupal-based features. Because Spotlight is intended to be a self-service application, I basically just asked the test participants to try to create a digital exhibit for a small group of their collection items, without any instruction. I watched and took notes of application errors, tasks users found confusing or failed to complete without help, and questions that came up while completing tasks.

Results from these formative evaluation sessions fed directly back into the development cycle. I created issues for bugs that users uncovered and, in some cases, for new features that were prompted by my observations during the test sessions.

Product Release and Impact

Spotlight has not yet been released as a version 1 product, but it is open source and publicly available on GitHub. We're slowly rolling it out at Stanford and currently have five public exhibits, with many more planned or in progress.

As we've discussed Spotlight and demoed it at various conferences and community developer meetings, there has been a great deal of interest from other institutions in adopting it as their digital exhibit solution. We know of at least a half dozen institutions that are working on integrating it into their digital repository environments and have realistic expectations of seeing Spotlight in production at other institutions in 2016.