The Bentley: UX Strategy and IA

The challenge

How might a storied historical library improve public access to its archival records?

About the Project

This project rebuilt the information architecture (IA) for the Bentley Historical Library, an independent library operating within the University of Michigan library system. Prior to the project, the library struggled to connect website visitors with its wealth of archival materials that comprised thousands of collections and millions of records.

Current site model, which represents millions of archival materials ( View image )

Both exploratory and evaluative research informed solutions, uncovering dozens of users’ and stakeholders’ unmet needs.

Analytical, competitive, and comparative audits identified entry points, standards, insights and innovation opportunities.

Users and stakeholder workshops helped define key problem areas and jobs-to-be-done (JTBD).

11k+

Collections

Interviews

Personas

Audits

Extensive research

Participants shared a remarkably cohesive set of insights and opinions. They said the website’s information architecture (i.e. organization, structure, and labeling) sometimes led to users’ confusion and frustration. Foundational concepts (e.g., finding aids) were occasionally misunderstood, causing confusion about available research methods. Some users expected experiences conforming to traditional libraries, whereas others learned to appreciate Bentley’s internal standards, for example, the archival material handling rules.

Confusion stemmed from users’ lack of awareness as well as misperceptions about the institution’s offerings – a classic case of what, why, and how. What is the Bentley? Why should I care? How does it address my needs? Frustration resulted when the website did not adequately answer these questions.

"Bentley’s interpretive content was rich and fascinating; yet, it was not widely known or fully leveraged."

Although the Bentley website’s challenges were considerable, the core of its service offering was both compelling and robust once recognized by users. The collections provide users a wealth of archival materials and unique ephemera, from soldiers’ rosters to students’ dance cards. Bentley’s interpretive content was rich and fascinating; yet, it was not widely known or fully leveraged. Users were often delighted by Bentley’s customer-service oriented reference archivists and helpful staff. But realizing such a positive sentiment required a visit to the building.

Key findings were distilled into an answers-first format, deferring the review of detailed reports to more convenient times. ( View image )

Personas encapsulated the insights, thoughts, motivations, and pain points of real-world users. ( View image )

Analytical, Competitive, and Comparative Audits

An analytics review described visitor’s activities across the system. It identified main entry points and the relative user-interest of existing content. Such quantitative evidence bolstered qualitative insights revealed in our subsequent workshops and testing. Competitive and comparative audits identified standards, insights and innovation opportunities.

The audits revealed that key pages were disconnected from site architecture, often severing entrance pages from higher-quality content . With ten different layout looks and feels, pages spanned the gamut. Thousands of pages lacked clear and consistent headers, footers, and navigation.

Comparative heuristic review of key tasks [Nielsen 1994b] ( View image )

Audits and Analysis video presentation to institutional stakeholders

Workshops

Workshops offered stakeholders and users an opportunity to contribute their unique perspectives to the project. A detailed script drove the workshops, covering a range of research exercises, prompts, and goals. The design-thinking exercises elicited participants’ feedback and opinions. Complementing the interviews, the workshop helped us uncover areas of alignment and opportunity as well as disagreement and contention.

MIRO board from the persona workshop ( View image )

Multiple frameworks were employed to define the problem space — empathy maps, assumption mapping, More-or-Less, Three Priorities, How Might We (HMW) exercises, Lightening Decision rounds, and MoSCoW.

Prototyping & Testing

User experiences happen whether we plan for them or not. Prototyping allowed the team to solve UX challenges before encountering expensive and time-consuming visual design and development obstacles. Low-fidelity wires served as the blueprints for the experiences we built, thereby reducing project risk, fostering team collaboration, and expediting subsequent implementation. User-centered organization and goal-directed labeling guided the way to an improved information architecture.

"Prototyping [solves] UX challenges before encountering expensive and time-consuming visual design and development obstacles."

The new IA provided visitors an intuitive research and retrieval experience. A handful of broad categories led users to numerous resources, including both digitized collections (i.e. online materials) as well as finding aids that described offline materials. Retrieval was improved by gathering online materials under a new section titled "Materials Available Online." Reorganized polyhierarchical relationships (e.g., university, athletics, students) reduced content fragmentation.

Wireframe of the Discover screen ( View image )

Prototype and testing video presentation to institutional stakeholders