Facilitating Expert Search through Information Architecture

Existing Site

Taking Inventory

Initial Discovery

I began this project by reviewing the brand's existing site. I immediately found that search did not function semantically, not allowing for genre-based search. Further, the site's curated categories were incredibly specific, broken down to typically contain a single piece of content, creating over 1,800 categories and a poor browsing experience.

Research

Mental Models of Browsing

User Interviews + Secondary Research

I began my research with user interviews and secondary research. I first wanted to learn what users/customers loved about the brand in order to improve their experience. My research focused on users’ mental models and expectations when browsing films and yielded some helpful results.

Research Methods

• Screener Survey

• User Interviews

• Card Sorting

Genres are...subjective

Card Sort Results + Synthesis

After creating a persona from my initial research, I wanted to learn how users would sort through and categorize the site's existing content and looked to the results of my card sort.

By the nature of VT's business, the majority of their content is unknown to a vast majority of users, making its categorization near impossible. To remedy this, I created a card sort using only genre/category identifiers to learn how users would group and sort films when content is stripped away. This card sort lead to some extremely valuable insights

Insights

• Even within users who are SMEs, the categorization of film based on genre and/or other factors is highly subjective

• Users’ mental models of hierarchy when browsing film incorporate multiple identifiers, creating a highly-specific search/browsing experience

• Users will only engage with curated lists of content with the knowledge they've been curated by experts

This card sort focused my entire design strategy. The results showed that the vast majority of users would sort and create categories with a high degree of specificity, narrowing down my offerings into categories of 5 items or less (of 35 total items). 

​

Along with specificity, users showed a high degree of subjectivity. There were almost no commonalities among the categories users created. Some operated with genre as their largest factor, while others did so by country of origin.

Design

Specific + Subjective Browsing

Design Strategy

These insights focused my overall design. I designed a search and filtration system for the site that empowers users to create their own highly-specific, multi-factor search criteria. I also wanted to allow my audience of subject matter experts the freedom to create categories for browsing based on their own subjective perspective.

Key Design Elements

• Multi-Factor Search

• Non-Mutually Exclusive Search Criteria

• Card Sorting

Empowered Expert Search

Information Architecture

My redesign was focused on empowering expert search. SMEs know exactly what they're looking for, whether it's a specific film or a multi-factored browsing experience.

Because these experiences and identifiers vary from expert to expert, I created multiple filter options, each with a wide range of non-mutually exclusive criteria. Used in combination, users can create browsing categories from the highest level down to a specific combination of niche genre factors.

Expertly Crafted Search Criteria

Takeaways + Next Steps

This project taught me a great deal. As one of my first UX-focused projects, my wireframes never progressed past the point of mid-fi. I focused my attention on information architecture and research. ​

​

I'm proud of the solutions I've designed, but am aware of the full technical overhaul of the site required for implementation.

​

In a live client project, my next would be to bring these designs to a point of high fidelity in collaboration with a visual designer, and conduct usability tests on my solution. Initial usability testing I conducted showed positivity from users towards search and browsing experiences. However, further testing of a more developed design would be necessary to move forward. 

​

​