Whisk — Retrospective

UXDI Project 2: Information Architecture + Mid-Fi Prototyping

Objective: Design a new version of a commerce retailer’s site
Duration: 3 Days
Solo Project

Premise

Whisk is a kitchen-goods retailer with a full range of kitchenware and tabletop items that aim to meet the needs of everyone, from the novice home cook to the professional chef.

The project brief called for designing a website for Whisk, with focus on the site’s information architecture. The main goal of the project was to organize 100 items (90 given and 10 chosen by me) of Whisk’s inventory into affiliated categories. How these items were grouped determined the structure of the sitemap and how the users navigated the website to find the items they were looking for.

The pedagogical objective of the project was not to re-design the existing Whisk website, but to learn how to structure a website’s information architecture within the context of a specified brand and appointed personas.

Personas:

3 personas with varying degrees of shopping habits and needs were considered in the designing of the website. The proposal had to address of the business goals of Whisk while satisfying the needs of these types of users.

Roland is the primary persona, who wants to quickly find specific items and easily browse for gift ideas. Most of the proposed website’s features revolved around the needs and pains of Roland.

The needs and pains to address for the Roland, the primary persona.

The needs and pains of two secondary personas, Daria and Trung, also determined some objectives that translated into features.

How we can serve the secondary personas

Contextual Inquiry

The project began with a visit to one of Whisk’s stores, located in the Flatiron District. There, I studied the layout of the store and observed the behaviors of how the customers shopped. It was apparent that the store was organized in what could be described as an organized clutter. Every available surface was covered with kitchen-goods — shelvings concealed every wall and boxes of items adorned every available tabletop. Also, none of the arrangements were labeled. Although it wasn’t too difficult for the customers to find what they were looking for, the arrangement of the store forced them to scan the entire store (which is probably by design).

Whisk doesn’t hide the fact that it has an extensive inventory of products. However, the physical store’s method of displaying does not translate to its digital counterpart. Without the visceral ability of scanning the inventory in its entirety, online shoppers must be provided a more organized presentation of the inventory.

The challenge was to take Whisk’s organized clutter and present the items in an intuitive and simple manner for the users to shop on the e-commerce site.


Research Method: Card Sorting

To determine how the 100 Whisk items were to be organized on the website, I had to first understand how people grouped these items and why. Using the method of card sorting, I facilitated 8 sessions where I asked sorters who were familiar with cooking to group the printed out items.

Open Card Sorting:

The first 5 sessions were open card sortings. The testers were asked to group the products in any categories of their choosing without any limitations to the number of groupings.

The types of groupings that emerged between the sessions varied a lot more than I initially anticipated.

Some testers organized the products by utility or types of functions, while others organized them by locations in the kitchen or even at a kitchen-goods stores that they would expect them to find. These varying approaches to sorting posed a challenge in finding reoccurring patterns of groupings.

The results from the sorting sessions were diagrammed with each groupings assigned a color. The groupings of each items were then averaged to derive a “final” category.

Open Card Sorting Findings:

15 items were toss-ups between two distinct groupings (the items with a blank cell as an average in the chart above). Perhaps this infers that such binary items should be found in both categories in the information architecture. Below are examples of these ambiguous items.

Closed Card Sorting:

3 additional sessions were administered as closed card sortings. For these, predetermined categories were given to the testers to organize the items that had ambiguous groupings. The results from these sessions were added to the average for the final groupings of the 100 items.

Sitemap:

By rearranging the items based off the results from card sorting, we can see the groupings of each items and the proximity of affinity between the groupings themselves.

These groupings are then arranged within a hierarchy to layout the sitemap of proposed website.


Research Method: Competitive Analysis

3 kitchen-goods competitors of Whisk, Sur La Table, the Brooklyn Kitchen, and Fishs Eddy, were studied to understand their user flows and good design heuristics that they implemented. Specific instances of flows and features that addressed the objectives of the primary persona Roland and the secondary personas were highlighted from these competitors. These instances were then adapted as features in the proposed website.

User Flow Analysis of Competitors:

Design Heuristics to Adapt or Avoid from Competitors:

Summary — Features and Design Decisions to Implement:

  • Offer a list of suggested items (homepage and product page)
  • Combine discounts / gift card code options together at checkout
  • Add supplementary gift options at checkout
  • Limit the number of subcategory or filtering options for items
  • Use simple and concise language for labeling navigations
  • Present “shopping cart” with number of items in cart and total price
  • Allow for quick overview of items in cart
  • Provide a page for gift ideas
  • Provide access to contact information in header and footer

Proposed Design

See the prototype in action! HERE


Final Notes

  • Although Roland’s need to discover gift ideas was covered in the proposed design, the “add gift option” features in checkout (gift wraps, gift message) should be implemented.
  • Conduct additional closed card sorting research to refine the hierarchy within the information architecture.
  • Prototype and conduct more usability testings to assess the effectiveness of features that were derived from Secondary Personas’ objectives, such as writing reviews and returning orders.
A single golf clap? Or a long standing ovation?

By clapping more or less, you can signal to us which stories really stand out.