I evaluated the initial features of Pinterest— search, save, collaborate—in order to test if its design is optimized for what it says it does.

Mixed Methods

For this research project, I evaluated Pinterest with a mixed-method approach. I designed and conducted both a cognitive walkthrough (usability inspection research method) and a semi-structured interview in order to collect data about user tasks and record user actions and comments.  I aided these with an exit interview to compare thoughts before and after the study. Users were divided into two categories–naive and expert– with five users in each category.

Cognitive Walkthrough: User Testing

Cognitive walkthroughs evaluate the system’s flexibility and ease-of-use because the user engages directly with the interface. It’s a structured inspection method as I assigned the what goals that the user should accomplish. This method allows me to assess if the system is intuitive and usable.

Because Pinterest is a collaboration tool, the key tasks that I identified were: 1. Search for an idea, and find three different ways to filter results, 2. Create a “board” and pin the picture, 3. Invite someone (me) as a collaborator on the created “board.”

Semi-Structured Interview

I created an interview guide with basic sets of questions in order to both have control of the questions and time and have the flexibily of conversation in case of divergence from the order of questions. Interviews help uncover hidden motivations and concerns in the form of personal expression and body language.

Some of the interview questions include:

• What are you trying to accomplish when you pin something?

• Of all the approaches you tried today, was there anything that was especially easy for you? Why?

• Are there any scenarios where you will use a different tool instead of Pinterest for a similar purpose? Why is that?

Analysis

Observational data were coded in terms of success rate, time on task, number and types of errors, number of navigations to help the user, and number of clicks to complete the task. Interview data were coded in terms of characteristics of the user, contexts that Pinterest is used (physical environment, social context, and organizational context), and tasks (scenarios and use cases).  Subjective recommendation ratings were also compiled for each participant as a measure of satisfaction.

The patterns of both the qualitative data and quantitative data for each participant painted a story of the motivations and concerns of the users and helped prioritize design recommendations in terms of severity.

Findings and Design Recommendations

The main strengths of Pinterest identified and executed by both naive and experienced users were the consistency of metaphors and the emphasis on recognition as opposed to recall. The system makes exploration and bookmarking intuitive, so the users do not have to wonder if an icon or action represents more than one task.  The visual representation of themes makes it easier for the user to recognize which pins they want to select.

The main weakness of Pinterest is lack of error prevention features that communicate consequences for the user’s actions. Users experienced frustration when there was no feedback when a pinned picture was saved twice or when it was deleted. Both types of users also expressed frustration in bulky filter options and expressed satisfaction with only having the search bar and recommended pins.

Desing recommendations include removing the cluttering number of filter options. The data suggests a diminishing return as more filter options are presented to a point that it frustrates both new and returning users. Onboarding and learnability for users can also be optimized by adding cues when a picture is saved in a board, if the picture already exists in the board, or if a picture is being deleted.