During Summer 2015, I had the privilege of interning at ThredUP, the world's largest online fashion resale shop for women and children, with the Product team. My role was a mix of Product Management and Product Design, which allowed me to learn the process of making design decisions with the end goal of increasing onboarding conversion rates. My main project for the summer was redesigning Personal Shops, a personalization feature that saves a style profile of the users' sizes and favorite brands to enhance their shopping experience.

Problem

Only a very small percentage of users onboarded to the feature (not allowed to disclose figures), yet its users have the highest average order values.


Role

  • Led a strategic redesign by analyzing data and conducting interviews that led to key findings influencing user flows and KPIs set.
  • Worked closely with Product, Engineering, Design, and Marketing in the process of developing specs and receiving feedback
  • Created prototypes in FramerJS and facilitated user testing sessions that influenced rollout to another platform.

Defining the Product Strategy

Identifying Existing Successful Trends

I first pulled current data on users with Personal Shops to find the top used search filters and looked at which platforms (web, mobile/web, iOS, and Android) were the most popular and had the highest average order values.

An In-Depth Understanding of Users

I surveyed users with over 10 Personal Shops to understand their behaviors for saving so many shops in the first place. I followed up with 7 users who consented to following up with me for a more in-depth discussion about their use of thredUP and other shopping apps.

Based on my phone interviews with the users, I created a mental model to understand the user's train of thought.

Mental Model

  • Can I quickly find relevant items?
  • How can I find great deals that caters to my family's lifestyle?
  • How can I easily shop my favorite brands?
  • Will these items fit me?
  • Will it save me search time?
  • How can we get coveted items first?

Style Profile

We also further refined the idea of creating a Style Profile for every account in order to set Sizes and top 10 brands, based on the findings of a high percentage of users caring about their the brands they purchase and wanting to discover more items to similar brands.

Iteration

Validating Ideas

The aim of the first iteration (prototyped in FramerJS) was to validate and answer:
  • Would customers be interested in seeing recent searches, suggested searches, and popular searches?
  • Does saving the search into a separate tab ("Pocket") increase engagement?
  • After conducting the first user testing session on-site with 6 users who had at least one Personal Shop, I gained a lot of helpful insights for improving the next iteration.

    Findings

  • Simplify the view - remove color coding.
  • The concept of "Pocket" was too confusing.
  • Users wanted to be able to click through a feed row to see the PLP.
  • Users are not interested in seeing popular saved searches.
  • Refining the Design

    Since I had limited time to conduct another user test, I tested this prototype internally with select employees from Product, Engineering, Design, and Marketing to gather feedback from multiple points of view. I updated my FramerJS prototype based on findings from usability testing for version 1.0.

    Findings

  • We should rolllout to web, instead of mobile first.
  • There were mixed reactions to the heart iconography

  • Execution

    Challenges

    • It was tempting to include many new features for the redesign of Personal Shops, but I learned to prioritize features and make tradeoffs given the scope of the project. This simplification allowed for a more simplified design that was easy on the eyes.
    • I discovered the technical limitations for mobile after speaking with the Engineering team, so I had to quickly find an alternative solution for web.

    Takeaways

    • Involve other stakeholders earlier in the process to learn about potential contraints.
    • Simplicity and elegance rules. Think about information hierarchy when considering the visual. Only emphasize what needs to be emphasized.