This feature came from unique beginnings. One of our backend engineers (sup, Jason 😎) took the initiative to build a Pinterest integration within our Trunk Club apps during a hack day project. We predicted that allowing members to share their Pinterest boards with their stylist would increase engagement and potentially conversion metrics. However, until Jason gave this feature legs, we were having trouble prioritizing the needed dev and design attention to take this product to production.
I acted as design lead and partnered with our, then, new design intern (hey, Jimmy 🤗) to provide design support during the development process. This project was a great feature for Jimmy to jump in on since he would be able to see the product through the entire development process.
Constraints and goals
Because the feature was technically wired up and functioning, our main goal was to capture the value prop for our member, and design a flow that would support our member’s goal of sharing their style aspirations.
We wanted to design and build fast since this project was one of the several major initiatives on our plates, but we needed to do our due-diligence in creating a delightful experience for Trunk Club members.
The target user and their needs
The core of Trunk Club’s digital experience happens through a messaging platform. When a member signs up they are paired with a stylist. From there members and stylists message back and forth about the member’s style and wardrobe needs.
With the help of our PMs, we were able to sort through real member conversations to find examples of member’s sharing images from Pinterest or other sources of inspiration. We found that member’s were taking the initiative to share links to their Pinterest boards with their stylist.
From this information we drafted a simple HMW statement:
How might Trunk Club make it easy for members to share their style inspiration with their stylist?
Improving the design
Cleaning up the flow
Pinterest’s API was pretty straight forward, so we didn’t need to focus our design efforts too much on the first sign up experience. Instead, we combed through Jason’s hack day flow and highlight usability opportunities to create a more engaging experience.
Where should we hook this into the user’s experience?
There were a few places the Pinterest integration could live. Most obviously, the member’s account seemed like a natural place to house it. However, I wanted to push our design to feel like it was part of the messaging experience.
We explored different entry points for the feature. We also explored how we might make the Pinterest feature more discoverable.
We hallway tested our explorations to test whether our designs were intuitive. The experience that felt most natural was adding a Pinterest CTA (call-to-action) to our messaging compose bar. We knew this would push the development constraints, so I hosted a cross-platform presentation where I demonstrated the possible solutions for our feature.
The team agreed that adding the feature to our compose bar felt like the most natural UX and would allow us to further enhance the compose bar in the future.
With Jimmy’s help, we iterated through final mock-ups of the cross-platform experience.
Where things got messy
The start of this project was closely scoped out and the design team had the bandwidth to work ahead of the development cycle. However, because everyone on the feature team was also juggling other large projects small details began to slip through the cracks. Technically we underestimated the amount of work needed. From a design perspective, we became too distant from the project and had difficulty explaining small design decisions the team once made several sprints ago.
Overall, this product shipped on all platforms and we saw a significant improvement in member conversion by users who connected their Pinterest account. Member conversion is 39% higher for users who connected their accounts!
In the future, we will better scope our work and document our design decisions to ensure a smooth development process throughout the product build-out.