Reimagining & streamlining customization for a social media app

Shelf is a social media app that connects people through their shared love of music, books, movies, and games. Users can show off their personalities by featuring their favorite media, choosing themes, and adding trinkets.
Problem

Shelf users wanted more ways to customize their profiles but struggled to discover existing editing features. Additionally, the add-item flow was becoming increasingly challenging for the engineering team to support due to technical debt.

Solution

We consolidated all editing functions into a single, streamlined interface. Key improvements included dynamic menus and status indicators, while removing infrequently used features to reduce clutter. We also tested a new emoji trinket feature.

Impact

4x decrease in duplicate entries; increasing data quality

70% of users customizing themes in onboarding

+55% in users adding multi-media items across categories

Item menus are now the 2nd most used edit entry point

Role

Senior Product Designer
(interim Product Manager)

Timeline

July 2024 - August 2024

Team

1 designer (myself), 1 community manager, 3 engineers, CEO/Founder

My involvement

Conducted landscape analysis and mapped existing user journeys

Created and tested functional prototypes with users

Led cross-functional workshops to align priorities

Authored product strategy and product spec docs

Designed and iterated on new customization flows with community feedback

Project
objectives

Project objectives

What were we looking to achieve?

Improve visibility of customization flow

Make it easy for users to find and use customization features, increasing engagement of Shelf’s core user flow.

Design a cohesive system

Consolidate all customization options together into a single cohesive, scalable interface.

Create a delightful experience

Provide an intuitive and enjoyable experience that encourages users to personalize their Shelves, which will help drive retention and sharing behaviors.

Resolve technical debt

Enable future scalability and innovation for upcoming customization features.

Early insights

What compelled us to work on this problem?

High demand for customization

Customization was the 2nd most user requested feature, specifically around themes, trinkets, and item descriptions.

Higher engagement and retention

Users who edited their Shelves were 30% more likely to share and had higher retention rates.

Usability issues

Users rarely customized their Shelf - only 11% of active users edited items in the last 30 days. Many also added duplicate items, indicating confusion with the interface.

Technical limitations

Existing engineering and design constraints limited app scalability, particularly in adding more themes and customization options. We needed a more streamlined approach to customization and Shelf editing.

Process

What were the steps we took?

Landscape analysis

I analyzed 8 apps to understand how best-in-class apps handle customization and curation flows: Tinder, Hinge, VSCO, Unfold, Linktree, Shuffles, Spotify, and Pinterest.

Comprehensive flow audit

Since customization flows were scattered throughout the app, I created a holistic view of the entire system as-is.

Workshop

I led a cross-functional workshop to help clarify and align on priorities.

With technical debt and varying opinions on prioritization, the workshop helped determine what was feasible while balancing essential and nice-to-have features.

As a simple experiment, we opted to surface emoji reactions on Shelf items to explore user interest in custom trinkets.

Iterative design & testing

Since there was no way to track and compare certain user behaviors, we relied on user interviews and direct observations while user testing to both validate decisions and refine design iterations.

Initial designs & research insights

We interviewed 5 net-new users to understand NUX comprehension, and also surveyed 10 existing Shelf users.
Improved navigability and user comprehension

Overall, we noticed users showed a better understanding of how to add, manage, and customize items through the new flow.

Edit mode confusion

The edit interface wasn't intuitive - users had difficulty understanding their available actions, even with exposed drawers and controls.

Next iteration: We designed the new interface to feature a canvas-style interface with clearer, more tappable buttons to help users understand they're in edit mode.

Confusion around what items they could add

Users were struggling to know what had been added to their Shelves or if anything had been updated.

Next iteration: We added alert modals and status indicators to clarify how actions in the add flow impacted users’ Shelves, along with red dots to notify them of updates or changes.

Tutorials were overlooked

Users tended to skip through tutorials, rendering them useless.

Next interation: Replace the carousel tutorial with a guided spotlight walkthrough that highlighted important UI.

Personalized emoji reactions were a hit

Users really loved the emoji reactions on items as a way to customize their own Shelves and used them as ways to add commentary.

Final shipped designs

Making navigation simple and straightforward

We simplified the editing flow with contextual menus and a unified system. In addition, users can now add items directly without having to replace existing items, making the process flexible and straightforward.

Experimental emoji trinkets

To test user interest in personalized trinkets, we surfaced reactions as visual badges.

Result: Since shipping, 10% of weekly active users add emojis to their own Shelves—5x higher than adding to others’ Shelves.

Potential next steps: Focus on features that add more personalization and reward social engagement.

Streamlining item customization

We streamlined item editing by removing rarely used options that were confusing and didn't serve Shelf's core mission. We also changed "Delete" to "Archive" and softened the warning message to reduce user stress.

Guided walkthrough tutorial

We added in a walkthrough experience to highlight different areas of the UI.

Challenge: Because of the technical and time constraints, we were forced to ship this iteration.

Potential next steps: Focus on redesigning the add flow to improve user comprehension

Cohesive UI and status indicators

We redesigned platform items to match how they’d look on a user’s Shelf. We also added action indicators to improve tap affordance. We integrated alert modals and toasts to provide guidance without using extensive tutorials.

Adding to Shelf: a unique challenge

Within the customization flow, we discovered that users struggled to understand the different types of media they could add, and how to add them to their Shelf.
Differentiating item types

A single flow for all items was too streamlined - users had trouble distinguishing between different item categories (search, auto-updating items via platform connections, custom links, etc.)

Platform connection confusions

Users believed they could only add items from connected platforms, and additionally many users thought they could only search within auto-updating items.

Future-focused design explorations

Due to timeline and technical limitations, we were forced to drop the project. With our insights in mind, I designed a few concepts to inform our future iterations.
Making item categories discoverable

In Exploration #1, I exposed all item categories in the default edit screen. While this risked overwhelming users, it allowed them to see all their options at once and take action based on what they wanted to do.

Giving equal weight to all item categories

I added more visible sections in the add item flow (Exploration #2) to highlight the different types of content they could add to their Shelf.

Takeaways & reflections

What did I learn from this project?
Test early and often

With technical and timeline constraints, we launched the project. In hindsight, I think we could have caught some issues earlier with simpler prototypes.

Early collaboration helps mitigate later conflicts

Aligning all stakeholders early on helped us balance different team priorities and made the whole project run smoother. This project really showed how important it is to prioritize, find middle ground, and keep communication clear.

Fix usability for better experiences

While tooltips and visual cues help teach users about customization options, I learned first hand that fixing the core usability problems is what really makes for a smooth experience.

That's the end of this project!

That's the end of this project!

That's the end of this project!

That's the end of this project!

That's the end of this project!
  • let's chat

  • ·

Lydia Ding © 2025

  • let's chat

  • ·

Lydia Ding © 2025

  • let's chat

  • ·

Lydia Ding © 2025

  • let's chat

  • ·

Lydia Ding © 2025

  • let's chat

  • ·

Lydia Ding © 2025