Letterboxd
Adding a collaboration feature to a social movie platform
Letterboxd is “like GoodReads for film”
Letterboxd is a social network for film enthusiasts whose goal is to promote film discussion and discovery. Its features allow users to keep track of films watched, want-to-watch lists, rate and review films, follow other users, and curate lists around a common theme.
ROLE
UX Design, UI Design, User Research
TOOLS
Figma, Maze, Zoom, Otter.ai
TEAM
Group critique & mentor sessions
CONTEXT
Concept project
Overview
Letterboxd has a strong focus on film-centered social interaction, but users can’t collaborate on movie lists even though it’s been a highly desired feature.
Problem
Letterboxd has a unique role as a place for film discussion and reviews, but the platform lacks several social features which users desire. Many users like the list feature as a way to categorize and curate films with themes or to plan their next movie night. However, there’s currently no way for multiple people to contribute to the same list.
Workarounds are used, such as having people comment on a list to prompt the list owner to add a film to it. Sometimes there’s more manual list making through text or group chat which gets difficult to keep track of.
Solution
Integrate a list collaboration feature using the better of the existing design patterns, allowing users to seamlessly contribute to film lists together. This also has the benefit of bringing new users to the platform, who may become paid users.
Research
Research Goal
To better understand the role of lists in movie curation and validate the hypothesis that there is a desire to collaborate on movie lists with other people.
Discovery
I researched what Letterboxd users were saying on Reddit and Twitter, especially about wanting a collaborative list feature. Because this conversation was so common among users, it confirmed my assumption that this is a desired feature and substantial enough for this project.
Letterboxd has an official feature request log that allows users to suggest and rank ideas, often with detailed replies from the company. “Collaborating on a list or watchlist” is ranked the #5 most desirable feature and marked “Under consideration” by Letterboxd.
Letterboxd has had interactions with users on Twitter about interest in collaborative lists. There are posts going back to 2014 with follow-up replies as recently as January 2023.
Competitor Analysis
I examined other popular film sites, and although they are very well-established, few had any features with a social focus like Letterboxd. Trakt is the only one with collaborative lists as a paid feature, so there is ample opportunity here to implement a highly desired feature that could bring newcomers and satisfy existing users.
I studied the list options each site offers - both watchlists and curated movie lists.
User interviews
I interviewed current Letterboxd users to learn more about how they use lists. I started with open-ended, generative questions about their movie watching habits and wrapped up by watching them use the Letterboxd platform to evaluate their current experience.
OVERVIEW
4 participants, aged 26-36
3 iOS app, 1 Android
INSIGHTS
All participants need to keep track of multiple lists of movies across formats and platforms
All participants receive movie recommendations from friends, family, and word of mouth
3 of 4 participants were not fully satisfied with lists how they are currently used
2 of 4 participants thought collaborative lists were already a feature on Letterboxd
TAKEAWAYS
Participants seemed to really struggle with how to best make use of the list feature when planning watching movies with family and friends. The current method of simply updating your own list with everyone’s suggestions is too clunky.
SURPRISES
There were unique suggestions for ways this could be used on the platform, such as a shared film watching challenge and having some well-known actors, directors, or critics create a themed list together. There was desire for the user to observe others’ curated shared lists in addition to their own, which is line with the platform’s focus on generating social interaction around film.
"I’ve wanted to add people to a Letterboxd list and I’m shocked it’s not a feature."
— Interview participant
User Persona
I gathered all research insights about who uses Letterboxd and created a persona.
Define
How might we improve the ways lists are used?
Problem statement
June struggles with making decisions
June needs to keep track of movie recommendations from various people
June is not fully satisfied with the way she uses lists and feels it can be improved
June needs to reduce the amount of time spent figuring out what to watch
How might we
How might we reduce decision making and cognitive load?
How might we streamline saving movie recommendations?
How might we improve the ways lists are used?
How might we make choosing what to watch faster?
Project goals
Ensuring alignment of business, user, and technical goals to ensure viability, desirability, and feasibility
✅ This is a highly desired feature
Users have been requesting this feature on Twitter and Reddit for over nine years, and it is ranked the fifth most desired feature on an official feature log forum. My user interviews also confirmed an interest.
✅ Collaborative lists would keep Letterboxd competitive
This would increase new users on Letterboxd, as current users can invite anyone to share a list but they must create an account. Once they start using Letterboxd, they may become a paid Pro member. This is only offered by one competitor as a paid feature so Letterboxd could stay competitive by offering this.
✅ It’s feasible to work within the existing platform
Collaborative lists can be implemented, especially if the existing design system and design patterns are re-used.
Flows
I created a Task Flow for the process of creating a list, to better see exactly which actions are taken.
I created a User Flow to start to see how collaborative list creation would work by either creating a new list or editing an existing one.
UI Design
Working with an imperfect product
Letterboxd has numerous visual inconsistencies, awkward design patterns, and dated-looking UI decisions. From the beginning, I tried a few approaches using existing patterns to try eliminating the ones that were the most problematic.
“[…] unintuitive and vague.”
— Interview participant
“It’s kind of clunky and looks old […] weird, tonal grey contrast and odd layout.”
— Interview participant
UI Kit
Here I documented the existing logo, font, and color choices, and app-wide components. Many different shades of grey were used which presented a challenge, so I represented the most common and the ones I used. There are many icons of various sizes, which I kept consistent in the design and had to replicate some of them in vector for my use.
Wireframes
I started with lo-fi wireframe sketches of how I thought a collaborative feature could be integrated into the existing app. Two main options appeared - creating a large icon to Share, or a link to Add People.
In the mid-fidelity wireframes, I went more granular to see exactly how each version might work.
Prototype
Testing
Usability Testing
After some iterations, I needed feedback from real app users to determine if I was heading in the right direction. I tested the two main layouts to see which integrated with the app better.
OVERVIEW
4 participants, ages 26-36, iOS app users
Moderated usability test
GOALS
Assess the impression and integration of shared lists
Understand expectations when sharing lists and inviting people
Observe any areas of hesitation, frustration, or pain points
TASKS
Create a new list and add someone as collaborator
Invite someone to a collaborative list
METRICS
All tasks were completed 100% of the time
Average ease of use rating was 4.85/5
“I wish Letterboxd had this!”
— Test participant
Outcome
I gathered all data points from testing into an affinity map where I grouped answers together and looked for patterns.
✅ SEAMLESS & CONSISTENT
Participants agreed this feature felt integrated with the existing UI, and they had the correct mental model before even seeing the design.
✅ LAYOUT CLARIFIED
There was a clear layout preference for the first version, which mimicks the way the platform handles movie entries.
❌ ROADBLOCK WITH FLOW
it became very clear that the modal-within-a-modal pattern was the biggest roadblock in the task’s flow. There was also some confusion with verbiage about what “shared” and “suggested” mean.
Iterations
Simplifying the People screen
A frustrating and common design pattern on the app presents the user with a blank screen instead of suggesting an action to be taken. To avoid this, I had the suggested section only appear if the list isn’t shared with anyone. Once a collaborator is added, they appear on the main screen and suggestions can be found when searching for someone to add.
Using a better design pattern
The modal-within-a-modal led users to feel stuck and unsure of what to do next, so I opted to use the existing pattern for creating a film entry.
BENEFITS
Feels more natural since users are already used to doing this when adding movies to a list
Allows for more granular settings like permissions, if desired
Optimizing the list view
Based on user feedback, I updated the list view to have more clear verbiage, arrangement, and to explore how multiples are treated.
Final Design
Create a collaborative list
Add a collaborator to a new or existing list so you can plan a movie night or have a shared watchlist.
Invite by email
Send an email invitation so anyone can be a collaborator. This makes it easy to start sharing films in one place, and introduces new users to Letterboxd.
Reflection
Adapting to an imperfect product
Letterboxd was a challenging product to work with due to its dated and suboptimal UI. Although I kept my work consistent, often the feedback I received was about the way it is and not my additions. Nonetheless, working with an imperfect product helped to refine my critical eye and made me a more discerning designer.
Get feedback early and often
The feedback about the modal-within-a-modal not working well was consistent and could have been iterated even sooner.
Next steps
With further time and resources, I would do the following:
Beta test with Pro users
Give real users the chance to try this feature and gather feedback and insights to continue iterating.Consider openly collaborative lists
In line with their goal to promote public discourse, the next step could be examining how to implement lists that anyone can contribute to, such as on GoodReads. There’s an opportunity for shared challenge lists and more.Perform a heuristic evaluation
The platform would benefit from a full UX audit and heuristic evaluation, as I encountered numerous issues and inconsistencies with user flows, verbiage, visual elements, and more.