A reel of feature suggestions for filmgoers at Letterboxd.

For love of film: I wanted to take on a suite of usability proposals to Letterboxd on the web, my favorite film site. I frequently watch movies while working and do my browsing, debating, and scouring for opinions on them online—and I'd love to expand on how Letterboxd can do that for myself and others.

Roles User Research & UX Design Visit Say hello to me on Letterboxd


Exploring the film social network

In 2019, I found myself increasingly drawn towards film, and was lucky enough to be surrounded by people who I could share this love with. One of my new favorite sites is definitely Letterboxd, a film-focused social networking service. First introduced to it from film Twitter circles by seeing joke reviews highlighted on Oscar-nominated masterpieces, the sense of community it offered finally drew me onto the platform knowing it wouldn’t solely be traditional film buffs.

Letterboxd has been an amazing tool that has moved me to be more introspective about the film I consume, but it took a while for me to fully get into it, and the more that I use it — the more I’ve been thinking about ways to improve some of its features.

The following is a survey of some of the main frustrations on the web version of Letterboxd that I’ve collected from regular engagement with film communities, and my proposed redesigns and feature introductions.

Have other frustrations on Letterboxd's web version? Message me and let's look into it!


Improving the Movie Logging Flow

As Letterboxd’s creators get right: the first thing we want to do after we see a film is talk about it. On the film social network, why is it difficult and unintuitive to access the film logging feature when many of its social features and engagement notifications rely on it?

At the moment, the only flow on the website to log recently watched movies is to search for the movie title, then be delivered to a results page, click on a “Review or log…” button, then click on a checklist to specify the date watched—then confirming to successfully log a movie.

Immediately after watching a movie, we want to talk about it.

There needs to be a more direct way to get there.

Issues—Movie Logging

After observing this problem, I talked casually with over a dozen users to obtain a deeper sense of this key action. (Thank you Film Twitter!) Here are the key insights I captured:

  • Prominence of log action — It’s strange that the most prominent actions are the least important ones. The most confusing part about the action prompt is that “Watched” is used as a backlog tool without giving any prompt for when the film was watched. I address the iconography in itself later, but here I wanted to zero in on how the “Review or log…” action could be highlighted further.
  • Homepage layout — One of my initial assumptions that was validated through these talks was that the homepage was heavily underutilized. There are no existing prompts for users to take action on their own logs, such as shifting through their watch list or logging a new movie. The sections on the homepage are all on outside engagements with lots of repetitions (e.g. the popular reviews section tends to be filled with reviews of the latest movies; and majority of Letterboxd users, users that don’t maintain an active “friends” list, will have the first section stagnant).
  • Extended and unintuitive user journey — All search actions on Letterboxd take the user to a results page, and the search action is only accessible from the navigation prompt. (Clicking “Films” brings the user to a list of films in they’ve watch in reverse chronological order.)
  • Missing features — On the web version, the log box has areas to input less frequently used features (tags), but excludes lists: one of the prominent user-generated discovery features on the website.


An active homepage

As a starting point, I wanted to make better use of the Letterboxd home screen.

My proposed flow from the homepage involves a direct interaction from searching for a movie title (or director/actor) to opening up the review/log prompt to allow users to easily add movies.

Current Home

With Watch Prompt

On the homepage, I replaced the previous greeting/friend message with a new, large section prompting the user to write down what they’ve “seen” lately. Copy is very similar to the original page with the added “Seen anything lately?”

Watchlist cues—Visually, there’s a captivating header image (similar to the ones executed in film pages) in the background selected randomly from the user’s watchlist. User watchlists are one of the main sections on user profiles, and drawing attention to these backlogs would be an interesting way to nudge users to get through the list — and return to the website to mark it as done, of course.

Direct search—With the homepage search bar, I took on a prominent user assumption where they would have liked to be driven to the page immediately; as searches are frequently direct. I added an area for filters that replaces the search icon on type that could display primary filters: Film, Director, Genre, and the like. The search would provide a dropdown of the top suggestions (if similarly named), or go directly to the film page.
This new interaction would bring the user to the film page with the review/log prompt available, and the current date selected as the watch date. Upon saving the entry to the user’s log, the user can then view the rest of the film’s details, top reviews, and interact with what other members of the community are saying about it.

Persisting Challenges

Inclusion of search results page—The Letterboxd search results page tracks film titles, directors, cast & crew, and review contents. In certain cases (reused movie titles, similarly-titled sequels), I worry about the interaction bringing the user directly to the film page — but from my studies, this is rarely the case. Plus, search queries are generally filled with film reviews (always showcasing 250+ results, the maximum number displayed) when no relevant titles are available; this is a huge waste of clicks in most instances. I would love to see data on search queries and what results are most selected from the homepage to refine this thinking process further.


Better Discovery & Engagements on Film Logging

After designing a revised flow for logging and presenting it to friends and peers in online film communities, I also set my eye on modifying the interactions available on the logging prompt itself.

Issues—Movie Logging

Limited interactions—The main actions available are writing reviews, specifying date watched and if the film is a rewatch, tagging, liking, and starring. Half of these interactions (tagging, liking) aren’t pushed into action for other users (there are no prompts to see that a friend recently “liked” a movie, tags serve more as keywords).

"Diary entry" is unclear—Currently, the review and diary prompts are mixed together, leading to a flow with mixed pointers. There needs to be more direction and clarity on the Letterboxd logging process.


A revised log prompt

Rating—Rating a movie is one of the easiest ways for us to share an opinion about it. It’s what the user takes action on first, too. I moved the rating upwards right next to the movie title/date watched to follow more into the regular reading pattern and establish prominence.

Tags—Next, I revised the tags area. I thought saving space here would be safer since from research, most of the users I talk to don't tag movies (or if they do, add jokes as tags)—and if they are, they don't usually require an amount that demands as much real estate as it presently has.

List—With the new space, I added in the list dropdown where users can be prompted to start a new list or select an existing one. Lists are one of the main means where users tell stories, share insights, and group together recommendations for the masses: putting it on this log section is important to get more people engaged on building them.

Favorite—Finally, along with the “Like” functionality, I added a favorite star. This would add the film (replacing the oldest selected one) to the user’s favorites on their profile.

From sharing these hi-fidelity mockups of the new prompt, I validated these assumptions and found that users started wondering why some of the existing actions were so scarce.

One of the main reasons why people go to Letterboxd is because of its interface and community/discovery tools. By improving the logging process, I’d love to see how this gets users to continuously return to Letterboxd as a go-to website for sharing their own film discoveries and taking part in the network.


Reviews: Discussion for all

There’s nothing more critical to the film experience than being able to articulate and share our thoughts. It’s why movies like Uncut Gems with polarizing audience and critic responses show you that ratings alone won't cut it.

The current Most Popular Reviews for Uncut Gems (2019)

At present there’s a clear divide between two large user groups of Letterboxd: casual filmgoers (e.g. film twitter and Timothée Chalamet fans galore) that usually share humorous, pop culture-inflected thoughts, and film reviewers intent on sharing more traditional criticisms and breakdowns of pieces. In essence, the primary personas I’m thinking of are the following:

  • Younger film enthusiasts who offer casual, humorous, and witty engagements about movies and pop culture — with a lot of knowledge on trending actors; definitely engaged in the social aspects of Letterboxd
  • Classic film reviewers and critics — these users head to Letterboxd over IMBD to explore the social aspects of film. They’re engaged in analyzing themes and longer thoughts on films.

Letterboxd’s approach to reviews on desktops present this way: on movie pages below summaries and details, friend reviews are listed, then the top three most popular reviews (from “likes”) and the top three most recent ones. Reviews along with lists are the main spaces for users to interact with one another on the website. I’m interested in exploring how we can potentially filter through comedic reviews and more informational or critical ones to cater to what users feel like discussing.
After all, post-Parasite (2019) I might want to decompress and just joke about the magnificence of another Bong (Joon-ho) hit, or work through a deeper criticism on all its commentaries. Better filtering through these conversations will help all types of users in the platform to engage in different types of commentary.

Exploring existing models

The first existing model that I wanted to lift from is the Steam Store; games are just as meticulously looked over through reviews, and are equally a space for even AAA games to be laughed over.

The way Steam approached reviews would then be my main foundation for exploring improvement on this feature.

There are several buttons to indicate whether the review is helpful, and a “funny” option. The copy on this is a bit awkward, so I opted for three bttons instead: retaining the likes, laugh, and insightful — buttons that are similar to recent, intuitive trends on reaction on sites like Facebook and LinkedIn. I decided on these three buttons a bit arbitrarily; though they’re also the types of content filters used on Steam games and based on discussion with other users. Would take a note to do more research on categorization in this area before potentially moving forward for implementation. These buttons help users gauge and filter reviews themselves while addding onto the types of interactions with them.

Moving onto viewing these reviews, I envision adding category buttons at the popular review view to sift through the three most relevant reviews categorized as needed.
By default, the more reviews view continues to just show the most liked, popular reviews — which at the moment inhibits smaller users or more diverse sets of reviews to be seen; these limits in visibility mean that users don’t get to see content produced outside of the followers that they already follow, and make it harder for less established users to rise in popularity.

By adding new categories to reviews and making review filtering a priority on film pages, we hope to achieve the following:

  • Satisfying all types of users who are looking for different types of information from reviews: whether pure humor, critics’ insight, or more
  • Allowing more room for top reviews to shift; continuing interaction with the film, especially as time passes
  • Opening more avenues for users to engage with other reviewers on the platform, increasing follower interactions

Continuing Questions

What kind of film network does Letterboxd want to be?

  • As always, there’s going to be conflict in shifting to a new system of review interaction. How do we get users to use these new buttons as a means of reacting and engaging with posts? By design and with two new categories (insight and funny), we know that every single film prior to the system is going to have more rotation in reactions with these new tabs. How can the rollout be conducted in such a way that these new conversations spark the most user engagement?
  • What type of reviews do users on Letterboxd love the most? The retention of the general “like” button is going to help dictate and give insights to this. Is Letterboxd as a platform encouraging more casual, witty dialogue around film — or are people looking for the same thoughts as they do in critics? These would be interesting numbers to use in surveying the types of social dynamics happening around the platform, bettering not only our understanding of why this polarization in content exists in the first but how to best expand on it.

Film is community.

Visit Letterboxd