Improving User Engagement on Duolingo through Refined Social System
Purpose
Kleiner Perkins Design Challenge
Role
User Experience Design, User Research, Product Strategy
Timeline
7 days in January 2024
Context
Whether it is waking up to the sound of my suite-mate clearing his daily quest or hearing “have you done your Duolingo yet??” being constantly tossed around in my friend group, Duolingo makes a regular appearance in my life. Despite not using the app much myself, I am interested in taking a deep dive into its UX and finding ways to make it better.
Methodology
My trainings at Tencent and Amazon helped me adopt a design methodology characterized by thorough research and vision-orientedness. For this Design Challenge, my process is as follows:
Competitor Analysis
A deep dive into the language learning app market helped me divide the apps into three main types:
Duolingo belongs to the visual-focused category, which requires users to interact heavily with their screens instead of listening to a podcast or having a conversation with another person. I pressed deeper into the most popular visual-focused apps in the market and summarized the information in a competitor matrix.
Compared to other successful apps, Duolingo does not focus on providing the most effective lessons. Reviews on
YouTube and Forums like
Reddit agree that Duolingo is most helpful as “beginner material.” Instead, the focus is on
App Engagement. Duolingo VP of design
Ryan Sims says, “The secret to Duolingo is that we’re not an education company. We’re a fun and motivation company. Fun is the most important part of the work we do.”
So, what makes Duolingo so popular, though there are many other apps that check equal boxes feature-wise? One thing that the competitor matrix above does NOT capture is that Duolingo
chooses their focus and commits to executing it extremely well. While many of the competitor apps above incorporate gamification, none other than Duolingo has been described as “addictive” by reviews on the internet. By taking its strengths to the extreme, Duolingo stands out from its peers and achieves the immense popularity it has today.
For the redesign, it makes sense to conform to Duolingo’s proven product strategy and further augment its position as the #1 most fun and engaging language learning app. This realization led me to craft the guiding principle for my work:
User Research
I conducted 5 interviews with the following Duolingo users—all college students at USC—with a specific emphasis on how much fun they have with the program and their levels of engagement.
The first thing I noticed is that 4/5 users only stay on the app for less than 5 minutes on a typical day. “The biggest incentive is to keep the streak alive,” explains Victor, “so the motivation drops off once that is accomplished.” Coi, a user with a 551-day streak, comments that “after doing the same thing for years on end, [I wish to] switch it up to break down the monotony.”
When asked about what would make them spend more time on the app, people’s answers varied, but ultimately shared a unanimous consensus (5/5 users) on an improved social system that allows for more meaningful interactions with real-life friends. This indicates a strong potential for a revamped Social System to improve engagement.
Pressing deeper into people’s thoughts regarding Duolingo’s current Social System reveals some key insights:
- People find the current way of interacting with friends confusing. They complete quests with friends in one place, see their activities in another, and check their friends’ profiles in a third location. 2/5 users interviewed do not know how to locate all the friends-related content and miss out on relevant information.
- People want to engage with their friends beyond just comparing stats. It is fun to compare learning progress to friends, but 5/5 users interviewed would like to do more meaningful activities with their friends—whether it is competition, collaborative learning, or so on.
- People long to have “a sense of camaraderie” with their friends. 3/5 users interviewed report wanting to feel more camaraderie and togetherness with their friends. While users can see friends’ major milestones via the Feed section, they wish to know more intimately how their friends are doing on a daily basis.
Define Vision
The Competitor Analysis and User Research allowed me to find a viable path for the redesign that both stays loyal to the product strategy and accurately reflects users’ desires. Below, I consolidated my vision and defined some success criteria to work towards:
Product Audit
I walked through the ins and outs of the Duolingo Social System via a thorough product audit. Through the process, I hope to become familiar with more details of the app and clearly understand what stands in between the current version of Duolingo and the end results I envisioned above.
I achieved the following via the audit:
Identified major UX issues regarding friends:
- Cluttered Navigation Bar. The current Duolingo has 7 icons in the bottom navigation bar. This increases the complexity of the app and explains why my interviewees have issues finding friends-related content—there are simply just too many options available.
- Inconvenience. It currently takes 4 taps to view a specific friend’s profile. It is time-consuming and inconvenient for users who want to check how their friends are doing.
- Finding a specific friend from a long list without a search bar can be tedious and frustrating.
Discovered new interactions within Duolingo.
- Although users still lack fun ways to interact with one another, they can do solo Duolingo Timed Challenges (XP Ramp Up, Lightning Round, Match Madness). These timed challenges are a nice change of pace from the normal exercises that many users see as boring and repetitive.
Create New Design - Starting from Paper
In the earliest stage of design, it’s all about coming up with as many ideas as possible. So I used the fastest method—pen and paper—to draw up all I could think of. Beyond addressing the UX issues discovered, I aimed for my ideas to press further toward the three visions I established.
After coming back from a break, I began carefully evaluating the pros and cons for each idea under the following guiding principles:
- Appeal to users. Do users see themselves using the new design? I brought in three of my interviewees to help me judge this part.
- Disruption of current UX patterns. How much does the idea change the current layout on Duolingo? Sometimes, even if a different design is better, users may dislike it due to cognitive loads of abandoning something familiar and learning something new.
- Engineering Efficiency. How much additional engineering would this project require? How much existing resource can it reuse to reduce the cost of implementation?
After careful consideration, I decided to further pursue the following concepts:
Creating New Design - Solidifying through Low-Fi
My ideas began to take shape in low-fidelity prototypes. I tried to use as many of Duolingo’s existing UIs as possible to maintain stylistic uniformity and reduce engineering hassle. (I kept everything black and white in the low fidelity version to maximize efficiency.)
1. NAVBAR REDESIGN
Simplifying the 7-item bottom navigation bar is essential for making features more findable and users less confused. I divided Duolingo’s large set of functionalities into six top-level categories. “Social” serves as one of the top-level categories to make friends-related features more findable.
For those languages without a different writing system (which don’t have a "Alphabet Practice Tool" icon in the nav bar), "Feed" is the only icon that was removed from the navigation bar. To access Feed, users simply tap on where it used to be (now "Social," the rightmost icon). It is still accessible by one tap.
For users learning a language with a different writing system, they can find the "Alphabet Practice Tool" under "Exercise". To make the new location obvious, I used the same icon but enlarged. It is still accessible with two taps.
I also made it easier for users to find a specific friend by adding a search bar in the Friends List:
2. LIGHTNING DUALS
To introduce more novel ways to interact with friends, I modified the fan-favorite Lightning Round to be a competition between two players. As a feature to deepen engagement between friends, it lives under the “Social” icon, accessible via two taps.
Once users press on the “Dual with a Friend” button, they can choose to send a dual request to friends who study the same language on Duolingo and are currently online.
One thought I spent much time thinking about was how to make sure the questions in the Lightning Round are fair for both the user and their friend given that everyone’s progress is different. I came up with many different ideas:
To make the process as straightforward as possible, I was tempted to choose the first idea. However, the amount of steps required could result in additional hassle for the users. The solo version of Lightning Round can be started almost instantaneously and I want to preserve that smoothness in UX. I ended up choosing Idea 2 for its quickness and engineering efficiency—just create a normal Lightning Round for the user with less progress and give it to the other user as well.
Once the matchmaking is finished, users can complete the Lightning Dual exactly like how they would complete a solo Lightning Round. Users can compare the time taken to finish all the questions and XP is rewarded to the winner. There will be a leaderboard for friends with the most wins to further strengthen a feeling of community.
3. EASILY SEEING FRIENDS’ ACTIVITIES
Responding to one of my interviewees’ most consistent requests—“to know how my friends are doing more easily”—I designed the “community” page, which is under the “Social” icon and 2 clicks away from Home.
Seeing all the content on the page requires light scrolling. It is broken down into four main sections:
By consolidating segmented information hidden in each individual friend’s profile into a single page, “Community” aims to communicate one message loud and clear: the user is not alone and many friends are spending time on the app alongside them.
Creating New Design - Refining Design to Hi-Fi
After carefully reviewing my low-fidelity prototypes, I began heavily referencing the
Duolingo Brand Guidelines to create high-fidelity designs. I specifically paid attention to the following aspects:
- UX Writing. I discovered that the app avoids words like “Challenge” and “User,” leading me to change the tab for Lightning Dual to “Games” instead. I also did some careful copy-editing to ensure that my text matches Duo’s motivating and positive style.
- Visual Design. To match Duolingo’s existing aesthetic, I made sure that all my UI elements and icons have rounded edges only. When making the new "Social" icon, I paid close attention to the idea of “rhythm” and “simplicity” while refining my designs:
Now, without further ado, let me present to you my redesign:
Final User Testing
PROCESS
I fully animated the high-fidelity prototype and brought it to test with 5 Duolingo users (1 previous interviewee and 4 new users). To test for findability, I provided users with the following tasks and observed if they could accomplish them on their own:
- Find the tool that can help you practice the Japanese Writing System.
- Find the Feed feature.
- Find the game that you can play with your friends real-time.
- Find the place to view a summary of your friends’ activities today.
After completing the tasks, users described their experience, gave feedback and completed a survey.
RESULTS
1) 5/5 users completed the tasks quickly without interference from the moderator. As a result, if the Friends Section is included in the navigation bar, users would most likely be able to find all the friends-related content easily, achieving the Findability and Convenience goal in Vision #1.
2) Users think the Lightning Dual game is fun and see themselves using it in the future. On a scale of 5, users respond with an average of 4.8 when asked about both how fun they think the Lightning Dual feature is and how likely do they see themselves using it. This fulfills Vision #2—the goal of creating new and fun ways for friends to interact with each other.
3) Users report feeling updated about their friends’ activities and feeling a sense of camaraderie in the redesign. On a scale of 5, users report an average of 4.8 when asked about how updated they feel about their friends’ activities via the “Community” page. They report an average of 4.6 when asked about how much camaraderie they feel with their friends in the redesign. This achieves the goal outlined in Vision #3.
UPDATING THE DESIGN BASED ON USER FEEDBACK
While users liked most aspects of the design, they report minor confusions on the bar UI in the Lightning Dual results page. Since bars like these typically indicate incomplete progress in Duolingo, users are led to think that they have not fully completed the challenge. I addressed this confusion by removing the bar visualization and placing more emphasis on the remaining time.
Takeaways
Thank you for walking through my redesign journey with me! It has certainly been an enriching experience, and I am happy to produce something that meets my initial vision in this one-week sprint. Below are some key takeaways:
- Creating a successful product is not about checking as many boxes as I can. Given a company’s limited resources, it is impossible to build a product that does everything well. Sticking to doing one or two things extremely well within a well thought-out market positioning is enough to win a lot of users over.
- Including users in the design process helps a lot. It is important to not fall into an echo chamber when I am designing—constantly getting user feedback is one way to achieve that. I brought users in to contribute to the design from as early as the brainstorming stage while everything was still on paper. The users’ feedback saved me much time and energy.
Things I wish to do better with more time:
- More thorough user testing. Given the time limit for the exercise, I was only able to interview 5 users during both the initial research and usability testing. I ideally would have interviewed many more to obtain a more accurate representation of the overall user base.
- More time to refine the visual design. I could have taken more time to create more ideas for the “Social” icon and iterate more to help it further align with Duolingo’s visual principles.
- More time to consider game balance. While I am certainly proud of the Lightning Dual feature, I wish I could have more thoughtfully decided how much XP a user would gain after winning each round. Duolingo is very much similar to a video game—I don’t want to accidentally introduce something too “OP”!
Thank you so much again for checking out my redesign! I would love to receive any feedback or comments. Feel free to reach me by emailing guoqing@usc.edu!
© Portfolio by Justin Zhang, 2024