Duolingo
Goal: To add a new feature to the app to language learners’ experiences on Duolingo so they could have a more intuitive, interactive, and immersive experience
About the project
I worked with a team to add a new feature to the language-learning app, Duolingo, based on research we gathered from real second-language users. This project was a group project through the General Assembly User Experience Design Immersive Boot Camp.
Our goal for the feature, based on our research, was to help users have a more interactive and intuitive experience while they were trying to learn a new language. We did not work with Duolingo on this project.
Team Members:
Beth McNair
Rachael Sekamwa
Bradford Schiebel
Time Frame:
3 weeks
Tools:
Figma
FigJam
Maze
Canva
Procreate
Duolingo
What were my roles in this project?
For this project my roles were the following:
User Interviews
Comparative Analysis Reseach
Low-Fidelity Designer
Low-Fidelity Prototyper
Mid-Fidelity Prototyper
Conducted Usability Testing at all stages of the project (Low-, Mid-, and High-Fidelity)
Initial Research
To get a better idea of what our group was looking at, our team had to complete secondary research. We completed a comparative analysis doing a SWOT model on Udemy, LinkedIn Learning, and Coursera against Duolingo; all of these companies have similar models to Duolingo, allowing users to learn relatively inexpensively. Additionally, we also did a competitive analysis using another SWOT model on Babbel, Memrise, and Rosetta Stone, all of which are language-learning softwares.
Competitive Analysis
Babbel
Memrise
Rosetta Stone
Comparative Analysis
Udemy
LinkedIn Learning
Coursera
Research Takeaways
Duolingo is the only app that has gamification features, like Duo the owl
Most other apps include listening and speaking lessons, but not Rosetta Stone
Duolingo is the most-used language-learning app
Coursera, LinkedIn Learning, and Udemy all have courses that cost money, while Duolingo is free
However, Duolingo’s database isn’t the largest, compared to its competitors
The four learning websites are inexpensive alternatives to 4-year degree programs, but don’t have the resources of an accredited college or university
Duolingo’s competitors do not have interactive lessons
Visual mediums are the primary method learners prefer to use when learning a second language
User Interview Takeaways
Problem Statement and HMW
Following the user interview takeaways and secondary research, our team realized that Duolingo would require an interactive feature that would help users put their new knowledge of a language into use. Unlike the Seven Hills Running Shop project when I had to broaden the scope of the project, our team had to shrink the scope. There were so many things our user interviews addressed, but since we had such a tight time constraint, we needed to narrow down the UX scope to address the key issue in our project, which was interactivity. Our problem statement addresses this dillemma. Based on our data, we decided on the following problem statement and “How might we” statement.
“Users are having a hard remembering what they’ve learned because they are unable to practice their new skills.”
— Problem Statement
“How might we help users retain information that they learned?”
— “How Might We” Statement
Persona
*
Persona *
Jasmine “JJ” Jones
Key Goal
To not lose motivation while learning
Key Need
To be able to learn in her free time
Key Frustration
Learning Style
Has issues when she isn’t able to practice what she is learning
Needs immersive environment, fun, access to native speakers, and incentivization without penalties
For our project, we had to analyze what journey Jasmine would take for her path through Duolingo. Knowing that she is a learner that needs interactive experiences and ways to practice her skills, she had a difficult time progressing through her lessons once she runs out of free content each day. Additionally, she knows that she somehow needs to interact with native speakers and be involved with other people as she learns. This is her retrospective journey before we decided to add our new feature to Duolingo.
Decide to Learn New Language
Journey Map
Complete First Lesson
Continue with Second Lesson
Lessons End
User Flow
Once JJ completed her original journey through the original Duolingo lessons, we took a step back to analyze the user’s journey through the current app. We created both a retrospective user flow and a proposed user flow with the new feature we were creating. The first user flow does not include our new feature, but the second user flow does.
Retrospective Journey Map
Proposed Journey Map with new Media Center Feature
Site Map
Following the completion of our revised user flow, our team needed to make some changes to the Duolingo “site” map.
Design Studio And Sketches
Following the changes we added to the user flow, our team created sketches based on our research. We hosted a “design studio” to decide on what feature we could focus on given our time constraints. We ultimately decided that the feature that would be the most cost-effective with the highest impact would be a Media Center. Our design process, based on the data we collected from our interviews and research, is included in a design matrix below.
The sketches helped us see what our new Media Center feature would end up looking like. We had to keep the desire for users to put their newly-learned skills into practice at the front of our minds, in addition to making the videos we would hope to add, had there been more time and resources. We added a tab at the bottom of the app screen to make the Media Center prominent as we continued with our project. The following are the sketches that turned out the best prior to creating our prototype,
Feature Prioritization
Sketches
Low-Fidelity Prototype
Following the creation of our site map, user flow, and sketches, we had a better idea of what our new Media Center feature was going to look like. In our low-fidelity prototype, we made sure to highlight the changes we were making for the navigation of the app. However, during our creation of the wireframes and the subsequent usability tests, we learned users wanted back buttons/arrows and a play button over the placeholders for the video screens. Additionally, they wanted to swipe through the reading options on the prototypes, in addition to underlined words with a translation to English.
Our team interviewed 4 Duolingo users, and we made the following discoveries:
On average, each user went over the allotted 3 minutes by 30 seconds
Each task on the prototype had a 75% success rate
Users had a 62.5% success rate without hesitations
Observations our users had:
Include navigation back from the reading page to help with freedom and control
Make the proficiency level of the user visible to cater for visibility of system status
The reading flow was not intuitive for either user. They tried to swipe up first and then swipe left like a kindle. None of them thought to click the screen until afterwards
The language that the user is reading/watching visible by putting the flag of that language was not visible as the users completed the reading section
Mid-fidelity Prototype
Following the usability testing we performed on our low-fidelity prototypes, we realized our designs weren’t as intuitive as they needed to be. Users wanted to return to the home screen to see their learning path before advancing in their lesson. That being said, our team decided to add a back button navigation to our “reading” and “watching” wireframes so our users could return to the game board if they decided to. Additionally, users were not able to see the underlined words properly translated into their native languages, which in this case, is English.
We interviewed 4 Duolingo users for this round of testing, giving us the following data:
Users completed all the tasks in the allotted time, which was 3 minutes
Users completed all tasks with a 95% success rate
No errors among users
5/6 overall hesitations
Observations our users had:
Some confusion remained over the icons for the primary navigation
Video feature didn’t feel straightforward with play and pause buttons not working.
The switch back-and-forth between the native and language being learned was somewhat jarring
This design felt more intuitive than our previous set of wireframes
Accessibility Concerns
Following our mid-fidelity prototype usability tests, we learned a lot! However, we also learned that Duolingo’s brand identity was not up to WCAG 2.1 accessibility standards. We conducted accessibility testing and discovered people with vision differences would have a difficult time seeing contrast between backgrounds and text.
We changed the primary Duolingo green to a darker shade so it was more accessible. However, we conducted colorblind testing and discovered those with achromatopsia and tritanopia will still struggle with vision differences, we made sure that our design would have suitable accomodations for people with this condition.
High-Fidelity Prototype
Once we had collected all of the necessary data for our design, we were able to finally develop our high-fidelity prototype! This one was a ton of fun. Using all of our research to create the final design, we were able to complete our final design 4 days ahead of schedule. Because of this extra time, we were able to perform usability testing and receive some crucial feedback on our design. Our stakeholders loved it!
Using that data, we made minor changes to the design to make it be as intuitive as we could. Additionally, we were able to create our own Duolingo character, Robin, to further enhance the experience of our users due to their desire for more interactive learning.
See the prototype below:
Observations our users had:
The lack of subscription levels was confusing for users
Users wanted more variety content to our media and video library
Users desired tutorials on how to use the new feature
Next Steps
Following our high-fidelity prototype testing, we realized there is always more work to be done. We learned a lot about Duolingo’s brand guide, which helped us to develop the Media Center feature in the app. We grew in our understanding of accessibility testing, standards, learned how to include business metrics; such as KPIs (Key Performance Indicators); into our research, and also realized there are more ways to make Duolingo an intuitive and easily accessible app. If we were to return to this project, we would take the following steps:
Build out subscription levels for our Duo Super and Duo SuperMax tiers
Adding more variety content to our media and video feature
Continue to test for accessibility
Add tutorials on how to use the new feature for new users
Closed captions on videos
Translations on readings for those with cognitive delays
Consider the following KPI’s:
Currently 19 million users
Goal >5% increase
Conversion rate:
Seek for 54% from free to paid subscription
Goal: >3% increase
Increase Duolingo retention rate by 5%
Seek a >15% improvement score in language learning