Duolingo       

Learn Through Interaction

Case Study

What is Duolingo?

Duolingo is a leading language-learning app founded in 2011 that teaches users through a functional and gamified approach to focus on real-life goals with short lessons for better retention. 

Project Scope

Duolingo is limited in providing users with a community to leverage conversational practice and tools needed to reach oral proficiency. As a team, We wanted to explore how we could make this possible.

Tools Used: Figma, Slack, Adobe Photoshop, Google Suite, Zoom

Team: Melody Li (Interaction Lead), Leslie Humphrey (Planning Lead)  

Timeline: 2 weeks 

Role: UX Research Lead

Goals

Understand how and why users may want to reach oral proficiency, provide a user-friendly method, and keep it simple.

Discoveries Through Research

We wanted to understand the usability of the current site. 

Through Heuristic Evaluation we observed

Let’s Compare

Competitors

We chose to analyze and compare two industry-leading competitors:

Rosetta Stone and Babbel, offer different approaches to their success.

  1. Rosetta Stone is backed by scientific research, versus a gamification approach, that teaches users to think in another language. It provides interactions and leading speech recognition technology to perfect your accent without translation. 

  2. Babble offers comprehensive courses to teach users in a natural way to speak their target language through live lessons and classes for real-life scenarios, versus random words and phrases, like Duolingo.

Comparators

  1. Comparing an app such as the Nike Training Club App, showed how an expert level of structured personalized training programs can obtain a user's achievements. All for free!

  2. Yahoo Weather App taught us how a single page can provide all the necessary components and precise information while immediately modifying and updating the information based on the user’s current location through GPS.

Target Audience

User Research                          

To gain valuable user insights, we began with screener surveys to study qualified participants for our user interviews. The goal of our interviews was to understand the hows and whys of learning or wanting to learn a new language along with their pain points to synthesize for our research. 

We conducted 12 remote interviews and gathered insights split amongst our team members. Each member synthesized their own findings through affinity mapping to discover different perspectives and trends. Then we came together as a team to discover deeper insights.  

Collaborative Mapping

Amongst our team, we discovered their challenges such as retention, lesson structure, and an overall understanding of their knowledge. Their methods for learning via textbook style, reasons for learning such as travel, and their goals. 

Key Findings

Time

Users preferred short and easy lessons since they found it difficult to find time consistently

“…I was busy with work so I didn’t set aside extra time and that stopped me…”

Knowledge

Users only learn the fundamentals of a language that is not useful for real-Life Conversations

“Sometimes it just feels like I’m memorizing phrases and not actually learning the language”

Interaction 

Users found it difficult to practice without any interactions

“When you don’t have someone to practice with, you don’t retain it…”

Define By Storytelling 

User Persona

Synthesizing our research and finding key pain points, allowed us to create a persona to understand our target audience with their needs and goals.

Sarah is a sophomore at the University of Minnesota who will be studying abroad in Barcelona next semester. She enjoys reading travel articles and visiting cultural museums and is excited to immerse herself in a new culture. She wants to learn and interact with Spanish-speaking natives to develop her conversational skills confidently.

Journey Mapping

Through Journey mapping, we can gauge the user’s retrospective experience through her emotions as she searches for a language-learning app by doing, thinking, and feeling through each of the 4 phases.

Search – credible recommendations - Hopeful 

Explore – browse and find topics that meet her needs – Excited 

Learns – Learns fundamentals and explores features – content

Realization – Wants to practice conversationally – disappointed

Opportunities

  • Building new relationships with fluent speakers

  • Network and meet new people who are also learning

  • Gain feedback from a community through interaction

Sarah’s Journey gave us insights to create the

Problem Statement

Sarah needs an interactive way to practice conversations in a new language with an expert because she wants to navigate a country confidently while getting to know the locals

Before entering our design phase, our team ideated by asking:

  1. HMW find an effective way to navigate the user toward interactions with fluent Spanish Speakers?

  2. HMW allow users to experience real-life conversations with others while learning a language?

  3. HMW provide users with a way to interact on their own time? 

SOLUTION

We created a Community and Message page to provide peer-to-peer interactions via video or messaging as well as an intuitive AI known as “DuoBot,” accessible 24/7 for quick responses.

Ideation to Design

Go With The Flow

Our team began a design studio to sketch out multiple iterations of user flows and wireframes to best navigate the user to our newly added features. Remember, Keep it simple. 

Since the current state of the app’s navigation, and onboarding process was clear and concise to users, we kept the existing interface and continued using Duolingo’s Style guide to avoid any confusion.

Wireframing.

Wireframes began to give life to our design by highlighting the key elements of our features.

Deliver with High-Fidelity

Figma

We begin by adding a CTA(Call to Action) above the current lesson plan on the home page to introduce our new feature with contrasting colors of yellow and blue as well as a new message icon int he navigation bar for familiarity

By introducing Duobot and our virtual chat feature, we direct our users to a community page where users can click on active(green dot) or inactive member profiles to view their  background, stats, skills and achievements or filter through language, and location, to decide which member to interact and become friends with. Or simply search within the search bar if the user is looking for someone specific. 

Each profile page offers the option to interact via video through a camera icon or chat via messaging through the same icon as on the home page. If messages are unclear, users have the ability to translate via audio or by text within each message. 

Within the Messages tab, the user may view new and old messages, and of course our AI, DuoBot which is accessible 24/7.

For Privacy and security purposes, the user may block or report both new and existing members for their own safety.

Check out the Prototype!

All feedback welcome!

Usability Test and Report

We conducted 7 remote moderated tests to understand what the participants experienced by prompting them to speak aloud as to why they made their decisions, how they felt, and what stood out to them so we may better understand our design. We gave them the task to find a Spanish speaker and to interact via video. 

Initial Results

  • Within our first 3 tests, participants were already confused by our initial “Play Icon” to access our video chat feature. All 3 participants chose the play icon out of curiosity as their final choice, not by intuition. These participants expected the virtual chat to be accessed within the messaging chat feature.

  • The community page also did not indicate any specification of their language or location which also really confused users. These two features in itself resulted in 8-15 steps which all three participants succeeded in completing the task.

Then…

To Keep it simple, we conducted a second round of usability testing with 4 participants to see how we may shorten the amount of steps and to provide clarity for users by changing the following: 

  1. Redesign the “Play” icon to a camera for familiarity and also add it within the message box for access without exiting the chat to decrease the amount of steps. 

  2. Specify the language of the community page

A Huge win!

We decreased the number of steps participants took with such simple changes from 8-15 steps to 3-4 steps to complete the task!

Reflection

Design

  1. This project helped us understand how effective designing with familiarity and specificity can be done through very simple, yet complex changes with great results.

  2. Understood users’ emotional thoughts through tests to enhance their experience

As a Team

  1. Our team did a great job of communicating with one another and taking over each other’s roles when needed. Each member demonstrated their strengths and weaknesses which we were able to learn from. We were able to organize our timing and influence without authority. Our efforts and goal were clear from the beginning. I could not have asked for a better team! Thank you all!

As the Research Lead

  1. Overall, it was not a simple task. There were many obstacles. Conducting research before any design has always been a bit of a struggle personally but this team has shown me many different approaches which I have learned from. Mainly conducting the usability tests and key findings to re-iterate our designs have been the most exciting as well as the interviews. I’ve noticed I have gained a lot of love for the research aspect of UX, no matter the difficulty.

Next Steps

  1. Our findings helped us realize how users became confused when landing on the community page. 

  2. The users were unsure of what the profile pictures actually displayed, whether it was to begin a conversation immediately or if they actually spoke Spanish. Which leads users to hesitate to click on their profile. 

  3. Next steps would be to explore this problem further by revising the community page and conducting A/B testing to compare the original prototype to the new iteration.