A mobile app to assist English learners in enhancing their speaking abilities through personalized feedback

A mobile app to assist English learners in enhancing their speaking abilities through personalized feedback

Time

5 months

Context

Solo capstone project for Springboard’s UX/UI Design bootcamp

Role

UX/UI Designer, mentored by Google designer

Tools

Figma, Miro, Marvel, Google Workspace

OVERVIEW

Fluent

As a teacher in Korea, I developed a public speaking program that was implemented at 3 campuses. Its success led me to pursue UX/UI Design, where I could make a bigger impact through tech, thus inspiring Fluent.

Problem

Despite spending billions (USD) annually on English education, South Koreans often struggle with Foreign Language Anxiety (FLA) when speaking, commonly referred to as “English nausea” when translated from Korean.

Solution

Fluent is a native mobile app that instantly connects English language learners with conversation partners from diverse, global regions to improve speaking confidence through customized feedback.

image

DISCOVER

FLA isn’t limited to South Koreans

I approached the problem space by conducting desktop research on English nausea to better understand the problem’s landscape, and to determine if it was worth addressing. I discovered 4 key insights.

image

Diving deeper into the root causes of FLA

Based on my findings, I expanded my research to include all English language learners regardless of national origin. I received 134 surveys from various Facebook groups and chose 5 people to interview. To provide depth, 3 were target users. For breadth, 2 were non-target (i.e. advanced in English, learning a language other than English).

Direct quotes from target users

“I’m probably making mistakes that someone taught me, and I’m (as an English teacher) teaching that to others.” — English language learner from Mexico
image
image
“Practicing English is fun with foreigners. In Korea, we only focus on grammar exercises. There is no chance to speak.” English language learner from South Korea
“I want to practice with others to improve my pronunciation, but there are not many chances because of COVID-19.” — English language learner from Vietnam
image

DEFINE

Sharpening my understanding

I transcribed my interviews and saw patterns around users’ goals, motivations, and pain points that helped me get closer to a solution. After organizing my interview findings, I discovered six insights:

image

Archetypes

After creating empathy maps to understand the backgrounds of our target users, I created corresponding archetypes to represent their thoughts and behaviors, which served as guides for my design process.

image
image

How might we…

To facilitate the next phase of solution ideation, I crafted “how might we…” questions:

💭
create a safe space for learners to practice speaking?
💭
provide tailored feedback to improve speaking skills?
💭
build confidence in speaking over time?
💭
allow learners to practice speaking in real-time?

DESIGN

Course curriculum with AI conversation partner

Initially, I outlined a curriculum on listening and speaking for various levels, incorporating an AI conversation partner.

image

However, this idea didn’t provide enough speaking opportunities, especially with real people. Moreover, the curriculum would take more effort to develop for an MVP.

Prioritizing user stories

Going back to the drawing board, I prioritized user stories to better align with the target users’ needs and goals for the MVP.

image

Revised user stories

“As an English Learner, I want to…

👤
engage in a fun, casual 1:1 conversation in English.”
👤
choose the person I talk to so I can feel safe in a conversation.”
👤
receive digestible and consistent feedback on my speaking skills.”

A new direction

My new solution would allow learners to:

instantly match with others in real-time to practice English 1:1
give and receive customized feedback after each call
image

This solution required less effort (compared to a curriculum) but was more valuable (speaking opportunities). I asked 20 English learners from FB groups about their thoughts on this idea and 100% thought it was the most useful.

Visualizing the product

Next, I created 3 user flows to imagine how Fluent would look.

Flow 1: Sign up

Questions were tailored to provide the most helpful information to share for choosing practice partners.

Flow 2: Conversation experience

Users scan profiles and accept a speaking partner. I also added games for structure and engagement.

Flow 3: Feedback exchange

Since users wanted feedback, I added a feature for a quick rating and another for detailed feedback.

image
image
image

Guerilla testing

I created a paper prototype to test my ideas with 5 target users for early feedback and discovered 3 major issues:

1. Concern over safety

Since users expressed safety concerns, I created a pop-up screen outlining community guidelines users must accept before participating.

2. Limited preferences

Since users wanted more preferences for choosing language exchange partners, I extended the options to region and user type (tutor, learner).

3. Lack of motivation

Since some users said they’d feel too lazy to write feedback, I created a coin system to incentivize users to leave feedback to make further calls.

Paper prototype through Marvel
Paper prototype through Marvel

Wireframes

The guerrilla test feedback helped me to create a seamless experience through wireframes.

image

HI-FIDELITY PROTOTYPE

The final design

Preferences selection

After agreeing to the community guidelines, users would choose their preferences and browse potential partners to begin their English language practice.

image
image

Conversation experience

To enhance engagement, I incorporated several conversation-starter activities within the call to provide a degree of structure.

Feedback exchange

Some users expressed concerns about forgetting to leave feedback, so I implemented a coin system: users earn more calling coins by providing feedback.

image

TEST

Usability testing

After conducting two rounds of remote moderated usability tests with 10 target users, I fixed 2 of the most critical problems to ensure Fluent worked as intended.

1. Improved home screen navigation

image

2. Simplified call preferences form

image

Retrospective

Users lead the way

Creating a new product from scratch can feel overwhelming and uncertain at times. However, I found comfort and confidence in my design decisions when they were guided and supported by prior user research.

Impact at scale

It was fascinating to discover that a problem I believed to be unique to Korea was experienced on a global scale. Witnessing how tech, combined with human insights, can impact millions of lives was truly inspiring.

Next steps

Given more time, I’d explore ways to expand Fluent, including premium features for chatting with native speakers, incentivizing more natives to join, and introducing an optional curriculum for added structure.