Multi-User Translation Solution : Case Study
Google Translate
COMPANY
Google Translate
ROLE
UX Designer
Industry
Machine Translation
YEAR
2024
Project Description
Google Translate helps people communicate across language barriers, but sharing a single device between speakers creates awkward interactions. To solve this problem, I designed and prototyped two new features: Duo Mode and Friend Mode. These innovations improve communication efficiency by allowing each person to use their own device, facilitate meaningful connections between speakers of different languages, and expand Google Translates' potential user base.
Problem
The reliance on a single mobile device for language translation among multiple individuals (Friends or Strangers) poses significant limitations, hindering effective communication and collaboration across language barriers.
Objectives
To design easier communication modes (Duo Mode and Friends Mode), within Google Translate, enhancing communication efficiency.
Role
User Experience (UX) and User Interface (UI) Design
Timeline
3 months
Team:
Solo project, supported by secondary research and peer reviews
Design Process
The design process followed a clear, methodical approach. Beginning with research and concept development, then progressed through detailed design phases, and concluded with testing to validate the solutions.

Research



Journey Map and User Flow
To further understand users experience while interacting with the existing product, a user journey map was created to visually illustrate user activities, feelings, pain points, needs and wants.

Research Insights Summary
Translation effectiveness increases when the device is closer to the speaker.
Difficulty using UI when sharing devices (text legibility, light/dark mode, accessibility).
Difficulty using translate in offline mode. (Flight mode and No connection area).
Tools and Methodology
Tools : Figma, Notion, Python, Octoparse.
Methods : Desk Research, Focus group and Diary Study.
Define
Problem Statement Definition
Tourists, international students, and business individuals face challenges in effectively communicating with strangers, friends, or colleagues who speak a different language, particularly when seeking assistance, making inquiries, or engaging socially in unfamiliar environments. This difficulty arises because conversations require passing a single device back and forth, disrupting the flow of communication and causing inconvenience during interactions.
Scope
The design scope for the project aims to create a solution that facilitates seamless real-time communication between individuals speaking different languages, eliminating the need to pass a single device back and forth during conversations.
Solution
Based on the problem statement and research findings, the next step was identifying key personas, developing user journey maps, establishing user flows, generating ideas, and creating high-fidelity prototypes that demonstrated potential solutions to address the identified issues.

User Flow
The first step of creating a solution was creating a user flow to visualize the idea concept, showing how the user will interact with the solution. After multiple iterations the finalized flow is shown below.

Wireframing
The Wireframe was designed to build the structure and architecture of the information and only then go to visual and interactions.

User Interface
The final design includes two communication modes:
Mono Mode (Current): A single device handles translation for both speakers. Though straightforward, it requires passing the device back and forth, creating an awkward and inefficient conversation flow.
Duo Mode: Both participants use their own devices with a temporary connection between them. This streamlines conversation, maintains privacy with no saved chat history, and doesn't require sharing personal information. Perfect for short interactions with strangers in different languages.
Friend Mode: Designed for budding friendships across language barriers. Each person uses their own device while maintaining a more persistent connection. This mode combines efficiency with relationship-building features, helping people develop meaningful connections despite speaking different languages.Read more
Results and Impact
It was challenging testing the translation aspect of the design as a prototype. A usability test was conducted measuring error rate, completion, and learnability. The total error rate was higher than expected at 73%, completion rate was 51%, while learnability scored 87%. Further research into the high error rate revealed that issues stemmed primarily from navigation and user flow problems.
These test results highlighted the need for design iterations to improve the overall user experience of the solution.