The Client
Kaki Kaiwa is a language learning app targeted to Japanese market. The center point of the app is a mascot Kaki, AI-powered English tutor that provides you with an unique learning experience. No more memorizing, just talking!
The Challenge
Designing a working MVP prototype, redesign logo, mascot and all the branding. Focus was to create an app that provides the user with an entertaining learning experience, which feels just like having a conversation with a friend.
Outcome
Fully working clickable prototype, ready for an app development.
Role
User research, prototyping, UI design, branding.
Clickable prototype:
https://xd.adobe.com/view/4d9a283c-6c12-41ae-58b6-6e560743affa-3bab/?fullscreen
Research
During the research phase, I sought to understand the current state of the English learning landscape. My intention for the research phase was to uncover the needs and frustrations of the target user to better understand how to appeal to them. I tried to understand obstacles user faces when if comes to study language and sustaining regular use of a learning app. As a non-Japanese I also researched motivations and preferences of Japanese users.
Research Goals:
1. Define people's common behaviors when learning language
2. Define people's common behaviors when learning language using an app
3. Determine what people like and dislike about both learning experiences.
Secondary Research
First, I conducted secondary research to understand user demographics, developments, opportunities, and challenges in the language learning industry. I collected information from articles and case studies to identify industry standards and user expectations.
Clustering interview responses allowed me to discover trends and patterns from which insights could be constructed. Through this process I identified several user needs:
- Fun: Users want to enjoy the learning process. App should contain game-like features rather than grading. Users prefer learning by talking rather than learning grammar and rules.
- Fast: Japanese work longer hours than workers in western countries. They prefer more shorter lessons/checkpoints, so the structure of the app should allow users to be able to use it during short breaks in work, in metro etc...
- Smart: App has to have intelligent phrase suggestion and preparation for the future AI implementation. Users were really interested in "chat" feature, that allows users conversation with Kaki similar to talking to Siri.
Personas Creation
Next, I used all of the qualitative data i gathered during the research process to create my user personas. Our main target are Japanese Uni students and working class, known in Japan as "office ladies" and "salary men. I created multiple personas for each group to be able to understand different needs within our target groups.
Define & Ideate
Now that I had empathized with target users and identified their needs, I needed to define the solution. I reflected on the business goals, user goals, and technical considerations to find a happy medium for all stakeholders. Once I had identified common goals, I could decide what product features were necessary for the prototype.
Technical Goals
1. Quick app flow that allows completing a lesson within one minute.
2. User interface that looks clean and intuitive.
3. Create a spot for a mascot Kaki and an interaction with him.
4. Create a system of premium features that encourages users to subscribe, but in the same time it doesn't discourage free users.
Initial Flow Idea
According to provided client's brief, original app flow idea was quite linear. The flow consisted of:
1. Suggested lesson
2. Video of a English phrase and explanation of use
3. Examination the user by audio input
4. Rating
4.1. (optional) Possibility to hear the phrase in various accents
Simplified flow is described in the diagram below.
Redesigning App Flow
Looking into the technical goals, this structure wasn't solving originally defined challenges. Such a linear structure will result in a boredom of the user and engagement will decline over the time (based on user research of our competitors). From another standpoint, we are designing a MVP which will be most certainly updated with a new features in the future. Proposed structure would make it quite difficult to add features in the future especially when it comes to AI-tutor experience, which is the USP of the app.
Chat UI
After some research we came up with a flow revolving around chat interface, similar to Whats App or Messenger. This solves problem of interacting with a mascot quite well. because the whole flow will be coming back and forth to a chat with Kaki. Also structure will be less predictable and flexible when it comes to adding new features in the future. Simplified flow is shown in the diagram below.
Design
Design part of the project included redesigning logo, creating branding, designing app UI and creating a spectrum of emojis based on mascot Kaki. Client insisted on branding color to be "kaki"-orange, based on the fruit's color and specified few apps that design should be similar to. Design mood-board is shown below.
Logo Design
I began sketching out ideas for a logo, aided by the inspiration I found on Pinterest. I wanted to play with ideas based on Japanese typography with in infusion of a modern simple design. Challenge of the logo, was finding and adjusting two fonts, one for a Latin letters and the other one for Japanese symbols. I settled on a comic book style font, reminiscing traditional Japanese manga comic books. As I explored different variations, after the user feedback, I came up with a few versions of the logo including the mascot Kaki.
I have to say, that in the end the final version, wasn't necessarily my favorite, but since I am not the part of the target audience, I considered the user feedback from Japanese users more important.
Style Guide
The logos were added to the Style Guide and UI Kit, a comprehensive collection of all site components and UI patterns. Both documents are constantly changing as the apewp and brand grow.
Results
App is currently being in the development. Launch date is scheduled for Q2 2020.
