Todaii Japanese: Boosting user stickiness via integrated learning
Descriptions
Todaii Japanese is a news app designed to help users master the language through reading.
Integrates supporting features like JLPT exam, dictionary, lesson for a better study experience.
Problems
Continuously adding tools to transform passive reading into an active "Read-to-Learn" experience.
Too many features integrated into a single screen increased cognitive load.
The actual usage rate of new features on the article detail screen remains significantly low.
After analyzing UX metrics via GA4, I found that both user stickiness & engagement faced a consistent decline over the five-month period from July 2024 to November 2024.
-17.1%
User stickiness (DAU/MAU) confirms fewer users are returning to the app daily
+12.4%
Bounce rate increase on news detail screen shows a rise in immediate exits
-0.21%
News detail engagement rate reflects a decline in feature usage during reading
To be more specific...
0.4%
Adoption rate for AI Analysis shows positive growth but remains far below the 5% target
-13.8%
Translation usage shows a decline in the app's most essential feature in news reading screen
1. Primary goal
Improve user stickiness & engagement by optimizing the core news-reading experience.
2. What to do
Identify pain points on the current interface that lead to higher bounce rates & declining feature usage.
Propose improvements to help users focus on reading & easily access core tools without distraction.
Since news reading is the core experience of Todaii Japanese, we adopted the User-Centered Design approach to ensure every optimization is rooted in proven behaviors. Rather than relying on assumptions, this process allows us to precisely align design changes with actual user data.
Review mining
Users complained about a cluttered & suffocating reading screen. Many even requested features that already existed—like Sentence Analysis—simply because they were buried under excessive icons and fixed bars.
Previous Surveys & Interviews
Synthesis of previous data confirms that adding learning-support features was a reasonable direction, directly addressing the real needs of our users. However, engagement rate of these learning features are low. Why is that?
80%
Users learn Japanese for entertainment & hobbies, followed by career growth (60%)
Heuristic evaluation
A usability audit where I evaluated the UI against industry-standard principles (Heuristics). By doing this, I could pinpoint UX bottlenecks that were causing user frustration on the reading screen.
Feature overload
Excessive buttons clutter the screen, distracting users from their primary reading goal.
Redundant information
Grammar and vocab lists are duplicated in the header & after the article.
Ambiguous icons
Abstract features are hard to represent with icons, making users miss existing tools.
Confusing buttons
Similar translation buttons placed side-by-side creates confusion for users.
Miller's Law: 7 ± 2 “Cognitive Load”
Source:
Based on past interviews and surveys, I identified two core personas: The Casual Reader, who needs a clean interface for quick news, and The Study-oriented Reader, who requires deep access to study tools for language mastery.
Skims through headlines and picks 1-2 interesting articles during lunch breaks
Uses quick translation for difficult words but rarely saves them
Fast and natural translations to understand the news quickly
Simple, clean interface that doesn't distract from the reading experience
Why are there so many buttons blocking my view?
The screen is so messy and crowded. Reading can’t be more mentally exhausting!
Constantly scrolls to the bottom to check and save words or grammar points
Spends a long time on one article to analyze difficult structures
A way to quickly jump from reading a sentence to practicing it
Save and memorize new words I encounter in each articles
I can’t find the grammar analysis feature they advertise
Why would I care about commenting on the news?
I used UJM to analyze the news-reading experience of each persona. Casual readers are distracted by too many study tools, while study-oriented group struggles to find & collect new knowledge due to a cluttered, unclear interface. The long, cluttered layout makes both groups feel exhausted, causing them to quit the app early.
Based on user behaviors & needs found in previous research, these solutions directly address actual struggles to create a simpler, more natural experience.
As is: Todaii wants to serve readers & learners simultaneously, but ends up failing both
As is: 12 features in one screen increase cognitive load & breaks the reading flow
Chunking Content: Progressive Disclosure
Source:
A research method used to organize information effectively. I collaborated with PM & Japanese content specialist to categorize 12 features based on user intent, aiming to simplify the interface into 2 clear paths: Read & Study.
Read
Study
I designed a structured 3-step learning path: Vocabulary, Grammar, and Sentences. By combining personalized selection with interactive practice, this flow ensures an active & tailored learning experience.
Showcasing side-by-side UI enhancements to highlight significant UX upgrades from the previous version.
Read
Study
#1: First-time entry
Shows the vocab list to start the first lesson instantly
#2: Returning Access
Displays the roadmap that allows users to resume lessons
A few other screens
A quick look at various screens within the new study flow.
As this feature is still currently in development, we plan to evaluate its success post-launch using a combination of quantitative & qualitative data to ensure both efficiency & user satisfaction.
Adoption rate (Goal: 15%)
Evaluate the appeal of the study mode entry point and its effectiveness in converting readers into active learners
Time on screen (Goal: +25%)
To measure how the integrated reading and learning flow drives deeper engagement
User stickiness (Goal: +20%)
To test if the bite-sized roadmap successfully builds a daily study habit
User reviews
To understand how users feel about the new study flow and if they find it truly helpful
Survey
Actively reach more users and collect a large volume of feedback on how the new feature meets their needs
Screen recordings
Observe real behavior & see if users struggle with the new solution, allowing us to identify & fix issues early
A quick look at the real struggles of balancing what the business wants versus what actually works for the user, & the lessons I picked up along the way.
