AI Story Time

Learning Chinese through Stories with Your AI Tutor.

My Roles:

Research | Product Design | Project Management

Shop_Accessories.png
AIS_Home.png

01

What is AI Story Time?

AI Story Time is a platform to help kids learn Chinese through stories. The target users are Pre-K to 5th Grade students who are learning Chinese as their foreign language. It has 1000 animated stories at 15 levels of language complexity. 

Each animated story is followed by a short Quiz and a Read Aloud practice where kids can get instant feedback from AI tutor (powered by speech recognition). Learners can earn points and badges across these activities, which contributes to their rankings on the leaderboard

AIS Icons-1video.png
AIS Icons-1quiz.png
AIS Icons-read aloud.png

                                               Animated Stories                             Quiz                                    Read Aloud

AIS Icons-1point.png

Points

AIS Icons-1leaderboard.png

Leaderboard

02

What was Wrong?

When the product was released in late 2018, we ran pilot programs with three school districts including 15 schools and 652 target users to test the market. However, the result was not ideal. After the free pilot subscription ended, only 6.9% of users subscribed to the product though they were provided a good discount.

 

Similarly, the subscriptions of AI Story Time in the retail market were not optimistic: 

Cache.png
Tooele.png
4.7.png

4.7%

16.png

16%

17.png

17%

                   Conversion                                             Cancellations                                               Deletion

           For Users Who saw the app on App Store                              For Users Who Purchased Subscription                                               App Deletion From Phone

* Data range: 12/1/2018-10/1/2019

03

What was users’ behavior?

* Data is demonstrated with the time frame of a week.

To explore the problems, I first analyzed the behavior of our users. The monthly active users accounted for 10% of our total users since the app was released. To better interpret our current user’s behavior, I generated an existing user profile and compared it to an ideal user we defined when designing the product:

 

 

The current user's profile has a big gap compared with an ideal user. From the data above, I found three problems:

 

 

Did AI Story Time fail to serve users' needs? Did our users encounter any obstacles when using our product? Were we wrong in defining our ideal users? With these questions in mind, I decided to interview our users.

Existing User

Ideal User

User Frequency

1.27 times                     3 times

Time Spent

3.38 min                        30 min

Story Watched

              1.2                                   5

Quiz

              0.3                                   3

Read Aloud

             0.14                                  3

Points Collected

 116 - 363                      930 - 4680

logo light.png
AIS Icons-1video.png
AIS Icons-read aloud.png

 AI Story Time was not being used frequently and the time users committed to for each use was comparatively short.

Users watched much less animation than we designed it to be.

The Read Aloud section was least accessed among all the 

three activities.

04

Why did they behave like this?

I conducted eight in-depth interviews with users including 4 teachers, 2 parents and 2 students from pilot schools, local schools, and retail channels in addition to the feedback from the product review page and our Customer Service team. Four teachers were chosen because they can not only tell us their usage feedback but also students’ feedback. In order to find out the patterns, I organized their major feedback in an affinity diagram:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

From this feedback, I identified the root causes of the three problems.

Affinity Diagram 4.png
Affinity Diagram 3.png
Affinity Diagram 2.png
Affinity Diagram 1.png
logo light.png
AIS Icons-1video.png
AIS Icons-read aloud.png

Problems

 AI Story Time was not being used frequently and the time users committed to for each use was comparatively short.

Users watched much less animation than we designed it to be.

The Read Aloud section was least accessed among all the 

three activities.

Causes

Most young learners lack interest in learning foreign languages unless motivated. However, the incentives (Leaderboard, badges) provided by the app was not effective enough.

Students had a hard time looking for stories that are age-appropriate and interest them.

 

Teachers found it difficult to search for stories that is associated with their curriculum.

Read Aloud was too long and overwhelming.

 

The autoplay feature of the English translation was annoying for users who did not need it.

05

Persona

After synthesizing our research finding and sorting out the behavior patterns of our target users, I segmented 3 user personas in order to inform our redesign and communicate with stakeholders and developers: 

Pain Points

1. He is not motivated to learn Chinese. It is just another task.

2. His parents do not understand any Chinese. He feels helpless sometimes.

Ben

Non-heritage Learner, 6

Learning Chinese is not fun!

Ben has been taking Chinese classes in school but the instructional time is very limited. Thus, his school wants to purchase some apps to assist students' learning at home.

Student2.png

I need to communicate with my family in Chinese.

Emma’s parents are first-generation immigrants to the United States from China and they want her to maintain the culture and speak fluent Chinese. Thus, Emma’s parents are looking for some apps to help improve her Chinese.

Pain Points

1. She finds speaking the hardest part in learning Chinese. 

2. Her school does not provide Chinese, so she has to learn Chinese in the afterschool program or sometimes by herself with support from parents.

Student1.png

Emma

Heritage Learner, 8

I’m looking for materials that can supplement my students’ different learning needs.

Tina is a second grade teacher who is teaching Chinese in elementary school. She is looking for some materials associated with the curriculum to supplement her teaching because the textbook cannot satisfy her students’ various needs.

Pain Points

1. Her students demonstrate various Chinese proficiencies. It is hard to satisfy different needs.

2. Her students make slow progress because the instructional time is limited. Thus, she hopes her kids can continue learning Chinese at home.

Teacher1.png

Grace

Chinese Teacher, 30

Before going right into the solution, I did research on our competitors and the most popular language-learning apps in the market and want to find out if their users are facing the same problems. How do they solve these problems? How can AI Story Time distinguish itself from other products? To answer these questions, I selected six representatives and examined them from 7 aspects: target user, content type, number of instructional units, leveling system, incentive, key features, and subscription.

brainpop.png

Key features

Subscription

06

Who are our competitors and what do they do?

Are our competitors facing the same problems? How do they solve these problems? How can we distinguish ourself from other products? To answer these questions, I selected six representatives and examined them from 7 aspects: target user, content type, number of instructional units, leveling system, incentive, key features, and subscription. Here are the three key insights:

Language Assessment

1. The majority of our research subjects assess users' language backgrounds during onboarding to match users with the appropriate content. The information requested including age, gender, grade level, and interests. For products with complex leveling systems, they assess users via a short quiz and/or include teachers to conduct the placement test and determine a level for users. 

 

 

2. Lacking interest in learning languages seems to be a universal problem. More than half of the research subjects have an incentive system. Their means to motivate learners include tickets and rewards system, avatar builder, room decoration, badges, and certificates. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3. Read Aloud is a unique feature of AI Story Time. Among the research subjects with speaking features, they only provide users a recording tool to mimic the audio and let users compare by themselves or submit their recording to teachers to get feedback. Thus, the Read Aloud will be the main selling point to distinguish us from others.

brainpop.png
83.png

83%

abc mouse.png
amazon freetime.png
epic.png
ichinesereader.png
Raz-Kids.png

Incentive System

Age

Gender

Grade Level

Interests

Quiz

Placement Test

Tickets and Rewards

Avatar Builder

Room Decoration

Badges

Certificates

amazon freetime.png
brainpop.png
67.png

67%

abc mouse.png
epic.png
Raz-Kids.png
ichinesereader.png

Speech Recognition

VS

Voice

Recording

07

Solution

After analyzing our users’ behavior, the root causes and our competitors, I finalized 6 features to resolve our user’s problems. 

logo light.png
AIS Icons-1video.png
AIS Icons-read aloud.png

Problems

 AI Story Time was not being used frequently and the time users committed to for each use was comparatively short.

Users watch much less animation than we designed it to be.

The Read Aloud section is least accessed among all the 

three activities.

Causes

Most young learners lack interest in learning foreign languages unless motivated. However, the incentives (Leaderboard, badges) provided by the app were not effective enough.

Students had a hard time looking for stories that are age-appropriate and interest them.

 

Teachers found it difficult to search for stories that are associated with their curriculum.

Read Aloud was too long and overwhelming.

 

The autoplay feature of the English translation was annoying for users who did not need it.

Solutions

Create a Game system in which users can earn coins to build their own avatar. 

 

Invite friends to read story together.

Personalize users content by giving them recommendations based on their age and interests.

 

Optimize the search page by providing theme-based search and search by voice.

Allow users to earn rewards in the read aloud in sentence level instead of the whole story level.

 

Provide users the option to turn the English translation on or off.

08

User Flow

new user flow.png

*The highlighted ones are the new flows. 

09

Wireframe

AST Wireframe.png

10

Final Design

1. Game System

The existing point system allows users to collect points by doing activities. The higher points they collect, the better their ranking on the leaderboard is. However, based on my search, our current leaderboard doesn’t seem to be an effective incentive for many users since users know no ones on the leaderboard and nobody knows them either. 

 

To better motivate users in learning, I added a coin system where users can convert their points to coins which they can spend to upgrade and decorate their avatar. Users can get enjoyment and motivation in three ways:

​​

  • The happiness of spending money

  • The ability to create their own avatar

  • The joy when their avatar can be seemed from the leaderboard

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

game system.png

Before

After

10

Final Design

1. Game System

To better motivate users in learning, I added a coin and shop system, parallel to the points and leaderboard system. Users can earn points and coins by doing activities and spend coins in the shop to upgrade and decorate their avatar.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AIS_Home.png
Shop_Accessories.png
checkout.png
AIS_quiz.png
AIS_summary.png
AIS_congratulations.png
AIS Icons-1video.png
AIS Icons-1quiz.png
AIS Icons-read aloud.png

                                                   Animated Stories                  Quiz                      Read Aloud

AIS Icons-1point.png
AIS Icons- 1coins.png

                                                                               Points                            Coins

AIS Icons-1leaderboard.png
AIS Icons-1shop.png

                                                                           Leaderboard                      Shop   

2. Friend Invitation

In order to increase the number of users and resolve the problem of "knowing nobody on the leaderboard", I designed a Friend Invitation feature. By inviting friends, users can not only collect coins but also read stories together with friends, which boosts the user's engagement with the platform.

AIS_Home.png

3. Story Recommendation and Search Optimization

During the onboarding phase, users' grade level and interest information are collected so as to recommend them the most appropriate content, which can be accessed under the "For You" level on the story page.

The search page is reinvented by adding "search by voice", "My Interests", and "All interests" features.

AIS_onboarding1.png
AIS_onboarding2.png
AIS_onboarding3.png
AIS_Library.png
AIS_search.png

4. Read Aloud

To take care of different users' needs, I add a step to ask users whether they need the English translation before the Read Aloud section. Also, to encourage the usage of the Read Aloud and help users make progress step by step, we reward users on their reading on sentence level instead of the whole passage.

Story page.png
Read aloud0.png
Read aloud4.png
Read aloud3.png
Read aloud2.png
Read aloud1.png

11

Reading Contest

In order to promote the new features and help more learners with their Chinese during the winter. I designed a reading contest called Read-O-Rama on AI Story Time. Learners can collect points by doing different activities. The top 133 learners with the highest points will be rewarded. The contest will end on March 1, 2020. I am looking forward to the result! 

Read%20O%20Rama_edited.jpg
Read%20o%20rama_edited.jpg