My Framer Site

TutorS

Capstone project, November 2022

Despite the rising demand for online tutoring during the Covid-19 pandemic, teachers and tutors faced challenges in finding students. By uncovering key user insights, I designed an MVP app, TutorS, a platform that matches tutors with students. Through user testing, I achieved a 36% increase in usability by the end of the project timeline.

My Role

Lead UX/UI Designer

Timeline

80-hour time constraint

Technical Constraints

Conceptual capstone project
Product constrainted to MVP

Splash screen to main login page.

Splash screen.

Key pages for the main flow.

Background: Addressing the Online Tutoring Challenge in South Korea

South Korea has a large audience of education enthusiasts, with tutoring being one of the most popular forms of education among parents and students.

However, despite the growing demand for online tutoring since the Covid-19 pandemic, many of my co-workers lost their tutoring jobs due to the outbreak. This sparked my decision to investigate the underlying user problem and design a solution to address it.

South Korea has a large audience of education enthusiasts, with tutoring being one of the most popular forms of education among parents and students.

Overview

— Identifying and Solving the Tutoring Challenge —

Through a combination of market research, secondary research, and user interviews, I identified a key problem:

Teachers were struggling to match their teaching styles with students' learning preferences due to the limited job opportunities and the influence of personal connections, like mom-networking.

To address this issue, I designed an app solution, TutorS, which allows tutors to connect with and reach out to their ideal clients directly.

Through a combination of market research, secondary research, and user interviews, I identified a key problem:

Teachers were struggling to match their teaching styles with students' learning preferences due to the limited job opportunities and the influence of personal connections, like mom-networking.

To address this issue, I designed an app solution, TutorS, which allows tutors to connect with and reach out to their ideal clients directly.

— User Testing and Improvement of Usability —

Due to time constraints, I conducted just one round of user testing, which revealed key user problems on the first day. After addressing these issues, the usability score, which started at less than 35%, improved to 68% by the final day of testing.

Skip to see final design details

01 Product Discovery

Secondary Research

VALIDATING FIRST ASSUMPTIONS

A blue ocean in the tutoring/education industry in South Korea.

I began with secondary research to investigate the growing demand for online tutor matching platforms. Numerous articles highlighted the rapid expansion of this industry in South Korea, confirming its rising popularity and potential.

“Increasing Number of Video Tutoring Matching Platforms Amid the COVID-19 Pandemic”

“As a result, "Kim Tutors" showed a growth in monthly transactions from 1 billion won in 2019 to 2 billion won in the second half of 2020. The number of users of the tutoring service”Seoltab” has also increased significantly, with the number of new students more than doubling since June 2020.”

An article written by Kookmin University - Korean/original (left) and English/translated (right).

Comparative Research

TAKING A LOOK AT THE MARKET

Despite the number of apps, none of them were the same.

The growing interest in online tutor matching platforms also brought significant competition to the industry. To better understand the landscape, I analyzed four tutoring app services specifically based in South Korea.

Key Findings:

Most platforms primarily focused on pairing tutors and students for offline tutoring.
Platforms dedicated entirely to online tutoring lacked a focus on specialized matching between tutors and students.

Below is a detailed breakdown of my findings for each platform during the research.

Key Findings:

Most platforms primarily focused on pairing tutors and students for offline tutoring.
Platforms dedicated entirely to online tutoring lacked a focus on specialized matching between tutors and students.

Below is a detailed breakdown of my findings for each platform during the research.

Competitive research notes. Highlights the amount of focus on online tutoring in the platform as well as notable points for each service.

Soomgo.

Incredible amount of options of services.

Extensive, but online learning was not the main priority.

TeachUs.

Assigns students with a designated coordinator who helps with scheduling etc.

Teachers need to be “hired” and only except students from top elite schools.

Ossam.

Kept their service to the bare minimum - the business model and their platform interface.

No way of teachers reaching out to clients.

Kim Tutors.

Online tutoring is introduced as one of its subcategories.

Cluttered UI.

User Interviews

TALKING TO THE DECISION-MAKERS

Teachers AND parents.

Conducting interviews was a crucial step in this project to understand firsthand user experiences and identify trends and patterns in how tutors find work.

I focused on two key user groups:

Teachers with tutoring or teaching experience – to understand their challenges and processes in finding tutoring opportunities.
Parents who have hired tutors for their children – because they play a pivotal role in networking and decision-making when selecting tutors.

Conducting interviews was a crucial step in this project to understand firsthand user experiences and identify trends and patterns in how tutors find work.

I focused on two key user groups:

Teachers with tutoring or teaching experience – to understand their challenges and processes in finding tutoring opportunities.
Parents who have hired tutors for their children – because they play a pivotal role in networking and decision-making when selecting tutors.

Through a messenger app, KakaoTalk, I went through sessions on audio calls & video calls.

02 Synthesis

Data Synthesis and Analysis

TEACHERS HAVE IT HARD MAKING DISCREET DECISIONS

THEME 1: POWERLESS TEACHERS

Networking Reliance: 5 out of 6 teachers stated that they primarily find tutoring jobs through close connections or mom-networking.

Scarcity of Opportunities: Teachers noted that these opportunities are relatively "scarce and rare" compared to those available through broker platforms.

High Acceptance Rate: All interviewed teachers reported accepting any offers they received, highlighting the limited availability of opportunities.

Quotes from teachers’ replies.

THEME 2: THE HOW AND THE LEARNING STYLES

Curriculum as a Priority: Both teachers and parents emphasized that initial or pre-sessions primarily focus on curriculum-related information.

Teaching-Learning Style Mismatch: 84% of participants reported encountering issues in tutoring due to mismatches between teaching styles and students’ learning styles.

Struggled due to lack of understanding of student’s learning styles/needs.

Wished to have understood the personalities and learning styles prior to committing to full sessions.

Struggled due to lack of understanding of student’s learning styles/needs.

Wished to have understood the personalities and learning styles prior to committing to full sessions.

Struggled due to lack of understanding of student’s learning styles/needs.

Wished to have understood the personalities and learning styles prior to committing to full sessions.

Struggled due to lack of understanding of student’s learning styles/needs.

Wished to have understood the personalities and learning styles prior to committing to full sessions.

Patterns discovered and clustered from replies.

03 Define

Key User Problem

PROBLEM: INSUFFICIENT COMMUNICATION

Theme 2 leading to theme 1.

After synthesizing data above, there was a question that was raised from theme 2.

This question helped me gain further clarity on one user problem to focus on.

After synthesizing data above, there was a question that was raised from theme 2.

This question helped me gain further clarity on one user problem to focus on.

The question:

Couldn’t they have communicated ’Theme 2 (The how and learning styles)’ before committing to sessions?

The question:

Couldn’t they have communicated ’Theme 2 (The how and learning styles)’ before committing to sessions?

Short answer: no.

It was important to note that this ties back to theme 1 where teachers often have less control in deciding whether to commit to sessions.

Short answer: no.

It was important to note that this ties back to theme 1 where teachers often have less control in deciding whether to commit to sessions.

KEY USER PROBLEM:

Teachers struggled to align their teaching styles with students’ learning styles due to a limited availability of job opportunities and an overreliance on personal connections, such as mom-networking.

04 Brainstorming

Main Task flow

BRAINSTORMING A SOLUTION

Enable teachers to go first!

To address the identified problem, I explored various user flows with a key focus on communication. The primary question was:

What shape or form should this “communication” take?

Initially, I felt stuck, as direct messages (DMs) and text-based communication seemed like the only viable options.

Then, I reframed the challenge: “If this approach is inevitable, how can it be improved or tailored to better meet user needs?” This shift in perspective opened the door to refining the solution.

To address the identified problem, I explored various user flows with a key focus on communication. The primary question was:

What shape or form should this “communication” take?

Initially, I felt stuck, as direct messages (DMs) and text-based communication seemed like the only viable options.

Teachers choose!

Sending quotes

Sending DMs

Tutor

Student

Flow A

Flow A:

Tutor

Parent

Student

Flow B

Flow B:

Simplified diagrams of how communication would be achieved in matching process. Direction of communication (top A and B). Methods of communication (bottom diagram).

Tutors first!

Whether talking to parents or the students themselves, communication starts with the tutors. This way, tutors are to seek & consider their options without having the pressure to committing to every single offer.

A more formatted form - quotes.

It went from DMs and text messages to forms, or quotes. This gives a clearer way of communicating additional details such as the non-academic information.

Fleshing out the flow.

Now was time to figure out the precise flow of how these quotes would be sent. Drawing inspiration from the apps I had researched earlier, I designed a streamlined flow that made the most sense.

Task flow.

Landing on the page of “request”s.

Here, teachers are given the advantage to pick and choose which clients to choose from at their own pace. Once signed-up, teachers land on a page where students upload their "requests" on the platform.

Sending quotes to kick-start the interaction.

This stemmed from the thought of who gets to see the options first and starts the interaction.

The teachers are able to take a look at a summary of the client’s needs and decide whether they want to teach the student. IA point to note was who gets to see the options first and start the interaction.

This stemmed from the thought of who gets to see the options first and starts the interaction.

Engage and report.

When the client further details, both parties can discuss through a chat function and once settled, the teacher reports session details in the platform.

Finalizing the match.

According to the recorded data, a service fee is calculated and the matching process is complete.

05 Branding + UI

UI Design

BRAINSTORMING FOR VISUALS

Let’s start with the name for this app.

This platform is to provide learning services online but more importantly, to focus on the perspective of tutors and their matching. In this sense, I wanted to go with a name that was straightforward and one that sends the users the message right away: TutorS.

Sketches considered and anatomy of the final ‘TutorS’ logo

Thoughtful Branding, Done Efficiently

I chose a deep blue color (2F1FE2) which gives the brand a meaning of trust, inspiration and intelligence.

Although the logos were personally designed, most of the icons were used from utilizing pre-made icon packs from the Figma Community to save extra time and abide by the time constraint.

I chose a deep blue color (2F1FE2) which gives the brand a meaning of trust, inspiration and intelligence.

Although the logos were personally designed, most of the icons were used from utilizing pre-made icon packs from the Figma Community to save extra time and abide by the time constraint.

UI kit. Consists of variations of logo, colour palette and icons.

Thought Designing, Done Efficiently

I decided to keep the strokes and lines of the overall UI minimal.

I put more emphasis on design specs to design components as clean and crisp as possible, paying attention to how the designs would be communicated for potential implementation.

I decided to keep the strokes and lines of the overall UI minimal.

I put more emphasis on design specs to design components as clean and crisp as possible, paying attention to how the designs would be communicated for potential implementation.

Buttons

CTA

Normal

Selected

Non- selected

CTA buttons

Radio buttons

Checkboxes

Clicked

Disabled

Typography

Ubuntu, Medium, 20

Ubuntu, Medium, 16

Ubuntu, Regular, 13

Ubuntu, Regular, 15

Questions, input guides

Message texts

Page headings

Request answers

20px

chayee678

English grammar

Grade 5

Female

Seoul, Korea

M/Tu/Thu/Sa

goldbake06

Culinary

30s

Female

Seoul, Korea

Tu/Thu/Sa

meraki78

English grammar

Grade 4

Female

Seoul, Korea

M/Tu/Thu/Sa

meraki78

English grammar

Grade 4

Female

Seoul, Korea

M/Tu/Thu/Sa

nosinikg1

English speaking

Grade 1

Male

Seoul, Korea

W/F/Sa

imkimmm

Korean

Grade 9

Female

Seoul, Korea

M/Thu/Sa

arinialim

Biology

Grade 11

Female

Seoul, Korea

Sa/Su

arinialim

Biology

Grade 11

Female

Seoul, Korea

Sa/Su

Requests

9:41

Requests

Messages

Profile

Settings

20px

16px

20px

We’re going to need some details to get you started!

WHAT

WHEN

Continue

9:41

Profile

*1 is required

Name

Nickname

I choose to show my nickname only

Gender

Year/month/day of birth

I choose to show age range only

Personality traits

*2 or more

Introverted

Extroverted

Critical

Calm

Humorous

Challenging

Warm

Bubbly

Earnest

Positive

Encouraging

Attentive

Check

Phone number

*contact info will be used when retrieving lost password

Based in

Time zone

CREDENTIALS

*This will help increase the chance of match

School

Campus

Major

Licenses/certificates

Other mentions

0/40 characters

20px

12px

8px

20px

32px

Variations of buttons and typography as well as a few screens to highlight design specs between elements. Gutters and paddings in multiples of 4px.

06 Testings

Testing out the product

LEARNING TO COPE WITH UNEXPECTED OCCURRENCES

A brutal number of people that gave up.

I utilized the platform, Maze, to gather more extensive analytics as well as garner more people to interact with my product.

To first share the final results...

FINAL RESULTS

19 responses (20 were asked, and 1 wasn’t able to participate due to scheduling issues)
About 68% were able to go through the process (direct/indirect),
the other 32% gave up.

To show you how I got to these figures, let’s dive deeper into the process.

I utilized the platform, Maze, to gather more extensive analytics as well as garner more people to interact with my product.

To first share the final results...

FINAL RESULTS

19 responses (20 were asked, and 1 wasn’t able to participate due to scheduling issues)
About 68% were able to go through the process (direct/indirect),
the other 32% gave up.

To show you how I got to these figures, let’s dive deeper into the process.

The first 7 responses out of 19. 4 give ups and 3 completions.

A more proactive monitoring was needed.

During the beginning of the testing period, it was clear that something was up, and I looked into why people were having such a hard time with the testings. I contacted some of those participants to find out what the problem was.

INSIGHTS

Found out that from the first 6 tests, more than half of the results came back "give up/bounce".
It turned out that they weren't paying close attention to the instructions for each "mission" and had a hard time getting around the prototype.

INSIGHTS

Found out that from the first 6 tests, more than half of the results came back "give up/bounce".
It turned out that they weren't paying close attention to the instructions for each "mission" and had a hard time getting around the prototype.

... I thought I had to fill out the form but it didn't let me. Did I have to?

Hi Sophia, just wanted to ask about the testings - were you able to complete the task all the way? ...

Excerpt from one follow-up with a participant.

Heatmaps screenshots from sign up flows (type of user on left and form fields on right).

Clicked on wrong user.

Participants(who gave up/bounced) were clicking on 'STUDENT' when the mission was to sign up as a 'TEACHER'.

Clicked on other buttons instead of CTA.

Many were clicking on input fields instead of the CTA button when they were instructed to just browse the input fields. (Mentioned in text before starting the tests.)

From less than 35% to 68%.

With the insights above, I had 2 options.

Redo Test OR Carry On

Although I would have to compromise varying test results, with the time constraint in mind, I addressed the problem by carrying on with letting participants know of the instructions once more.

Having addressed the problem, what showed less than 35% of usability score on the first day of testing was able to reach 68% on the last day.

With the insights above, I had 2 options.

Redo Test OR Carry On

Although I would have to compromise varying test results, with the time constraint in mind, I addressed the problem by carrying on with letting participants know of the instructions once more.

Having addressed the problem, what showed less than 35% of usability score on the first day of testing was able to reach 68% on the last day.

Screenshots of further instructions sent to participants who have yet to complete the task.

Utilizing what surfaced earlier to iterate

THE LACK OF INTERACTION IN CTAs

Interactions matter.

One of the problems I came across during testings was when users weren’t able to get pass the sign up form fields. The expected flow was for them to browse the form and eventually find the CTA to proceed to the next screen.

I initally thought this was primarily due to the user error of them not understanding the instructions. This was true to some extent, but I also discovered one crucial flaw in the prototype.

What prototype resulted in heatmaps
* wait ‘till the end!

Very little clicks on CTAs

Majority of clicks were in first viewport - users trying to click on what looked like clickable buttons

What prototyped used in testings resulted in heatmap analysis.

Fixed buttons in viewports.

In hindsight, this was a no-brainer. However, I was happy to see that test results resulted in discoveries of these mistakes. I definitely learned a lesson.

The CTA button that was at the bottom of the viewport was fixed in prototype settings so that users are able to view the “main” click point instead of panically clicking on other buttons.

In hindsight, this was a no-brainer. However, I was happy to see that test results resulted in discoveries of these mistakes. I definitely learned a lesson.

The CTA button that was at the bottom of the viewport was fixed in prototype settings so that users are able to view the “main” click point instead of panically clicking on other buttons.

Before

After (Iterated)

CTA button is fixed to bottom of frame

Prototype settings

Before (left), after (middle) and prototype settings in Figma (right).

07 Final designs

Showcasing final designs!

A TUTORING MATCHING PLATFORM FOR TUTORS

Signing up on the platform as a teacher.

Click to see full prototype

08 Reflections

Reflections

IN HINDSIGHT...

Here’s what came to mind after wrapping up the project.

Another user testing session needed!

Although I got a mini-panic attack with the first user testing results, it was very interesting how that shaped the way I had to think after the “release” of the product. Similarly, I would love to see what different results the iterated prototype reveals.

Different country, different research?

During the majority of time I spent on this project, I was in South Korea and hence, the user research was mainly catered to those living in the country. Towards the end of the project, I am also keen to see how my designs would turn out if research was done in Canada or in a different country.

Next up: