FangFang is a virtual assistant that collects both housing and roommate information for struggling renters


Project Objective:

The project was done for 58.Inc. I was responsible for this project, and I collaborated with data analysts, UI / UX designers, researchers, design leads and product managers in the company.

Project Details

  • Project Duration: almost 3 months

  • Project Type: Internship at 58.Com Inc

  • My Role: UX design intern

  • Approach: Primary research, ideation, UI, prototype, video editing.

    ( With the help of data analysts, researchers and other designers )


The Challenge

It’s really hard for renters to find good roommates and a nice home at the same time:

1. People gradually lose their patience throughout the process

2. Passive communication brings problems between roommates

3. People feel helpless and in pain but there is nothing they can do


The Solution

Fangfang is a virtual assistant that collects both housing and roommate information, and create groups for struggling renters. Our goal is to help them find a comfortable home in this bustling city.

It’s an efficient and simple way to offer you a perfect match.


Key Features

set up.gif

Intro & Sign in

We provide the overview of the application, so it’s easier for you to understand the mental model of FangFang.

By connecting with your social media, you have a higher chance to match with your friends.

2 search+view.gif

Type location & search groups

Type any address, city, zip code or neighborhood, and we help you match with appropriate groups.


Join the group chat and check group info

You can make your considerations by chatting with the rest of the group members and browsing the housing info.

By clicking the “list” button, you can check out the price and amenities of each room.

4. others' info.gif

View other user’s profile

If you are curious, click on the group members’ profile picture to learn more about them.

Besides their basic info, you can see your similarities, their lifestyle, personality, interests, and their expectations on the profile.


Make your decision

To improve efficiency, you have a limited time to consider, after that you need to confirm this group, move to the waiting list instead, or quit this group.

Group 10.png

Work distribution & timeline:

  • Collaborate with Inc: market research, user research, system diagrams, ideation ( 2018. 6.25 - 8.15 )

  • Individual work: final idea, user flow, UI, prototype, usability test, video editing ( 2018. 12.24 - 12.28 )


1.Background Research


Influenced by national policies and multiple factors, the rental market will continue to grow in the future. It is estimated that by 2023, the rental population will reach 248 million. Newest policy is beneficial to further promote the development of the housing rental market.

Also, according to a new paper from the World Resources Institute, 330 million households, or roughly 1.2 billion people in the US lack access to secure, affordable spaces to call home.


2.User Research

Empathy Maps & Semantic Profiles

We sent out 200 surveys through our 58.Com platform. Based on participants’ info, we divided them into different groups, and we selected the same number of participants from each group to interview. Empathy maps, which organizes the information, made it easier to communicate with other researchers.

For the semantic profile, we divided the interviewees into three different user groups based on our analysis.

Empathy maps.png

Semantic profiles measure peoples’ attitudes about products, services, experiences, concepts, and similar entities. They are used to compare how different user groups think and feel about entities.



Persona info: age + city / hometown / personality / lifestyle / attitude towards house and roommates.


Analysis of Target Users:

The mental model and ecosystem map helps me to understand our target user group better. I reorganized the information I found from the interviews into 4 concentric circles — Who, What, Why, How might we/Insights.

I also separated information based on people’s behaviors, thoughts, expectations, and realizations. Interestingly, I found some gaps in between, which generated a lot of helpful insights.

Mental model of target users



1. People gradually lose their patience throughout the process

I started from searching by myself. Then, I downloaded 5 apps and also posted my information on social media. It took me too much time... I ended in asking agencies to help me.
— Mr. Zhu
At first, I could keep browsing the house for a few hours. I even thought about decorations in the living room, but gradually I didn’t know which to choose… The more choices I had, the messier my brain got.
— Miss Si

2. Passive communication brings problems

I told them one time, but they kept doing this and we didn’t know how to deal with it.
— Mr. Zhu
I just entered society. Even though I don’t like what they did, I want to avoid conflicts. Living in the same house, it shouldn’t be too awkward.
— Miss Jiao
I don’t know how to say it... my male roommate always reveals his upper body... It is really embarrassing.
— Miss Fan

3. People feel helpless and in pain but there is nothing they can do

What can I do? This is not my house... I rent it.
— Mr. Li
It’s always the landlord decides who will live here. Well... as long as you give him money, he will allow you to move in. I mean it’s his house, I can’t choose roommates...
— Mr. Chai
There is no way you can solve the problem. Even though you tell the landlord, he will just ask you to fix it by yourself.
— Mr. Zhu

3.How Might We


So, how might we help users to decide on housing and roommates?


4.Ideation & Iterations


I organized 3 ideation events in the company. Over 60 people participated and devoted 300+ ideas including researchers, UX designers, UI designers, design leads, and product managers.


Asking for feedback is an efficient way to iterate concepts. By presenting our findings and ideas to people, we gradually simplified our strategy — connect housing and roommate info to improve users’ searching efficiency. I then highlighted more plausible ideas.


Visual Design

“What makes you feel like home?” I asked.

“A place that I know where I can go back even in the midnight. “My interviewees told me.

So I decided to use dark colors to represent “night“ and light colors as metaphors of lights.



Style tile


Concept Development


5. Final Idea - FangFang

Walk Through


User Flow:

Final wireframes & user flow

Concept model

Next Step & What did I learn?

I still need to run a few rounds of usability tests to fix some small problems.

This was the first time that I got to know the real world, talk to real people and solve their real problems. I still remember at the end of the interviews, my interviewees asked me whether their voices would be heard. I said yes. I was really proud when I presented my research results to the company. Everyone was really into what I found out. I made it, and I not only allowed their voices to be heard by the company, but also helped millions of users in China.

Working with data analytics, researchers makes me realize the magic of data and how to build empathy with our customers.

It was a really unforgettable working experience for me, which I am proud of.


Thank you so much for your time. I hope you enjoyed it!