Sqft is a passion project I worked on to test out a product idea that aims to redefine house hunting experience under the specific case of Hong Kong.  Through a mix of system suggestions leveraging smart filters and user customization, Sqft helps with user’s searching process and better informs the decision-making. 

User research, Wireframing, UX/UI design  |  Nov 2016 (5 weeks)  |  Passion Project


"With an average property price 19 times more than the median household income, Hong Kong ranks as the least affordable city to buy a house globally.”

House hunting has never been an easy task. In the case of Hong Kong where buying a property means mission impossible due to the sky-high housing price, renting a place is a decision as important for people as buying a house.


It is also the first thing to deal with when people just move to the city and know nothing about the renting landscape. So how can I help make the house-hunting journey less stressful and the outcomes more satisfying for the non-locals here? 


I first talked to the non-locals and new comers around to get an idea about their house-hunting approaches.

What I found - 

  • The social nature in Wechat made it easier for posts looking for roommates/houses to go viral and matched the right people quickly.

  • If Facebook page and Wechat where people couldn't filter any post were still popular approaches, did that mean existing house-hunting apps could not do a sufficient job?

  • House agents reduced the searching burden, yet there were communication barriers and trust issues. 


To get more in-depth insights about people's renting experiences, I interviewed 8 people including college students/graduates who moved out from student dormitories and expats who had just been relocated to Hong Kong. Their decision-making process during the searching immediately drew my attention.

Existing filters could not help with user's sifting process in browsing listings - information still overwhelmed users. 

1. Criteria went from clear-cut yes/no to uncertainty

People were very clear when they first considered location, budget, room type...But when I presented them some listings online and asked how did they feel, their attitudes became vague and couldn't tell exactly what they did not like about them.

2. Choosing from abundance led to frustration

Basic filters in existing platforms did not significantly help with the sifting process. Listings that met the ‘hard criteria’ were plenty, leaving users the burden of going through hundreds of posts to make further comparison. This overwhelmed their cognitive loads and led to their frustration.

This means,

The sweet spot to tackle the question lies in the searching and filtering process. 

User Journey Map


What do we mean by better searching and filtering process? The pitfall is to add more filters so as to reduce the number of choices. Yet, as shown by the interviews, what was really time-consuming were the soft, subjective criteria that couldn’t be easily captured by traditional filters. 

The questions we should ask, that would help approximate to a more successful house-hunting experience are,

How can we inform user's decision-making?

How can we learn more about user’s preferences?

1. Location filters tapping into user’s social profile 

Location is an important factor to narrow down user’s searching scope at the very beginning, but a simple map is not conveying the real important information - where exactly should they live? What would be a better location choice for the user?

2. Leverage visual filters to better understand user's 'soft' criteria

Natural lighting, layout of the room, levels of renovation…some of these factors are the 'game-changers' for user to decide whether they want to look into the listing. Yet, they cannot be captured by traditional text filters. Sqft uses visual filters to learn about these 'soft' criteria and preferences and update its recommendations. 

Visual filters provides a new way to measure user's soft criteria that cannot be captured by traditional filters. 

3. Soliciting in-the-moment feedbacks to improve recommendations

The bookmarked listings can tell a lot about what a user likes, so as the listings that user doesn't like. Soliciting user's feedback right after they like or dislike a listing can help better gauge user’s personal preferences and return more accurate results. 

Swiping a recommended list away triggers a just-in-time survey to ask about user's feedbacks and thus improves following system recommendations. 


Pinpointing the problem to solve was the true 'AHA' moment for me in the project. There were a lot of potential improvements that could be made in the house-hunting journey if I had more time to spend on it , for example, better communication channel for users and listers, more features helping user compare and decide among a few options. 


Yet, with time and resources constraints, designers often have to decide on what would be the most impactful problem to approach, which feature would be most value-adding to have. Sometimes, priorities do need to be set and solutions do come with tradeoffs.   

© 2019 designed by ann peng.