Roommates matching app
Helping people find their ideal roommate.
Design challenge for a UX role I was pursuing. Took me 7 days to research, design and document.
Design challenge
Design a mobile product experience that appeals to millennials that makes it safe to find the ideal roommate in New York City. Design the experience from the perspective of the person who is looking for a roommate as well as the one who is looking for a room in an apartment.
Design Process
I jot down every random idea I get, and any information I can find. I then use a tool called workflowy to organize my thoughts and information.
In this article I will be discussing my processes in the steps as shown in the image below.
Research
My research sources were from several YouTube videos highlighting the experiences and pain points of finding roommates in New York City. I also looked for tweets on the topic, and recollected my personal experiences as well.
Understanding the NY rental scene
Expensive
“NYC is great but I wouldn’t be willing to live with no background checked roommates just for the chance to pay through the nose for a small room and work myself to the bone daily.”
Tenancy laws and regulations
Security deposits are limited to one month’s rent and landlords must return it within 14 days of the tenant ending lease.
Application fees are limited to $20.
Interested tenants should meet rent requirements, which includes credit score, background check, income check. Many people have to rely on a “millionaire aunt” or a rich relative to be a guarantor and co-sign the lease. If they don’t have a guarantor, they are forced to pay several months’ rent upfront to be able to lease.
We could help users by educating them about rental laws in the region, so they don’t get cheated for lack of awareness.
Brokers and fees
People tend to avoid listings posted by brokers, as brokers charge one to two months of rent as their fee.
People also don’t trust that agents have tenant’s best interests.
“This broker is in charge of putting you and 5 other strangers on the internet on craigslist to live a harmonious life in a 5 bedroom remodeled apartment”
We could highlight if a listing is by a tenant or an agent. We could also highlight the rating an agent has, so as to build trust and comfort with users.
User pain points
Some of the major concerns and pain points were related to:
Compatibility
A lot of friction occurs for cleanliness, food, noise levels, and overnight guests.
Safety
Bombardment of communication
Competition analysis
To find roommates, users heavily use craigslist, some use facebook groups, and roommate matching apps. I checked out several competitor apps, such as Roomster, Roomi, No Broker, Spare room, Badi. I also checked out several hotel booking apps such as Booking.com and Airbnb.
Positives
Negatives
- Critiques on app stores suggest people are very dissatisfied with search quality.
- Numerous complaints about some apps riddled with 90% agent profiles who scam people with fake photos and entice them to share their numbers and constantly bombard them with spam messages.
- A user was unhappy as they couldn’t use LGBT in keywords to search for receptive homes.
To evaluate
Lister and finder modes?
Airbnb and Badi let you switch modes, from hosting/offering a room to looking for a stay/room.
At this stage, I decided to not have listed and finder modes, for the following reasons -
- User profiles may change frequently. A user who is offering a room to rent, might at a later point seek a room in a house once their current lease is over.
- Some users might be simultaneously looking to offer a room, as well as find a room to stay. For eg. a user looking to sublet their room, and finding a room in another city where they plan to move.
At a later point, we can have an interface dedicated to fit agent needs.
What would be our differentiators?
Superior search and browse experience
Ratings and reviews on competitor apps indicate users are unhappy with search results. Listing cards don’t show enough information, certain important filters are missing, sort functions are limited. Applied filters don’t reflect in results.
Posted by agent?
None of the competitor apps differentiate who posted the listing — a tenant, an agent or a landlord. An agent posting might incur agent fee. Clarifying who posted the listing on the listing card, and allowing users to filter search results to exclude such listings from search results.
This idea would have to be discussed and debated over with all stakeholders, as it might work negatively for agents, who might be major contributors to the revenue generated by company.
Dealbreakers
People may be compatible on many aspects — but there might be some factors that would be deal breakers, no matter how much they match overall. For example, some strict vegetarians in some cultures wouldn’t want to live with non-vegetarians. In New York, people have very strong political inclinations, and will ignore any nice attributes of anyone who favors the opponent candidate.
So in user profile forms, people should be able to define deal breakers. And apply them as search filters when scanning for matches.
For eg.
Trump or Hillary? (user chooses Hillary)
If a potential roommate chose Trump, would that be a deal breaker for you?
Yes or No
User Personas
There are two broad categories of users — those offering a room, and those looking for a room. Listed below are select examples from both categories.
Living first time without family
Looking for a place with a group of friendly and helpful roommates. They are usually looking for roommates in their age group, in the same school they are studying in. They may also seek people from similar cultural backgrounds such as theirs.
Last minute panic
Tend to message dozens of people. Willing to compromise on preferences. Safety takes back seat.
Subletter
Subletters want to move out of their apartment and are looking for someone ASAP to take over their lease. They usually don’t care about who they are subletting to, compatibility isn’t a concern.
Agent
Has a huge volume of listings to manage, and has to arrange logistics to set up meetings. Goal — save time, manage reputation.
Landlord
Needs to find the right tenants who are respectful of the property. Special tenant requests before signing a lease. They tend to prefer professionals with a stable income.
Defining design goals
There are several categories of users with their own unique needs. There are some requirements such as compatibility, efficiency of search, which are common across all personas.
Let’s solve for Katherine, whose roommate has broken the lease, and has to find a new roommate in two weeks!
Katherine’s journey
Designs goals to help Katherine find her roommate
- Relevant and custom search
- Efficient screening of matches
- Compatibility
- Trustworthiness
- No stale listings
- Recommendations for a match
Feature analysis
Here are the features needed to support the design goals.
Search
An efficient and user friendly search process that helps users filter relevant results, is crucial to the success of the app. It impacts all the user personas.
List search
In this design version, we show a bare minimum of most important information on the listing card. Tapping on the listing card will go to the listing detail. Clicking on the down arrow button will expand the card to reveal more information about the listing, and user can continue browsing the results.
Alternative designs for list search
We could do an evaluative research initially with a group of users to understand which design they find more user friendly, and serves their needs best.
Full screen search
Just to experiment — we could test a full screen search result mode, where the most relevant information shows up first. If the user is interested, they can scroll further. Or they can just swipe up to see the next listing.
Map search
People search
People offering a room can look for profiles who are seeking a room, sort them by compatibility and moving dates. They can filter results based on gender, age group, verification status, if they are students or working, last seen activity, and ‘dealbreakers’.
How might we reduce the number of stale listings?
- Remind and request users to update the status of their search — rented, found a room, or still looking?
- Allow users to filter and sort listings based on recency, or last activity.
- If a lister has been unresponsive to emails and communication for more than two days, drop the rank of the listing. Bring the rank up again if the there is activity again.
- For people search, drop ranking of people who haven’t been active on the app for 3 days. (Time span can be reduced or increased based on observations.
- Show last active on users’ profiles
Listing details
Forms
Create listing
- Easy to navigate form sections through steps on a backdrop layer.
- Show completion progress
- Allow user to save and exit form (could also auto-save)
Break down the create listing form in several sections. Each section will have relevant fields that will give finders a good idea about the room, and also search relevancy through filters.
User profiles
Compatibility quizzes
Finding a roommate is almost like dating! You want the perfect match! When users initiate a chat with someone new, they are hesitant to ask too many questions, as it is ‘too soon’. Quizzes give answers to burning questions and move things forward faster! :)
“Do you have a roomate”=”do you have a boyfriend” TOO EARLY TOO EARLY
How do we do things differently?
Most apps ask users to fill out a form, such as are you clean or messy? What could be clean in one user’s perspective, may not be so in another’s. More so, people will always suggest they are clean and tidy.
In the quiz designed below, we show users pictures of the same room. That levels the grounds on cleanliness standards. We also challenge the users to quickly take a picture of their room. That will minmize the “My roommate isn’t as clean as they claimed” surprises to some extent.
Our quiz will be a mix of various formats of questions, so users stay engaged. It could be an ever growing library of questions. A well gamified quiz could be shared on Facebook, “Which roommate type are you?”, and could help with growth and marketing.
Public view
Public profile view gives other users a quick understanding of compatibility between them, and their rooming preferences. It builds trust about the person, to see a video, how long they have been on the platform, and their verification status.
One feature we could add in profiles is a ‘status’, which states, “I am currently looking for a room”, or “Already found roommate.” This will help reduce the number of stale profiles.
Private view
The private profile view will have similar content as public profile, plus some nudges to complete their profile and verification if incomplete.
Chats management
Often users are bombarded with dozens of messages. (agents even more so, as they list multiple properties. Hence, chat management becomes important. Some features to consider:
Canned responses
Standard message templates for common questions or scenarios, such as application rejection template, request for background verification template, and so on.
Overall culture goal
- Common etiquette- don’t lead, keep your word, don’t spam.
- Educating how to avoid scams
- Educating about max application fee and other city rules
- Market research on area — what’s the average rent?
More features in the future
- Community feature, where users can post statuses of their search and ask for help.
- Ability for users to connect with other users looking to rent different rooms in the same apartment. Groups feature would serve handy.
- Deposit escrow? We could think of holding deposit in escrow, so seriously interested renters are easily distinguished.
- Ratings? I am unsure about ratings for tenants, as the number of roommates who can vouch for the person will be few and infrequent. And some reviews could be unfair. Ratings for agents on the other hand is a must, and they will have sufficient volumes as well, coz of the rapid business.
- Lease contracts and other legal documents
Information Architecture (Navigation)
User flow and navigation for person looking for a room
Initial low fidelity (with audio explanation)
High fidelity
User flow for a person offering a room for rent
Initial low fidelity (with audio explanation)
Final high fidelity
Success
Factors that can determine success
Number of listings
Craigslist is still the most used platform for roommates search, for the sheer volume of listings on the platform.
Get the job done
At the end of the day, people should successfully find a place to live, and compatible roommates. Search, filter, forms, profiles should be user friendly, efficient and with minimal bugs.
Genuine profiles
Real users with transparent and verified profiles will help build an environment of trust and safety. We should implement means to identify and eliminate catfish accounts, scammers and sexual predators.
How do we measure success?
- Growth in number of users, listings, and transactions and activity occurring on the platform
- Good reviews on app stores
- Referrals and recommendations from satisfied users
- Users opting in for our paid services
Retention
- First time success and satisfaction will give a reason for the user to come back again in the future when they are looking for roommates.
- Having user profiles already filled out, verification and background check already completed on the platform, will give more reasons for a user to return.
- If a user has built a certain reputation on the platform (specially for agents and landlords), they are more likely to make more successful transactions on the platform, which will help retention.
How could we monetize?
Boost listing
Canned responses
Bulk messages
Video listings
Show 30 unique messages, charge to unlock more
Ad in Chat (like Linkedin inmails)
Looking for a room max Interest messages limit
