(Don’t ignore) the “ELLY” in the room.

Buying your own place (and living) in Vancouver is like Mission Impossible. Properties are overpriced and the cost of living is high. Not being able to afford, many Vancouverites resort to renting. They rent rooms and share spaces with others. They do so to cut the cost of living and to enhance their desired lifestyles.

Sharing space with others also possesses problems. Living with someone else is always a challenge. Difficulty increases when that someone else is incompatible. They ignore issues and avoid confrontations. They start living with “the elephant in the room”.

To help these struggling Vancouverites, we wanted to create an app. We wanted them to stop ignoring the elephant in the room. We want them to be able to live the life they deserve. We would match them with their ideal living space and to the ideal people they will be sharing their space with.

Room/mate “Tinder”


The first thing we did was to look at similar competitors on the market. We wanted to make sure there was no other platforms offering similar solutions. Craigslist, Airbnb, PadMapper, and Kijiji were amongst the many room-finding competitors we analysed.

Popular room-finding competitors on the market

None offer the type of “matching” feature we intended to offer. Most of the platforms provide large volumes of postings. Yet, few assists the users in filtering long lists of results. Fewer supports efficient save features. There is definitely an opportunity for us to enhance the user’s experience. We wanted to offer a more efficient alternative as we help them achieve their goal.

Storyboard of user interacting with our app

Jane lives with a roommate she feels disconnected with. She often dreams about all the fun she could get if she lived with people similar to her. She uses our room/roommate matching app and finds rooms tailored to her needs/wants. She successfully rents her new room and feels welcomed and comfortable with the people she lives together with.

To further solidify our idea, we created a storyboard. It illustrates the possible interactions our potential user would have with our app. Which also helped us explore the possible solutions we could offer.


We had a rough idea of the solutions and product we could create for our users. We researched and interviewed people in our intended user group to validated those assumptions.

Affinity diagram indicating key insights | Top criterions renters value
Lack of options to accurately locate desired postings


Many users experienced a lack of options on existing platforms. They are unable to filter the long list of postings down to a more tailored search.

Even if there was a filter system, it wasn’t specific enough. It didn’t offer the enough options. They were unable to find rooms with the right qualities. There are no means of verifying whether the roommates would be compatible either.

Waste of time and effort with no results


Too much time and effort is considered wasted when using existing platforms.

Users have to apply to each posting individually. Every time they find a desirable room, they have to repeat the same processes over again.

Contacting each landlord and arranging viewings are exhausting tasks. They require a lot of time, attention, and effort. And to do that multiple times until finding a match is frustrating.

Applying to rent a room is like applying for a job. Many users described long “interrogations” with potential landlords. Identifying the right qualities for a match — reliable candidate, culture fit etc — was high priority for the landlords.

Too many unreliable postings


Existing platforms are unreliable. A majority of our potential users frequent certain platforms because of the large volume of postings. Yet, they suspect a large percentage of that to be fake.

They have to risk running into scams. Some are easy to spot, others are not as obvious. Current competitors do not evaluate and block suspicious postings. They require their users to deduce and filter themselves.


From the key insights of our research, we created a user persona — Jane. Jane helps us better understand the behaviours and goals of our hypothetical user group.

JANE, 27. Recruitment professional


  • To pursue her interests without any money constraints — enjoying her ideal lifestyle
  • To live in a space of her dreams — beautiful design, private area, amenities etc
  • To connect with other people


  • Efficient use of time and money
  • Like-minded people to share her passions and interests with
  • External elements of the space she’s in — budget, location, proximity, accessibility etc


  • Lack of security and safety
  • Not being able to be herself, freely express herself — not belonging
  • Confrontations, miscommunications with other people
  • Unreliable platforms and people around her

We needed to identify Jane’s goals and motivations to be able to create an app tailored to her needs. We pinpointed the frustrations keeping her from reaching her goals. Our goal is to motivate her and to eliminate the barriers so she can successfully reach her goals.


Efficiency and effort were two of the main frustrations of our hypothetical user group. We wanted to make our users’ lives as simple as possible as they are using and navigating through our app.

We decided to introduce an A.I. The A.I. would not only guide our users to navigate the app, but also make the learning experience easier. When encountering any kind of problems and frustrations, the user would be able to go to the A.I. for assistance. The A.I. could provide answers to the user’s questions. The A.I. could perform a specific search. The A.I. could transport the user to another desired screen.

The main goal of the A.I. is to make the frustrating room-seeking process easier for our room-seekers. It would interact with the users in a conversational approach. The intention is to make the users more comfortable through casual narration.

The A.I. needed a form. It has to be tangible. If it is too abstract and intangible, the users would experience disconnection towards it.

We explored animals living in packs/herds. Our users are not only looking for the right rooms, but also the right people they will be living and sharing their lives with. Our app will help them find their right ‘herd’.

After looking into some potential animals, we narrowed down to a few. These animals have the qualities we needed in our app and A.I. to possess. Elephant was amongst the few in the list.

Choosing and validating the elephant concept

“The elephant in the room.”

There was a strong connection. It was our initial idea to “not ignore the elephant in the room” and “find the right room/mates”. We wanted to address the problem. We wanted our users to “go to the elephant”. And the “elephant” would then help them out and solve their problems. Despite the strong connection we felt, we had to validate our concept. Our users need to be able to make the same associations.

We listed the characteristics and qualities we wanted our app to have. We asked potential users to sort these characteristics under the animals they associate with. We provided a few animals from our original list of potentials.

Our testings confirmed some of our assumptions. Many people associated elephants as being comfortable, trustworthy, safe, helpful, understanding, friendly etc. These are the qualities we wanted our A.I. and app to have.


After deciding on the elephant, we began to brainstorm a name for our A.I. mascot. We knew what we wanted — something that is simple, and easy to remember. It also has to be easy on the tongue. We want our users to be able to effortlessly say and pronounce the name.

Inspiration of “ELLY”

We eventually decided on “Elly”. Elly stemmed from the first two syllables — e-le — of the word elephant. The “ee” ending allows users to sense friendliness and approachability.

Aside from being the name of the A.I., Elly is also the name of the app.


Having decided on the name “Elly”, we needed a face to go with it. We started rough sketches of Elly’s head.

We decided to use simple shapes (circles, squares, triangles) to construct Elly’s face. Being an A.I. mascot, we wanted Elly to represent the digital world. Simplicity mirrors the digital universe. Complexity overwhelms. The absence of complexity helps highlight the amiability of Elly.

After initial sketches, we started digitizing the faces. Digitizing allows us to visualize and complete our design. Besides to the simple shapes that makes up Elly’s face, we decided to round all the corners. Rounded corners diminish the rigidness and sharpness of the character. They help the mascot to become more engaging and approachable.

Finally, we painted Elly’s face in a light shade of pastel blue. The subtleness of the colour helps calm and ease the frustrations of our user. The blue not only relaxes, but also aids the mascot to establish trustworthiness.

Pencil sketches of “ELLY” | Various versions of digitized “ELLY” | Finalized logo/face of “ELLY”


One of the biggest challenges we faced was to create an app that is reliable and trustworthy. We needed our users to feel safe and secure using our app. Our users should feel safe — being able to comfortably be who they are and do what they want. They should also feel secure — trusting us to match them with rooms/roommates closest to their ideal.

The mood our app tries to capture

To establish the above, we continued with our use of the colour blue. We used a variety of shades of blue.

As previously mentioned, the colour blue creates a sense of soundness and tranquility. It is a key component in helping us build our brand reliability.

To accompany the solid blue theme, we included a tint of mustard yellow. Yellow generates positive, happy emotions. We are matching our users to their ideal living space and lifestyle. We want them to be enthusiastic about the dream life they are about to enjoy by using our app. Yet, we don’t want to overwhelm and distract them from the trustworthiness created with the blue scheme. We decided to use the yellow exclusively — as a highlight to help the user focus on information important to them.


Since we have Elly as our A.I. mascot, it is important to present the app in Elly’s voice. Our users should be comfortable and feel like they are casually conversing with Elly.

A sneak-peak at ELLY

The typography and language play important roles in how Elly would be perceived by our users. They have to somehow embody Elly’s character and personality.

We revisited some of the key characteristics we wanted our users to find in Elly — helpful, friendly and trustworthy. We eventually decided on Quicksand. The roundness of Quicksand softens the harshness of Elly’s tone. It makes Elly sound more approachable and friendly.

We have identified the opportunity for our app in the renting/housing market. But other than our room/mate matching concept, how are we different from other room finding platforms out there?


As mentioned previously, our main feature is to filter the massive amount of postings to those compatible to our user. Current platforms require a lot of user effort in narrowing down their search. The filter systems lack options and are not specific.

Our model allows the users to define their preferences once, and our system will do the rest of the work. Our main difference is we match our users with not only the rooms meeting their needs, but also with the right people sharing the same values. Our users would be able to enjoy a desired lifestyle. Living with people who are compatible, allows them to be comfortable being who they are, freely expressing themselves.


Current platforms offer no verification process. They require their users to apply their own critical thinking skill to filter out the fishy and unreliable postings themselves.

Along with declaring their profile and preferences, our users would also upload their certified background/credit checks. Yet, these confidential documents would not be accessible to other users. Completion of these procedures would only be shown as verified. The details would not be disclosed to the other party unless required in the final step of renting.

Our extra layer of verification allows our users to feel well protected in a secure environment.


To find the right match, our users have to declare their preferences so our system knows what/how to filter for them. With current platforms, users have to define and refine their searches each time. Our users would only have to do that once.

The background/credit checks mentioned above would only have to be uploaded once (until expiration). The users would not have to supply the documents again and again, each time they apply for a room they are interested in. Repeating the same procedures every single time is time consuming, effort heavy and frustrating.

Our one-time-form model allows our users to save time and effort.


Many users expressed frustrations in keeping track of the postings they have viewed before or the ones they have applied to.

Similar to Tinder, our users would be able to swipe and declare their interest in the matches filtered for them. The system automatically track all their decisions so they don’t have to view the same results over and over again. The system would track and show only the ones saved as interested. It would also help track the rooms the users have applied to.


Arranging viewings and communicating with landlords requires a lot of attention and work from the users. There’s a lot of back and forth between the room-seekers and the landlords. This causes a lot of miscommunications, consequently creating confusion and frustrations for both parties.

Similar to uploading verified background/credit checks, our landlords would be able to define their availability beforehand. The room-seekers would then be able to view and choose from the defined availabilities. This decreases the number of back and forth communication. It also minimizes the chances of error from ambiguous deliveries.

These key features we offer are unique to our app. We tackled and solved many frustrations and pain points room-seekers currently experience. Our goal is to enhance their overall experience in finding the right room/mates, and building better lives for them.

Embedded prototype of ELLY

To build a desirable product for our users, we tested and iterated our designs with potential users. We tackled the repeated stumbling stones identified by multiple users. We wanted to improve their experience interacting with Elly.

Iteration — defining your budget


The on-boarding experience sets the mood(atmosphere) of the app, creating expectations. It has to be simple and easy to use. Users should not experience difficulty and frustration, leading them to leave our app.

Our app has to be friendly and approachable. We wanted the interaction to be unique. We wanted to stay away from long boring forms.

We designed a few versions of how the “defining your budget” page would be. A common similarity across all the versions was the predefined budget range. We wanted our users to be able to choose from a selection, rather than trying to recall their own range.

Yet, after a few testings, we quickly discovered a problem. Many users had trouble choosing a predefined range. Some may have really specific budgets, and our range was too big. Some were more flexible and their range falls into two or more selections.

With that feedback, we decided to allow our users to define their own range. It allows the users’ preference to be more specific, thus allowing our system to find matches that have higher compatibility.

Iteration — information shown on cards


After the system has filtered out matches, the results would be presented to the users in card form. Each card would show a brief summary of a match. If they are interested and want to know more, they can access the details by proceeding to the specs page.

Yet, our users should have the option to decide of whether not they need more information.

Iteration — review versus verification


— establishing sense of trust & safety

Vancouver is a diverse city. The city attracts people from all over the world. Many users interviewed have mentioned using community platforms such as Vansky (Chinese) and Tonarigumi (Japanese). Language can be a major barrier. Ease of communication and cultural familiarity are primary influencers.

Elly is an app solving the problems of room/mates finding specifically for the renting community in Vancouver. Options (eg. other than English) for the system’s display language and communication tools could definitely be improved to accommodate to this large population of potential users.