
Airbnb & Me
HCI+ Interaction Design + Prototyping User Interface + Usability Testing
CONTEXT
The INFO I 441 course, led by Marty Siegel, l collaborated closely with Airbnb in researching and redesigning their interface and how their users interact with their services. My team's goal was to design a unique companion that drives personalized experiences and invites new users into the Airbnb culture. We wanted to emphasize on the user journey because traveling can sometimes have pain points. To address the pain points of traveling, we decided to focus on Gen X users who are undecided travelers and are new to Airbnb. Further reasoning on our rationale for choosing this particular group can be found in the documentation. This project had a quick turnaround time of 2-3 weeks.
MY ROLE
My role in the team was to perform primary research through interviews and on-site observation, creating the screens, prototype and the hi fidelity screens, as well as write up and format our documentation. My team and I spent many hours fleshing out constraints and affordances in our redesign. Our redesign ended up placing top three in the class. The overall feedback from Airbnb was positive and our redesigns will hopefully prove helpful to their future iteration of their services and products.

For this project, we followed Jake Knapp's book, Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days. The project duration was about 3-4 weeks.

With our goal being to help new users figure out where they want to go and what they want to do, our team mapped out what we thought were main stake holders (Hosts, New Users, and Local Businesses) and mapped out the experiences and interaction between stake holders from when they start the trip to when they are on the trip.

Outlines what our client, Airbnb, is looking for. We had to break apart what the given tasks and figure out what they really wanted. Figuring out our problem space, my team set a goal which we referred back to constantly to make sure we were on task. We mapped out the Airbnb interactions and expanded on how people, did primary (interviews with travel agents) and secondary research, as well as took general notes on our performance as a team.

This picture showed our final stage of our brainstorming. Here, I am reiterating through each process of the interaction and the rationale behind each step of the user's interaction with the app to my team mate so that the team was on the same page. As the person making the screens, it was important to me to be able to reflect what everyone has agreed on.

Gamaliel Garcia, Reem Alturki, Andrew Faggin, Tasha Lim. (up to down) We got second place out of 7 teams! According to the Airbnb Feedback, we were trying to address one of the harder Life Cycle of Decision Making ( Undecided*, Had a destination, booking, already booked). They also asked for a clean documentation that "got to the point," and so another driving factor to our rank was how straight to the point our documentation was.
OUTCOME
In our research, we found that travel agents, welcome centers, as well as concierges all utilizes visual stimulation as their strategy to figuring out what their customers want to do. Using that insight, we designed a image-based quiz that would factor in qualities in the pictures that a New User chooses and produce suggestions based on those qualities. This process would ideally be backed by an algorithm. For a more details about our research, please refer to the documentation. Also check out our prototype.
To flesh out our design, we created a persona named John. John is an accountant in his mid 40s. He has a family and his children are at the age where they are either in college or they have their own family. He currently lives with his wife, but since their youngest child is about to graduate and already has a job offer, he is thinking about retirement. Many of his friends are traveling the world, but he is not quite sure where he would want to go or what he would want to do.



Since the travel agents we interviewed said that they asked clarifying questions in order to hone in on what experiences and locations a traveler would like to have, we designed a “quiz”, that would help us determine a user's (in this case, John's) preferences and be the centerpiece of our Airbnb & Me design. We also wanted to make the user aware that they can alter their choices at any time. The quiz is comprised of various images and the user can select as many images that interests them. We decided to use images instead of tags and text to cater to our travelers who do not know and cannot verbalize what they are looking for, but unconsciously know what they like. To cover all grounds, we have also added an area to put in tags for any other descriptions that our user might feel strongly about that the quiz may have not covered.


Once John is done with the quiz, using the data collected, an Airbnb algorithm will allow the home page to produce listings for homes and experiences that are catered specifically towards his preferences. Listings that cater more towards what John indicated in the quiz will appear closer to the top of the list. For an example, if he selected pictures related to the ocean or beaches, listings related with the ocean would appear closer to the top. However, John still has the capability of browsing the rest of the listings if he so chooses. This would suggest that listings appear in an ordered list, however it would beneficial for both hosts and users to have the listings slightly randomized so that John can get exposed to potential hidden gems and uncover new interests.


One comment that we received during our research on Airbnb was that users were not always successful in searching for experiences. The initial text in the search bar usually says "Try '[insert location name]'" and we saw that this could add stress for travelers who do not have a destination in mind yet. Therefore, we suggest that the search bar alternates between locations, experiences, and potentially restaurants. If John wants to check out a hot spring but does not know where he should go, Airbnb will be able to help him find all the hot spring experiences all around the world. Utilizing the data from John's quiz, the experiences will be listed based on what characteristics in an area he indicated.


Once John has found the experience he would like to have, he can easily be redirected to a page listing homes available in the area. Our team decided to make the location on the top of the experience page one of the ways a user can be redirected to a listing of the homes in the area of the experience. We also factored in some feedback we received during our usability test on how it would be easier if a list of homes in the area was also included at the bottom of the experience page. Having house listings available underneath the experience mimics to how a user might have multiple tabs open on a computer. At the moment, something like this exists for "Experiences others viewed in "[the location]," but we think adding a row of available homes would also be helpful for users who use Airbnb to organize an experience first and housing second.


After John has returned from his trip, he might not want to see the same types of suggestions he received when he first signed up for Airbnb. However, John can always take the quiz again by accessing his "edit profile" page. Just as you can edit your profile page at any given time, John will have access to update his preferences at any given time as well. In this way, John can choose what he is looking for in his next trip and Airbnb will be able to once again provide him a personalized experience because it will incorporate his new and previous preferences by using the powerful algorithm.
FEEDBACK & FUTURE ITERATIONS
One comment we got back from the people at Airbnb was that the documentation was as they had requested it: Straight to the point. They did not want to hear the nitty gritty of our research-- they wanted the insights from our research and how it would affect them. One of the feedback said that our design was "inspirational," however another feedback said that they would've like to see us connect our design and user journey with our insights and rationale more. This was a good lesson on balancing what goes in a documentation.
​
For future iterations I would like to revisit the questions we ask in the quiz. I feel like the three questions we have cover the three categories of Airbnb, however there is an art to how you ask questions. I also want to look into how the user would scroll through the images. My team and I played around with the idea of having the screen bleed off the screen to suggest the users could continuously scroll through the images, but there is probably a better way of carrying out this task. It would also be beneficial to see the first things to come to mind when seeing an image in order to help us understand how to interpret the data we're given.
​
Overall, working with Airbnb was a great experience and we hope that the suggestions from our class could help inspire future iterations of the Airbnb app.