Airbnb
In-Home Accessibility
The IHA team working to make travel accessible for everyone
Overview
TIMELINE
June – September 2019
CONTEXT
The In-Home Accessibility team’s mission is to make travel accessible for everyone – enabling and empowering our hosts & guests with disabilities. The Accessibility Evaluation Tool is an important asset in scaling accessible home listings by 1) increasing the accuracy of the accessible features being audited, and 2) creating Airbnb’s own internal tool to evaluate the listings and recruiting qualified evaluators from the people with disabilities community.
ROLE
I worked on the end-to-end experience for the pilot program, working in sync with cross-functional partners throughout to collaborate on product strategy. I ensured a web-responsive design and seamless handoff for implementation as well as supporting the UX research during testing.
Getting started screen of the pilot guide
Why do we need an evaluation tool?
GOALS
The launch of the pilot program for the Evaluation tool was vital to the success of ensuring more availability of accessible homes. It was one of the biggest projects for 2019 on the In-Home Accessibility (IHA) team.
✧ Build trust on the Airbnb platform
✧ Reach out to the disability community
✧ Increase accessible listing accuracy from 80% to 100%
CHALLENGES
Previously, Airbnb has been using a 3rd party company to tag all accessible homes that are listed on Airbnb. However, with the lack of accuracy and the need to scale, it was necessary to take the steps to create our own platform.
✧ Developing the tool on 3rd party software
✧ Inconsistent accessible feature definition and training
✧ Calibrating the nuances of accessibility
Going from 0 to 1 designs – key research
INSIGHTS
Working closely with the UX Researcher on the initial context-gathering research, the sessions revealed a core theme of how people tend to assess based on their specific personal needs as well as relying on their assumptions of navigating a home. This led to the next question of how quickly and efficiently will the evaluators be able to go through the evaluation of each home accurately with minimal bias.
More questions arose – What will the journey be like after the evaluation session ends? What is the sense of progress going to be like? How many photos will be shown at a time? All these considerations helped to form the prototypes to test with users in the next round of research.
For the initial research, we asked about their current experience with the guides we currently offer today like this page.
An example of some of the accessible features that the homes are evaluated on to be considered accessible
Making mental models tangible
INSIGHTS
The previous user research insights informed my perspective on testing the overall flow that users would prefer. 1) listing based (the whole home), 2) features based to understand what would be the most effective way of evaluating multiple homes efficiently.
One of the variations – viewing by features
Reframing through technical pivots
INSIGHTS
Originally, the pilot design goal was to have the platform of the tool be on Airbnb. However, there was a roadblock with the technical features that required us to move to a 3rd party platform, SageMaker. This gave way to another design overhaul to fit the constraints of the new software base.
The initial external UI without the visual branding (but following the user flow we aligned for the guide)
A more objective and focused assessment for the evaluators
INSIGHTS
While discussing with the team the next steps for the design and solving the challenge of creating an objective guideline, the checklist was born. Based on the previous guidelines the operations specialist gave to the evaluators, I incorporated and designed the checklist system.
To ensure that the evaluators had high certainty of their decision, it was important to consider the error states and how much friction this would to the experience.
By adding an additional element of interaction, the guide helps to emphasize to the evaluators on their final confirmation and checking with our own standards