Mable
Mable is a healthcare marketplace platform that connects people from aged care and disability with various support workers. It allows independent support workers to offer different types of support services.
Problem
Many aged care and NDIS clients face significant challenges in effectively communicating their support needs. Despite spending considerable time navigating the platform, these individuals struggle to articulate their situations and lack clarity on how to best leverage their funding for achieving independence.
As a result, 80% of the users posted only 1 job in their lifetime. This was becoming a huge loss for the business.
My Role
-
Empathize with the target audience and understand different personas
-
Conduct explore and ideation workshops
-
Design high-fidelity UI designs in Figma
-
Create components & maintain the Mable design system
-
Follow WCAG 2.2 accessibility guidelines
-
Perform usability testing with real users
-
Follow data driven iterative design process
-
Collaborate with stakeholders & engineers
-
Measure success against key metrics
Human-centered lean design process
User Research
I conducted several user interviews and went through some pre-existing research documents to understand the user types and their pain points in the existing job posting process. I used Calendly to send interview invites, dovetail to store all the interviews, and Qualtrics to gather instant qualitative feedback. ​
Pain points:
- For some users it takes up to 3 hours to share their need which results in lot of frustrations and demotivations
- They had to repeat the information everytime they are posting a job
- The old process was lacking that human touch
User groups identification
User type 1: I am sure what support services I need
This user group wanted a process where they could share all details regarding what type of support they need without repeating their situation every time they post a job.
User type 2: I am unsure what support services will help me live an independent life
This user group wanted guidance on what support services can help them achieve their support goal without lot of thinking from their end.
Conduct ideation workshop
After having a clear idea of the user pain points, I conducted an ideation workshop with the team to brainstorm on different high level solutions.
High level solution
I worked on high-level concepts, i.e. to simplify the lengthy job post process by allowing users to write down their support goals or upload their support plan. Using AI we recommend a support plan, which users can fully adopt, modify or choose to create their own plan. I shared these concepts with stakeholders and gathered feedback.
MVP release (Beta version)
I collaborated with the PM and engineering team to divide the solution into different releases. For the beta version, I designed a solution where we are recommending different support plans with minimal customisation. We released it with a small group of users.
Beta version
Study Data
Study Data
I worked with the data analysts and was monitoring the user interactions through Datadog, Tableau and Amplitude. Based on the data I worked on various iterations to improve the experience of the Beta version then worked towards the final solution.
Final designs
After various iteration and releases, here are some final designs of the new job post process. I followed the Mable design system while creating prototypes.
Measure
Success
Our key metric was to increase the number of jobs per client in a single session. After multiple releases, now 70% of the clients who are coming into the flow are posting more than 1 job in a single session. Also, the form completion time has reduced from 1 -3 hours to just few minutes. We achieved this by following the human centered designing process and rapid iterations by closely monitoring the data and feedback.