
Uncovering bias through analytics.
Design Lead
Untapped's mission is to make the world more equitable by broadening access to transformative careers. I joined the Analytics and Sourcing Pod in February of 2021.
About Untapped
Getting started
I began by reading any information about the company and scheduled meetings with my product manager, and engineering team. I also reviewed all existing user research and competitive analysis that will help me understand the market.
The problem
I learned that traditionally, recruiters use self-reported information such as race, ethnicity, and education to examine candidates. The top educational institutions tend to be preferred because of the skills developed through their resources. However, not all great candidates have access to such resources, this creates a gap in the hiring process where unseen bias can start to appear.
The solution
Through analytics, we can examine self-reported data, compare it to job-hiring industry benchmarks, highlight any discrepancies and provide actions to fix those discrepancies.
Discovery
My discovery phase continued with a brainstorming exercise. The goal was to share, vote, and discuss ideas in 1 hour. I also scheduled calls with a few existing clients to gather product feedback.
Feedback:
Broken usability was frustrating.
Unclear insights made it hard for clients to meet their hiring goals.
A disconnected experience between Analytics and Sourcing didn’t provide data-backed actions.
Ideas:
Explore merging Analytics and Sourcing into a single experience.
Design a recruiter to recruiting manager experience.
Providing value through insights and building a relationship with Sourcing.
Definition
We aligned in providing value by building a relationship with Sourcing. I put together a few concepts to quickly start testing for usability, validation, etc.
Initial graph concept exploration.
I got a chance to dive deeper into the recruiter experience and discovered the following while testing:
Companies want to hire top diverse talent but finding that talent is extremely hard.
It is very hard to see an overview of each job and how candidates are moving through each step, specifically drop-offs, warm candidates, and passthrough rates.
Companies don't know how they're doing in hiring diverse talent because there are no guidelines or information they can reference.
User walkthroughs and interviews
Development
Testing went well and it was clear what worked and what didn’t so during the next few weeks I got a chance to refine ideas, get input during design critique, feedback from clients, and share progress with the rest of the company.
"That button (View Candidate Details) was very clear... If I'm able to click this (candidate details pagination) I assume I'll be able to see who the other candidates are"
- Vivan
"These types of graphs are great but can become more complex depending on the workflow."
- Stefan
Research Synthesis was presented at a brown bag lunch.
After validating some concepts and hypotheses, I continue to refine designs.
Delivery
A proper roadmap, early scoping and research findings allowed me to put together a proper handoff for engineers to start working on the first version of this massive update.
Quality Assurance and Shipping
Engineering set up a dev environment for us to test and Q.A. new features. While this was happening, I worked with marketing and customer success on a release plan that involved blog posts, email campaigns, and training.
We kept an ongoing QA document which was split into JIRA tickets.
Designing for accessibility
Making the world more equitable through design means utilizing methods in the design process that allow people of all abilities and backgrounds to use technology. Inclusive design is something that was kept in mind.
Color
I spent a couple of weeks developing and testing color schemes for color vision deficiency.
Patterns
I created patterns to help identify sections of information regardless of the color scheme.
Separation
I created rules to help re-enforce patterns and color schemes, which is particularly helpful with smaller pieces of information.