Unmatch Survey
PRODUCT OVERVIEW
Indeed Hire is a recruiting agency within Indeed that handles sourcing, screening, and hiring on behalf of enterprise-level companies looking to improve their talent acquisition flows.
ROLE
UX/UI Designer
TOOLS
Figma, Optimal Workshop
DELIVERABLES
Card sort user research, UI designs
User problem + persona
Within Indeed Hire, users called Delivery Specialists are responsible for determining when a candidate is not a match for a job they interviewed for.
Delivery Specialists did not have way to record and assess why candidates were being “unmatched” from jobs, and expressed to our team that this was a major gap in their ability to tell their clients how to improve applicant match quality and decrease the number of candidates unmatched.
How might we structure the way we collect and access data that tells us why candidates are not a match for a job?
Process
Planning
When presented with the user problem, I first investigated what workarounds the team was currently using in order to record when candidates were not a match. Users were using a survey UI that was owned by another team that didn’t contain any relevant “unmatch” reasons for Indeed Hire. They were manually typing in notes for each of the 100s of candidates per day that were rejected from various jobs on Indeed, which was taking up the majority of their working hours.
We couldn’t update the existing survey UI because it was owned by another team, so I created a plan to design our own experience that could be easily updated, scaled, and managed by our team.
User Research
Focus group with users. Led group discussions with users to uncover all possible reasons why a candidate might not be a match for a job.
Card sort study. Worked with a content designer to design a study in Optimal Workshop using the list of 25 different “unmatch” reasons collected during focus groups. This helped us better understand how users categorize the items and guide the UI design.
Designs
Using the results from our card sort study, I designed an in-product survey that appeared when a user marks a candidate as “Not a match.”
Features added include:
5 top-level categories for improved readability + discoverability
An “other” category to catch common responses that may need to be added to the survey
Optional field to add notes to the candidate profile while completing the survey
“Not a match” candidate status and corresponding reason automatically added as a private note to the candidate’s profile
Prototype walkthrough
Screens
Click to expand:
Post-launch
After handing off the annotated Figma designs to our team of developers, who launched the new survey experience quickly using a combination of Indeed components and Qualtrics survey capabilities for data collection, I continued to keep a pulse on user feedback.
Next steps:
Monitor for commonly submitted “unmatch” reasons in the Other category and consider adding them into the survey
Design and launch data visualizations that show Unmatch survey response data, with ability to segment by job title, location, client, and date.