The existing homepage. Upon login, job seekers were sent to Communities, a part of the product that most didn’t know how or why to use and usually didn’t encourage applications, a closely-watched business metric. The page they landed on needed to provide immediate value to them and ideally our partner companies.
Defining the new homepage. I hosted a brainstorm with our product managers and designers to determine the goals of the new page. We voted on which would bring the most user and business value and cut out any that would add significant engineering time.
Focusing on profile completion. We decided a profile completion “helper” is the highest priority item on the page because having a more complete profile ranks job seekers’ profiles higher in recruiters’ search results. Asking this amount of information from users generally isn’t desirable, so we took inspiration from 23andMe’s research questions to ask quick, 1-by-1 questions, which makes filling out information more enjoyable.
Pulling valuable elements together for a full page. The final page pulled in that profile completion helper, featured jobs, SMS opt-in, an ultra-light application tracker, refer-a-friend module, and a list of RSVPed events. With limiting engineering bandwidth, we reused existing components, like our job cards, as much as possible. After launching, we continuously evaluated the success of this page and each section within it.
Post-launch improvements to profile completion. After launching, profile completion scores increased from an average of 56% to 92% for active users. However, due to engineering bandwidth, we didn’t have time to create a profile completion helper for every field on the profile page. This led to user confusion; our support tickets of “I don’t know how to get 100% profile completion increased.” A few months after launching, we introduced levels of profile completion, explained what they meant, and showed users how to achieve 100%.
Summary
This project was an anomaly. While we usually have a more defined hypothesis going into a project, the openness of “Let’s give job seekers a more useful landing page” allowed us to test a variety of new hypotheses.
Ultimately, this new page was a huge success and exists (with some enhancements) today. It increased profile completion by nearly 40%, validated that featured jobs brought in significantly more applications, and proved our hypothesis that job seekers wanted some version of application tracking.
