Increasing screening filter engagement and eliminating support friction

Recruiters were spending most of their day "fighting" with the screening screen: to find the right candidates, they had to navigate through huge lists, filter everything manually, and remember configurations in their heads.

Because of this, filters were rarely used, search barely helped, and the screening flow didn't make it clear what the next step should be, resulting in slow, repetitive, and inconsistent screenings among recruiters, increasing support tickets with questions about how search and filters worked.

Based on usage data and interviews, I mapped this daily routine, identified key friction points, and redesigned the screening screen as a true decision dashboard: contextual search aligned with recruiter vocabulary, smarter and more predictable filters, cards highlighting only critical information, and clear candidate states.

After the redesign, search and filters became part of the real screening routine: there was a 73.8% increase in search usage, 14x more filter usage, and a reduction of about 26% in average screening time per session, making the process faster, predictable, and sustainable for the recruitment team.

Beyond the numbers, recruiters reported that it became much easier to understand what stage each candidate is in and to set up appropriate filters for each job.

Design process

GDPR and Compliance

During ideation, we conducted an audit of filter options, eliminating color criteria and adapting gender identity options to mitigate legal risks for clients and promote a selection process more focused on skills, reducing unconscious biases in screening.

Design System

The Handoff of this functionality was done with the usability guidelines and UI standards of the new Design System we were implementing. Through integration of Figma MCP with IDEs and a shared design tokens file, it was possible to further accelerate development and maintain platform design consistency.

The technical complexity remains,
but now behind the scenes.

Result: 73.8% increase in search usage.

Result: filters received 14x more interactions.

Search Field

Filter Chip with Counter

New Candidate Cards

Despite potential value, we decided not to include the AI filter suggestion function in the first release.

Usage estimates indicated high token consumption, generating a recurring cost incompatible with the product margin at that time.

We opted to prioritize the redesign of manual search and filters, solutions with low operational cost and high immediate impact, leaving AI automation planned as a roadmap evolution.

Results showed a clear impact: 73.80% increase in search component usage and greater perceived platform value, proving improvements made functionality more useful and relevant for recruiters.

Besides search, screening filters also saw a major leap: usage of this essential tool for effective screening went from 4 to 57 clicks in the same period — over 14x more interactions.

This growth shows the new flow made filters more accessible and relevant in recruiters' daily lives, helping decision-making with more speed and precision.

During discovery of this feature, we identified that talent pool filters also needed redesign: the complexity of these advanced filters hindered usage causing usability issues. So for next sprints we should work on Talent Pool screens and flows.

Also, we conceived an AI automatic filter version, but opted not to include it in the first release due to high token costs, which would make the solution unviable today. This resource remains on the roadmap to be re-evaluated when cost per use is more sustainable.