Experiment filters

Redesigned Amplitude Experiments' top-level filters to help users more easily discover and act on important experiments

Challenges

  • The filters lacked affordance, leaving users unsure how and why to engage

  • Didn't match filtering patterns in Amplitude, creating an unfamiliar experience

  • Were cluttered and lacked visual hierarchy, making results difficult to scan

Process

  • Analyzed filter adoption, including repeat usage and related behaviors, to understand use cases and discover design opportunities

  • Led a critique to establish shared context, present potential directions, and capture feedback

  • Scoped the work to a focused change, while identifying and documenting ideas for future iterations

Solution

  • Introduced interaction patterns and states to clarify affordance

  • Aligned with a filtering pattern in Amplitude and applied design system variables for consistency

  • Removed underutilized and confusing elements to improve scannability and increase vertical space for results

Impact

  • Increased weekly top-level filter usage by 50%

  • Increased individual filter usage for "In Progress" (2x), "Completed" (3x), and "Reached Stat Sig" (+70%)

  • Held custom filter usage steady, confirming the redesign didn't disrupt existing workflows

Before

After