Distro 2.0 - Web Platform

UnitedMasters distributes music for independent artists to DSPs globally. The web distribution flow is the primary entry point - artists submit releases through a multi-step form covering metadata, artwork, pricing, and platform selection.

The existing flow had accumulated years of incremental additions without structural redesign. Error rates were high, rejection rates created operational burden, and the metadata pipeline downstream was noisy.

Submission completion sat at 11.87%. Artists would start a release and abandon mid-flow due to unclear requirements, confusing validation, or inability to recover from errors. Rejected submissions created manual ops overhead.

The form architecture itself was the issue - steps weren't ordered by dependency, required fields weren't surfaced at the right time, and error messages were generic. Artists couldn't distinguish between blocking issues and optional improvements.

The new flow had to support the same metadata schema required by DSPs - no simplification of required fields was possible. Backward compatibility with existing releases in draft state was required.

Engineering capacity was limited to one sprint cycle for the initial build. The design system was nascent - this project would establish foundational patterns for form components, validation states, and multi-step navigation.

Reordered steps by dependency chain: identity → content → distribution → review. Each step validates before advancing, with clear blocking vs. warning distinction. Added inline validation with specific remediation text.

Designed a persistent progress indicator showing completion state per step. Added draft auto-save with recovery. Built the form component library (inputs, selectors, date pickers, multi-select) as reusable design system primitives.

Worked with data engineering to map every field to its downstream consumer, then used that mapping to write contextual help text explaining why each field matters - connecting artist action to platform outcome.

Submission completion increased from 11.87% to 25.0% - a 2.1× improvement. Subscription churn dropped 55% as artists who submitted through the new flow were significantly less likely to cancel, suggesting the improved experience reinforced the value of their membership.

Metadata quality improved at the point of entry, reducing ops escalations for fixable errors. The component library built for this project became the foundation for subsequent web features. The validation pattern (blocking vs. warning with remediation) was adopted across Content ID, Waterfall Releases, and internal tools.

Phase 2 focuses on auto-approval logic using fingerprint data, artist flags, and membership tier - reducing ops review for high-confidence submissions. Expanded metadata fields (producer credits, lyrics, Canvas video) are in planning.