Media & Entertainment · Case study 09 of 12

Media Platform

A regional broadcast media network needed to replace a legacy scheduling system that required 3 full-time coordinators to operate. The new platform had to integrate AI-driven schedule optimisation with ad sales forecasting — two functions that had never shared data before.

Media Platform — Media & Entertainment interface design by Yellowfirst
Client
Media Network
Year
2022
Industry
Media & Entertainment
Role
AI Product Design · Scheduling UX · Contextual Architecture

The challenge

Content scheduling and ad sales were run by separate teams using separate tools with a weekly sync meeting as the only integration point. Schedulers couldn't see ad inventory availability; sales teams couldn't see content commitments. This created a recurring pattern of conflicting bookings discovered at broadcast time. The legacy scheduling system had no concept of audience data — prime slots were assigned by seniority, not by viewership intelligence.

The solution

We designed an AI-enabled scheduling canvas that shows content, ad inventory, and audience data in a unified timeline. The ML model suggests optimal slot placements based on historical viewership patterns and current ad demand signals. Sales and scheduling see the same calendar in real time. Contextual data architecture means every content item carries its full performance history, and every ad placement shows its predicted reach — eliminating the guesswork that caused conflicts.

Our process

  1. 01 Workflow Ethnography

    Shadowed scheduling and sales teams for 2 weeks. Documented 23 recurring conflict scenarios and their downstream costs.

  2. 02 Unified Data Model

    Designed a contextual data architecture that associates content items with audience segments, ad categories, and historical performance.

  3. 03 AI Scheduling Surface

    Designed the UX for ML suggestions: confidence-ranked slot recommendations with reasoning visible and overridable.

  4. 04 Calendar Canvas

    Built a high-density scheduling canvas supporting drag-and-drop across 6 channels with conflict detection in real time.

  5. 05 Sales Integration

    Designed the shared inventory view so sales and scheduling see live availability with soft/hard lock states for in-progress negotiations.

Outcomes

  • 3→0.5FTE coordinators required to operate
  • 94%Reduction in scheduling conflicts at broadcast
  • 22%Increase in prime slot ad revenue
  • Real-timeSales and scheduling data synchronisation

Technology

  • React
  • Python ML APIs
  • GraphQL
  • AWS
  • Elasticsearch