Business Situation
- A leading American mass media and entertainment conglomerate faced significant challenges, including fragmented data sources, manual processing, high infrastructure costs, and inconsistent insights that impacted ad sales forecasting and revenue goals.
- The conglomerate had structured and unstructured data, including first-party data from internal sources and third-party data from data providers such as Nielsen, which needed appropriate integration into one unified source.
SGA Approach
Technology
- Implemented a robust data architecture with an enterprise-wide data lake with AWS and Snowflake
- Leveraged Databricks for advanced analytics and processing
AI
- Developed AI models for enhanced competitive programming analysis
- Implemented AI-driven master data management (MDM) architecture for improved data accuracy and introduced AI-powered tagging for shows using reinforcement learning mechanisms.
Data
- Consolidated diverse data sources into a unified architecture and built a scalable enterprise-wide data lake as a single source of truth for analytics needs
Key Takeways
- Established a unified data architecture, enabling comprehensive audience and channel segmentation, facilitating $1M in annual infrastructure cost savings
- Empowered ad sales teams with automated insights, improving productivity and forecasting accuracy, achieving a 20x reduction in the time-to-insight for ad sales teams
- Enhanced competitive programming analysis and audience demographics understanding using AI models with a 50% decrease in data science model run time and a 15% efficiency gain
