Streamlining Ad Monetization with Prebid Server Deployment on AWS

Customer Problem

• More Omni-Channel content providers are connecting directly with ad platforms.
• Advantages of owning your own connections with ad platforms:
  • Transparent auction mechanisms
  • Ability to modify decision logic
  • Full ownership of log-level data for real-time optimization and decision-making
• Challenges faced by customers:
  • Difficulty in deploying and operating high-volume, low-latency prebid server infrastructure
  • Lack of experience in operating such workloads
  • Desire to build their own SSP (Supply-Side Platform)

Introduction to Prebid

• Prebid server is the most popular unified auction solution in the market.
• It is based on open RTB, runs a transparent auction, and is completely open-source.
• Prebid server is configurable to suit the needs of large and small companies.

Prebid Server Deployment on AWS

• Introducing a solution for deploying prebid server on AWS in a scalable, cost-efficient, and production-grade manner.
• Benefits of the solution:
  • Scalable to handle high-volume, low-latency workloads
  • Cost-effective deployment
  • Full control over decisioning logic
  • Ownership of transaction data
  • Simple integration with BI tools and AWS services
  • Open-source and community-driven codebase

Solution Architecture

1. Client-side (prebid.js) sends bid request to prebid server through CloudFront.
2. CloudFront and AWS WAF provide security and low latency.
3. Load balancer distributes requests across the ECS cluster hosting prebid server.
4. Prebid server conducts the auction, fans out bids, and returns the results to the client.
5. Prebid server metrics are stored in an S3 bucket, processed by Glue, and made available through Athena for reporting and analysis.
6. CloudWatch provides monitoring and alarms for the solution.

Getting Started

• Deploy the solution using pre-generated CloudFormation templates or the CDK project on GitHub.
• Configure prebid server with your bidder adapters and other settings.
• Integrate the client-side (prebid.js) and ad server with the deployed prebid server.

What's Next?

• Incorporating customer feedback to improve the solution.
• Adding support for video, AMP, and mobile ad formats.
• Developing a transactional data layer to enable real-time decisioning and optimization.
• Exploring traffic shaping and other advanced features to optimize ad monetization.