RTB Ad Exchange Platform Development To Manage Digital Assets

Gone are the times when media owners were happy with a 50% fill rate for their digital assets. Today, they would be interested in leveraging the other half as well. The starting point for the same would be to explore alternate approaches/channels that help maximize bookings and optimize fill rates. And this has led to the increased adoption of open bidding & auction platforms that enable media buyers to list & sell their digital assets faster!

Learn how we assisted a UK-based global DOOH agency with Real-Time Bidding (RTB) ad exchange system development to monetize their digital inventory & maximize efficiency.

Project Overview

A Europe-based DOOH advertising agency operating in more than 15 countries wanted to build an exchange platform that enables real-time bidding for optimal usage of its ad inventory & increasing profitability. They were using an inventory booking solution to pre-book their digital inventories based on ad requirements but were unable to leverage the locations that drove fewer bookings, resulting in revenue loss.


  • No in-house technical resources to execute seamless RTB platform integration for digital ad bidding
  • Difficulty in tracking where & when the ads are showing
  • Inability to handle multiple requests for different order executions simultaneously
  • Inefficient tracking of unutilized ad inventory
  • Inability to remarket ad inventory to drive more sales
  • Capacity constraints for processing data in real-time & storing it in a data lake


We developed an Exchange Platform that integrates an RTB system that allows multiple media owners to bid on ad inventory, with the winning ad being displayed on a digital screen.

Following the micro-services architecture, our team developed the following three components

  • Product-Adapter: Fetches inventory listings from different supply-side platforms (SSPs) and converts them for processing within the trading engine. It also notifies the SSP platform about the final status (when ad inventories are sold) once the auction is completed.
  • Trading-Engine: Retrieves processed data from “product-adapter” to optimize inventory listing based on various factors and publishes the “Exchange Supply Message” for the auction stream on Kafka. Further, it consumes “Exchange Demand Message” to prepare the final sales data for the SSP platforms and share it back with the “product adapter”.
  • Auction-Stream: Consumes “Exchange Supply Message” and sends the bid request to the demand-side platform (DSP). Then based on the auction, decides the winner, and alerts the DSP with a win or lose notification. It also publishes the sales data in the “Exchange Demand Message” for the trading engine.

The data team utilizes Exchange Supply Message & Exchange Demand Message and finally aggregates them in the data lake. From here the data can be queried from an SSP interface to generate reports that can be viewed by media owners.


  • ~30 million digital auctions for OOH ads under 30 min with the new platform
  • 50% reduction in latency with high execution times
  • 100% utilization of available ad assets with hyper-precise targeting by location