Introduction

In this case study, we will explore how our solution, the Social Media Ad Optimizer, helped an American start-up automate the implementation of advertising campaigns on various platforms while ensuring precise targeting and effective budget management. The client was seeking a solution based on large data sets that would enable quick processing of advertising offers and provide full traceability of spent funds.

Client Profile

The client is an American start-up focused on digital advertising. They were looking for an innovative solution that would automate the implementation of advertising campaigns across multiple platforms, including the internet, smart TVs, taxis, and cash registers. Their goal was to reach diverse customer segments, from small and medium-sized enterprises (SMEs) to corporations, with granular targeting capabilities.

Challenge

The main challenge was to design an architecture that could process large data sets in real-time, with response times below 30 milliseconds. The solution needed to support efficient advertising campaign management, precise target group definition, and comprehensive spending tracking. Additionally, it had to be scalable, cost-effective, and capable of handling data stored in various formats and locations.

Solution

Our solution, the Social Media Ad Optimizer, comprised several modules designed to address the client’s challenges effectively:

  1. Client Application: We developed a client application using React and Redux, providing an intuitive interface for managing and launching advertising campaigns. This application enabled users to define target groups, select advertising platforms, and track campaign performance.
  2. Backend Infrastructure: The backend of the client application was built using Scala, Akka-http, and Akka-streams, ensuring efficient data processing and seamless communication between modules.
  3. Analytical Module: We implemented an analytical module using Spark, Hadoop, Hive, and Athena, enabling the processing of large data sets and generating actionable insights for campaign optimization.
  4. Decentralized “Pixel-server”: We created a “Pixel-server” using Scala, Akka-http, and Redis, which served as a tool for real-time monitoring of ad displays. This module provided valuable data on ad impressions and engagement.

The solution utilized a microservice architecture, allowing for scalability, flexibility, and easy integration with other systems. We implemented industry-standard practices such as code review, unit tests, and automated tests using tools like Jenkins, Liquibase, and Docker. The entire platform was deployed in the AWS cloud, leveraging geographic regions for efficient resource management and continuous monitoring.

Results

The implementation of our Social Media Ad Optimizer provided the client with a cloud-based solution that empowered them to conduct multi-platform advertising campaigns independently. They gained full control over their campaigns, precise targeting capabilities, and comprehensive metrics for measuring campaign effectiveness. With the ability to precisely define target audiences and monitor ad performance in real-time, the client could ensure that their advertising budget was optimized and not wasted. Additionally, the solution was made available as a B2B “white-label” product, enabling the client to offer it to other businesses seeking an efficient and effective advertising solution.

Conclusion

Our Social Media Ad Optimizer empowered the client, an American start-up, to automate the implementation of advertising campaigns across multiple platforms. With precise targeting, efficient campaign management, and comprehensive performance tracking, the client achieved better control over their advertising efforts. By utilizing a scalable and flexible architecture, coupled with advanced analytics, the solution provided actionable insights for campaign optimization. The client gained a competitive edge in the digital advertising industry and expanded their offerings by making the solution available to other businesses as a white-label product.