How to precisely target the television audience? How to plan and execute an effective campaign and measure its performance? Such questions and concerns arise with regularity in many organisations. Evojam provided technical support to help a New York–based ad-tech company address these critical issues using a data-driven approach.
Data-Driven TV Ad Targeting and Optimization
In the client’s words:
“We offer a complete suite of advanced, data-driven measurement and analytics solutions based upon one of the industry’s largest matchable TV datasets.”
The unique suite of services offered by our client includes:
Audience Analytics
Campaign Planning & Optimization
Measurement & Attribution
“Viewership data from more than 30 million devices across 21M+ households. More than 5,000 consumer attributes to pinpoint the most targetable audiences.”
Custom Data Processing Engine
The datasets used for the project are covering more than 21 million households. Analytics on such a level requires advanced data processing, so our client developed its custom data processing platform in-house. It was designed to match viewership data from multiple providers and ad networks.
The data processing pipeline was built on top of Spark cluster-computing engine to process 2TB of data monthly. It is a cost-effective and versatile storage for 40TB of data that is being used by data scientists for analytics and machine learning.
High-Quality Software Development Wanted
Having achieved market success, our client needed to speed up the development process. The skill set and capacity of their existing team proved insufficient to reach this goal, so the organization decided to contract outside assistance. It was not an easy task to find the right technology company due to the demanding nature of the project. What was essential to succeed and thus expected from the external vendor were both rare and unique skills of tech specialists and high-quality service.
The Launch of Cooperation with Evojam
Because of these special requirements, the company decided to reach out to Evojam. We started the cooperation by developing data extraction-transformation-loading pipelines (ETL) in Scala.
As development progressed, the collaboration extended to a wide range of domains. Finally, the Evojam team devoted to this project took on the following tasks:
infrastructure configuration and maintenance (DevOps)
frontend development services in Angular
project management
advising on data-lake repository design, tech stack, and implementation.
The solutions that Evojam delivers have allowed to combine raw data from multiple sources into a rich database. The information regarding the audience and their viewing preferences, such as what and when they are watching, is the input for machine learning algorithms and ad recommendations.
Techstack: Apache Spark, Scala, Angular, AWS S3, AWS EMR, AWS Redshift, AWS CloudFormation, Apache Airflow, AWS Glue, Apache Hive, Apache Zeppelin
Project Management Consulting
With growing teams on both sides of the partnership, our services have expanded beyond software development. Currently, Evojam shares expertise in managing complex software projects as well. We help to streamline the project management processes of our customer by setting up the project structure in Atlassian JIRA. In addition, we advise on the software development process and data architecture.
“After hiring Evojam, we had a 400% increase in our deliverable dates.”
