Empowering brands to create targeted TV advertising

Key Challenges

Our client dealing with large volumes of data from multiple sources required a tool to ensure high performance and speedy querying.

Key Details

Providing data to drive targeted advertising across TV

Our client’s company, founded in 2005 have been empowering businesses with targeted TV brand advertising for almost two decades. Their end clients range from marketing agencies to media owners looking to create target TV ads for their various audience segments. Our client collects a wide range of data TV data about audiences across cable, digital and more. They have a network of more than 90 cable networks, 200 MVPDs and 1,100 broadcast stations, so they deal with a large volume of data.   

Business challenge

Our client were collecting large volumes of data from a variety of sources which presented issues with performance, data accuracy and data cleanliness. They required a high-performance data solution that would effectively collate, process and clean their data for speedy querying for their data analytics and reporting purposes.

Sophisticated data indexing tool

Our highly skilled data engineers built a sophisticated indexing tool which essentially collated the data from third parties, cleaned it, ensured it passed the validation checks and processed it. The data was then fed into the client's reporting tool for them to gain insights from clean, processed data. A user-friendly User Interface was created in line with the client’s brand to produce a number of data reports including panel summaries looking at metrics such as the number of TVs and households.

The data processing was carried out on an AWS-driven infrastructure, to automate the processing of the data for the client with serverless architecture without manual input. Based on our recommendations, for this project and the data at hand, AWS was the technology of choice. AWS technologies are exceptional when it comes to automating scaling and dealing with big data.

Our data engineering team faced some technical challenges for this project due to the scale and the type of data. However, our engineers have over a decade’s experience in delivering excellence through our data engineering solutions, so they were able to identify and resolve these challenges efficiently. These are the challenges they faced:

• Processing variety of data formats – They were able to overcome this by setting up cleansing rules within the architecture.
• Lack of consistency in data deliveries – As data came in from multiple sources, it came in at inconsistent timings. Our data engineers were able to set triggers to start the processing on an hourly basis.
• Repetitive data – There was a lot of repetitive data that was coming in from the third-party sources, they were able to avoid duplicate data by setting up rules to ‘ignore’ this data.
• Lengthy query time – We employed two databased systems, Aurora and Athena. Athena was used for time-consuming queries whereas Aurora was used for quicker queries to speed up the query process.

Efficient, automated data processing

We worked with our client on a project basis, we built and delivered the infrastructure with a year of post-project support to ensure that the solution was working well for the client. With the indexing tool, our client were able to improve efficiency with automated processing of the variety of data coming in from several sources. This enabled them to offer their end client a targeted way to approach their audience segments.

If you are looking for expert engineers to provide secure, scalable and robust data pipeline development service, get in touch to find out how the Ardent team can help.