Vevo selected Aqfer’s Marketing Data Platform as the basis for its MELODY big data analytics platform, which was jointly developed by the Aqfer team and Prohaska Consulting.
About the Client
Vevo is the world’s largest all-premium music video provider, offering artists a global platform with enormous scale through its distribution partners. Vevo connects artists with their audience globally via music videos and original content, working directly with them to find unique ways to bring their music to life.
Challenges
In 2018, an influx of 18+ million new subscribers to and 25+ million additional monthly global views on the Vevo YouTube channel led to an exponential increase in inventory demand (resulting in $600 million in revnue). But Vevo’s publisher lacked an advanced and scalable data platform to fully optimize growth and take advantage of its recent successes. The team faced mounting risk from inefficiencies in its infrastructure.
Vevo’s reliance on inflexible and outdated systems and processes:
- delayed verification of premium inventory via Comscore and Nielsen audience data
- required numerous resource-intensive manual resources
- made them heavily-dependent on expensive, limited scale registration data
- caused ad server analytics to be housed in disparate systems, complicating target audience authentication
Vevo recognized that in order to continue to grow its business and productivity it needed a dramatically more advanced solution. The team was determined to invest in a data platform that provided them with the flexibility to data mine and that was scalable for future performance and profitability.
Rockin’ with AWS: Amplifying Video and Music Streams for Chart-Topping Performance
Scalable Data Lakehouse
AWS Simple Storage Service (S3)
Used to onboard various types of data, we store impression data and order metadata from first-party and third-party sources such as GAM, Freewheel, Nielsen, and Operative Staq using enterprise connectors that run on schedule to ingest data.
This data also follows the ETL process to adhere to Aqfer standard schemas to make data collation across partners simpler.
Intuitive Query Interface
AWS Athena
The data in S3 is further collated and summarized for granular access based on predefined parameterized filters. Allowing access to delivering and completed orders with accurate data across providers. These are possible through a complex CTAS query that help maintain an upto date report of ongoing orders.
High Performance CDN
AWS CloudFront
A custom application built to handle specific business requirements, as well as access custom access control is hosted via CloudFront, this application embeds in the Aqfer Portal and allows Vevo stakeholders to access their recent data by running SQL queries on top of the daily summary report. The MELODY application makes it easier for non-technical users to query data and pull historical as well as ongoing delivery data for orders.
Scalable Serverless Compute
AWS Lambda
Primary communication from UI to other AWS as well as Aqfer services is handled via a thin layer of APIs built by leveraging Lambdas to be cost-efficient at scale. These are also used to populate up-to-the-minute impression data from both GAM and FW into the MELODY UI every time a user queries the daily report.
Workflow Management
AWS Step Functions (State Machines)
Step functions run complex business workflows that relate to a separate tool in the application that deals with booking inventory. These make it possible to monitor granular steps and allow for long-running workflows.
Fast and Powerful Big Data Processing
AWS Elastic MapReduce (EMR)
A central part of the application is generating recommendations based on delivery data from first-party and third-party sources. EMR helps generate these recommendations based on a decided business logic across various ad servers.
Results
Vevo turned to Prohaska Consulting – experts in MarTech and AdTech strategy and execution – to find a solution that would best fit its needs. Together, they determined the company needed a scalable big data platform that could drive its first-party machine-learning strategy. After an extensive search, Prohaska advised Vevo that Aqfer’s Marketing Data Platform was the solution best suited to help the company achieve its goals and overcome its infrastructure inefficiencies.
Working jointly, Aqfer and Prohaska developed and delivered MELODY, an advanced, flexible, and scalable first-party data analytics platform custom-built for Vevo to help it dramatically increase premium inventory, improve productivity, and grow its revenue. MELODY allowed Vevo to aggregate data from sources including Comscore, DFP, Nielsen, Operative, and YouTube into a single platform and apply machine learning processes to perform simulations and optimize directions based on its desired goals. Built-in audience and campaign dashboards within MELODY are integrated with its machine learning tools and direct optimizations from the dashboards to Vevo’s ad server of choice.
Key Metrics
Premium Inventory Increase
Productivity Improvement
Revenue Growth
“The custom-developed MELODY platform improved productivity, enhanced our brand, and gave us a longterm, first-party proprietary data mining platform that will drive business for years. This next-generation solution is transformative for our business.”