As Qubole Data Service is a source of strategic advantage, this client has asked that its name not be used.
An award-winning, full-service digital advertising agency based in the United States. The digital advertiser leverages Big Data to help its more than 3,000 clients make smarter decisions.
Evolving digital advertising dictates increasing levels of personalization designed to scale to millions of users. To remain competitive, it needed to deploy advanced contextual and behavioral targeting and micro-segmentation to help its clients boost the return on investment of their online efforts.
Architectural Overview: The advertiser was relying on a traditional data warehouse to deliver insights based on historical consumer data such as purchase history, campaign responses, and customer profiles. However, the growth in the importance of online behavioral data such as user clicks, page visits, contextual activity, and social and mobile data led to Big Data requirements that were well beyond the capabilities of its data warehouse.
Deeper personalization required integration of data from a complex digital advertising ecosystem, including AppNexus for advertising management, Google Analytics, third-party data suppliers such as Axciom and Experian, offline CRM systems such as salesforce.com, social media profile data, and mobile data streams.
Although the digital advertiser recognized that the Hadoop platform would provide the ability to meet the high-volume, velocity, and variety of Big Data required for digital advertising, it found that Hadoop was difficult to use and had trouble getting started. In particular, the company could not build a data pipeline, the steps going from data acquisition to transforming data into Hadoop to production deployment of algorithms, in a repeatable and consistent way.
The digital advertiser estimated that its internal implementations of Hadoop would require millions of dollars in upfront investment. Failure placed not only this investment at risk, but the advertiser was concerned that being unable to deliver on business objectives in a timely manner would be even costlier.
QDS tackled the digital advertiser’s tough structured and unstructured data integration challenges using its pre-built connectors for on-premise, cloud and social media data sources. No changes to the digital advertiser’s on-premise infrastructure were required.
Qubole provided a managed Big Data service that made it simple for the digital advertiser to prepare, integrate and explore Big Data in the cloud. Using QDS, the advertising could turn its attention to its business requirements rather than having to deal with the complexities of Hadoop, managing clusters, or scaling its infrastructure. The QDS Amazon S3 loader and Sqoop-as-a-Service delivered an automated and reusable data pipeline.
Not only did the digital advertiser deliver its project in record-breaking time, but it also saved a few million dollars of upfront investment by going with QDS. The company found that its monthly could bill was half the price of other options.
With QDS, the digital advertiser has been able to use Big Data insights derived from QDS to:
“We realized that we had to implement Hadoop quickly to keep up with digital advertising trends, but found the platform overwhelming. Qubole Data Service helped us get up and running quickly so that we could deliver the micro-segmentation, contextual and behavioral targeting and optimized campaign performance our clients were demanding.”
Free access to Qubole for 30 days to build data pipelines, bring machine learning to production, and analyze any data type from any data source.