TubeMogul Teams Up With Its Big Data

Print Friendly, PDF & Email

BigData use caseIn organizations today, more teams than just IT are learning to access their big data and use it for their own purposes and projects.

TubeMogul, which created an enterprise software platform that brands and advertisers use to buy TV and digital video advertising in real-time, was facing the challenge of making data easily accessible to non-engineering groups and cutting down the engineering resources that their fixed clusters and Amazon EMR required. Using Qubole Data Service (QDS), a big data analytics platform, TubeMogul made data easily accessible to non-engineering groups across the organization and reduced the amount of engineering time necessary to accommodate users and manage clusters. As a result, they began processing twice as much data in half the time.

According to Gartner, CMOs and other marketing leaders often ask analytics teams for deeper cross-channel insights spanning a wider range of marketing and advertising activities. Such marketers understand that they can use this data for insight into what drives their markets, what their customers want, as well the best next steps that need to be implemented. At TubeMogul the marketing team uses its organization’s data to build out best practices and understand trends in viewability, inventory availability, price and other details. Extending this, the organization’s business development and intelligence teams use other areas of big data to work more effectively with publishers and data partners and to gain visibility into business metrics, brand safety and inventory analytics. Finally, the machine learning team used data to make its advertising more relevant by presenting the right ad to the right person at the right time.

Like many organizations, big data was in high demand across business functions at TubeMogul. The 30-user, enterprise-wide rollout of QDS doubled the amount of data processed in half the time as well as provided capabilities for scaling up to meet demands of queries against large data sets while users run queries simultaneously. With this, TubeMogul is now focused on defining insights from faster-run queries.

Data access is often delayed due to the considerable amount of time required by engineering teams to manage an organization’s fixed Hadoop clusters. Without proper cluster management, users don’t have access to the right amount of computing power for projects. At TubeMogul, only engineers were capable of setting up and managing clusters, knowing which specific nodes to run, accessing keys and completing the necessary account management tasks. Like TubeMogul, other organizations are looking to eliminate the need to architect, employ and manage their Hadoop clusters so efforts can be focused on deriving insights from their data. Traditionally, this has required an engineer, consumed valuable engineering time, and often required authoring Python scripts – consuming even more engineering resources. QDS eliminated that need.

With a range of big data tools, organizations now have the ability to quickly and easily run queries on their big data, eliminating the barriers and necessary expertise traditionally required for big data analytics implementation. With data that is easily accessible organization-wide, non-engineering individuals can now tap into entirely new areas of customer and client information. Proper data analytics tools are allowing organizations to focus on deriving even more insights than ever before.

TubeMogul’s 30-user rollout across marketing, business intelligence, machine learning and engineering teams helped reduce cost, drive a five-fold increase in execution of queries with auto-scaling and a doubling of data processed in half the time. With TubeMogul’s new grasp on their data across the organization, its focus has shifted to where it belongs – helping brand advertisers drive results.

Contributed by Jonathan Buckley, Senior Vice President Marketing at Qubole and Adam Rose, Chief Technical Officer at TubeMogul (NASDAQ: TUBE)

 

Sign up for the free insideBIGDATA newsletter.

Speak Your Mind

*