Improving online advertising with Hadoop big data tools

The online advertising marketplace has become more complex as the proliferation of consumer data in the space has increased demand for targeted buys. A number of startups have emerged around the practice of real-time bidding (RTB), which allows advertisers to buy spots in real time auctions and spread purchases across multiple publishers. To analyze and deliver billions of ad impressions every day, these RTB startups are relying on big data analysis tools and NoSQL database technology such as Hadoop HBase.

A recent InfoWorld article explored the importance of NoSQL database technology in the RTB industry, noting that real-time analysis of the information in more than 500 million consumer profiles is key to the success of one such vendor, AcuityAds.

"Big data technologies are critical to drive the RTB platforms efficiently," Tal Hayek, AcuityAds co-founder and CEO, told the publication. "To stand apart from the crowd, our real-time bidding system needs to be as fast as possible. It became absolutely critical that we identify a database, which would enable us to respond to opportunities in less than 50 milliseconds even as we analyze billions of ad impressions each day."

NoSQL tools such as Hadoop HBase are ideal for the type of rapid processing required of RTB engines. HBase enables random, real-time read/write access, which is ideal for the type of constant querying needed to handle massive volumes of transactions. RTB systems must pull data thousands of times per second, ruling out the batch processing approach of many Hadoop tools such as Hadoop Hive.

As RTB startups look to deliver reliability and speed in their effort to re-engineer the online advertising space, the ability to rapidly communicate with customer information databases is critical to success. With big data technology such as Hadoop HBase, targeted ad sales can be pushed closer to real-time responsiveness.

Categorized by :

Leave a Reply

Your email address will not be published. Required fields are marked *

Hortonworks Data Platform
The Hortonworks Data Platform is a 100% open source distribution of Apache Hadoop that is truly enterprise grade having been built, tested and hardened with enterprise rigor.
Get started with Sandbox
Hortonworks Sandbox is a self-contained virtual machine with Apache Hadoop pre-configured alongside a set of hands-on, step-by-step Hadoop tutorials.
Modern Data Architecture
Tackle the challenges of big data. Hadoop integrates with existing EDW, RDBMS and MPP systems to deliver lower cost, higher capacity infrastructure.