OpenX Enterprise Technology
OpenX's high-performance technology infrastructure enables networks to customize their ad server, control their targeting data, and efficiently deliver billions of ad impressions across a wide network of publishers
Advanced Optimization and Targeting
For publishers to maximize their advertising revenue, they need to precisely target the delivery of advertising to the right audience, at the right time. An ad server platform must provide an easy way to describe a target audience, and then deliver ads to that audience with pinpoint accuracy. And since every business has different ways of describing their audience, an ad server should flex to accommodate different targeting methodologies.
In addition, companies must continually decide how to optimize the delivery of both premium and unsold inventory to maximize revenue. OpenX Enterprise leverages advanced algorithms to help publishers optimize their ad revenue, and offers unique options for targeting users.
The following features make it easier for publishers to get the most value for their inventory while helping marketers find the precise audience they want to target:
- Multivariate forecasting: allows publishers to get an accurate snapshot of their available premium inventory, which helps drive premium revenue
- Real-time prioritization: prioritizes ads in real-time to manage the optimum delivery of contract campaigns and unsold inventory
- Profile targeting: enables publishers to create segments of users and target them across their inventory; or bring their own data and create customized segments for user targeting
- Revenue optimization: automatically adjust the delivery of ads based upon real-time revenue calculations, helping you optimize the value of unsold inventory
These key features are critical to maximizing the value of publisher inventory. Only with OpenX can customers can bring their own customized data to create a unique solution for their business and target audience.
High Performance Hybrid Cloud Infrastructure
Another challenge facing Internet marketers is the importance of minimizing technology costs while maintaining the performance and reliability of their ad serving solution. The advent of cloud computing has sparked a renewed interest in pay-as-you-go solutions that can scale according to a business’ growing needs. With the right tools and expertise in place, companies can benefit from tremendous cost savings as well as a scalable, high-performance solution. OpenX uses a hybrid cloud architecture serving a majority of OpenX owned servers in multiple high performance data centers around the world.
In addition, OpenX utilizes cloud computing providers such as Amazon to tap into additional data or delivery performance automatically if needed. OpenX also takes advantage of the cloud by using select Amazon locations (Singapore, Ireland, etc.) to complement our own data centers to ensure superior performance worldwide. OpenX hybrid cloud infrastructure provides a flexible, scalable, efficient, and redundant on-demand computing framework.
OpenX has employed industry experts with experience at companies such as Google and Yahoo! to design, develop, and implement this infrastructure. This best-of-breed computing infrastructure translates not only into rock solid reliability and performance, but also significant cost savings, which are passed along to customers of OpenX.
- Geographical traffic routing: provides redundancy across datacenters and Internet service providers to deliver ads from the location that best optimizes performance
- Redundant load balancing: eliminates all single points of failures in the system, and ensures high performance by distributing load across a cluster of multiple servers
- Content delivery network: delivers media objects to users from optimum locations across the world to ensure your ads are delivered quickly and efficiently
- Automated deployment: automatically forecasts and allocates additional servers to the cluster to compensate for high traffic periods or machine outages
- Redundant design: Anonymous services and zero points of failure mean that multiple hardware outages have no effect on application availability
- Advanced monitoring and instrumentation: provides real-time metrics and advance alerting, enabling engineers to react before problems occur
These are just some of the features that provide the high performance foundation on which OpenX Enterprise is built, while also reducing costs.
Massively Scalable Platform
Internet-scale technology solutions have a particularly challenging requirement to scale massively under load while still delivering high performance. With the rapid growth of Internet traffic, just adding faster and faster hardware is not a solution; these systems must be designed from the very beginning for high performance at scale. This is especially important for an advertising technology solution, since revenue is on the line.
OpenX Enterprise is designed using scalable, proven technologies that ensures the platform will scale along with customers’ business needs. Many of these technologies are in use today in mission critical systems at companies such as Yahoo!, Google, Facebook, and Ericsson. OpenX has hired experts to adapt these proven technologies to the OpenX infrastructure, and developed additional open source tools to ensure reliability and scale.
The technologies OpenX uses include:
- Hadoop data grid: a powerful open source batch processing system that uses the MapReduce framework and is able to process multiple petabytes of data per day
- Cassandra distributed hash table: a scalable, redundant open source system used to store and replicate advertisements and targeting data for ad delivery, serving up data in milliseconds
- Erlang optimization core: a massively scalable, open source concurrent programming framework used to optimize advertisements in the OpenX system, able to optimize millions of ads per hour with extremely high availability
- Lightweight event system messaging: a scalable open source messaging framework optimized for connectionless messaging of structured data
- Distributed persistence: objects are stored across multiple machines to ensure high performance and availability, and are backed up regularly to ensure data integrity and security