August 18, 2011
Mobile Data: Catalyst for Change in BI and Analytics
Topic: Experts in BI
Set at a suburban distance from the craziness of the Las Vegas Strip is one of the wonders of the computing world and a stunning example of efficient electrical power consumption on a massive scale: Switch Communications' SuperNAP, a 100-megawatt, 407,000-square- foot data center campus that is still growing. Although the heavily guarded and structurally reinforced facility hugs the ground, it is about as close as one can get to the physical reality of cloud computing.
SuperNAP serves government and private enterprises, including the ATM transaction infrastructure for much of the Strip as well as gaming facilities around the world. EMC, HP, and other cloud-computing providers have co- located systems at SuperNAP.
Stationary systems such as ATMs or point-of-sale devices generate plenty of data, but global expansion in the use of mobile devices is a huge driver behind demand for larger and more powerful data centers to handle both transaction and analytic workloads. It's not inconceivable that some day nearly everyone in the world will have a smart phone, tablet, or other mobile device with them at all times.
People will be using the devices for communication, collaboration, and social media access as well as for commercial transactions, healthcare needs, and more, which will generate enormous volumes and varieties of data. Gaming concerns, of course, are ramping up for continuing growth in the use of mobile devices for gaming activity. Organizations in nearly all industries need to prepare for the onslaught of mobile data and its rich potential for business intelligence and analytics.
EBay, which has massive data management, enterprise data warehousing, and Hadoop analytics systems running at SuperNAP, is seeing some of its fastest transaction growth coming from mobile devices, particularly through eBay's application running on the Apple iPad. The company's mobile-device-generated transactions are expected to reach from $2 billion to $4 billion in sales this year. The company's SuperNAP-hosted analytics are pushing the Teradata technology envelope to gain business insights from the real-time data inflow. Experimentation being done at SuperNAP by eBay will eventually find its way into future Teradata (and Aster Data) product releases.
During a tour of the SuperNAP facility given to industry analysts by Switch and Teradata, Oliver Ratzesberger, eBay's senior director of analytics platforms and delivery described how the company is implementing agile methods. Ratzesberger said that eBay is converting all analytic application development to scrum-oriented, two-week sprints to improve the timeliness of big-data discovery and keep up with new and iterative questions that arise during the process of discovery.
Organizations will need the speed to value that agile methods can potentially deliver so that they can meet demand for tools business users need to discover insights from big-data sources such as mobile device activity. In addition, organizations will need to reexamine traditional data warehousing policies that have limited query activity for performance and cost reasons. In the emerging era of big-data analytics, organizations must aim for faster analytic application development and increase the value of those applications by creating policies and technology strategies that open up, rather than clamp down on, ad hoc queries.
Spatial Data Analytics: Time is Ripe
Data generated by mobile devices will reveal new avenues for organizations to explore to understand the "paths" of customer behavior over time and across different channels as they interact with products and services. One type of data that will certainly garner more attention is spatial (or geographical) data.
Spatial data analysis is already important to firms engaged in "micro" marketing, which involves digging deep into relationships between demographic data, online customer behavior data, and location information. Organizations are also using spatial data analysis for real estate and store-location decisions as well as to decide what merchandise to put in specific stores. The availability of data from mobile device activity could be what propels organizations to seek functionality in BI tools and analytic applications that can give decision- makers a location dimension to add to their business analysis.
When it comes to BI, analytics, and mobile devices, much of the industry attention has been on implementing BI and analytics tools on smartphones and tablets or through Web- based applications that implement HTML5. This is indeed an important trend that will grow in significance as business users acquire tablets such as the iPad and demand BI and analytics on the go.
However, organizations should also focus on the other side of the explosion in mobile device adoption: the data that devices are generating. This source of big data will eventually cease to be exotic; it will come from smartphones and tablets that nearly everyone will be carrying. So get ready: Mobile devices are pushing BI and analytics into a new dimension.
David Stodder is director of research for business intelligence at The Data Warehousing Institute (TDWI). David can be reached at firstname.lastname@example.org
Copyright 2011. TDWI. All rights reserved.