Real Time: Real Insights

BrandPost By Stan Gibson
Jun 16, 2015
Big Data

Big data analytics are turning into a killer app for the cloud

The cloud is an ideal vehicle for big data analytics, thanks to cloud’s ability to provide large amounts of computing and storage capacity as needed. All types of big data have been considered fair game, with the exception of real-time data. The cost of sending it to and from the cloud and the timeliness of the cloud-based data insights just didn’t add up. Until now.

Microsoft Azure Stream Analytics is one of several real-time big data analytics services to be rolled out in recent months, driven by the changing economics of performing real-time analytics on large data sets. Analysts say the trend is momentous.

“This is a killer application for cloud. This is the way that cloud is going to change business,” says David Linthicum, senior vice president and cloud architect at the consultancy Cloud Technology Partners.

“You can use real-time data analytics to find out things you never could afford to before,” asserts Brian Hopkins, vice president and principal analyst at Forrester Research.   

The case for gaining business analytics insights in real time is compelling. After all, everyone talks about business agility, and you can’t get more agile than to respond to events as they are actually taking place.

Internet-of-Things (IoT) applications are a natural fit. Think about operational data in a factory. In the past you might have to put that data into a data warehouse to analyze it for trends. After a month, you would have the information you need to make adjustments to increase operational efficiency. But in the meantime, your factory has been running inefficiently for an entire month.  In contrast, the ability to create a dashboard and get metrics in real time enables you to make adjustments in the moment, eliminating inefficiency more quickly.

Data about customers and their behavior is, as always, a prime candidate for analysis as well. The IDG Enterprise 2015 Big Data and Analytics survey found customer-driven sources of data the most common sources of data at most organizations.

In one of the better-known examples of customer analytics, Netflix continuously examines the movie download patterns of all its customers to better understand their preferences and to pitch them on follow-up downloads. Think about how those predictive capabilities might work for your business.

There’s more. Hospitals can begin to aggregate, anonymize and analyze patient data so that treatments might be better understood as they are being administered, leading to potentially life-saving changes being made in real time.

For some applications, near-real-time analysis will do just fine — ten minutes of latency might be more affordable and just as good. “You have to determine how much you want to pay for an answer, and when you need it to make a difference,” says Hopkins.

But the trend toward real-time analytics may be irresistible. Linthicum thinks all big data analytics will head that way.

“It’s been a long time coming. With traditional data warehousing, you never got information in a timely way. Now we have the technology that can be bought per drink,” says Linthicum. “Nothing delivers a higher ROI than the ability to use data effectively. [With real-time data] you’re dealing with perfect information.”