When Hurricane Irene threatened to pummel New York, Phil Green couldn't have known that what was about to happen would change his perspective of the technological landscape forever.
His biggest concern centered on a park called Flushing Meadows, home to the U.S. Open. The tournament, the final of the four tennis tournaments that make up the Grand Slam each year, was well underway. But Green wasn't worried how high winds and heavy rain might affect the game. Instead, as senior director of advanced media for the United States Tennis Association (USTA), he was focused on something much more basic: electricity.
"We were hearing things like, 'no power for days,'" he said.
For an employee who helps to oversee USOpen.org, a power outage was like a death sentence. During the two weeks of competition, the website experiences its peak traffic for the year. Those visits are dependent on fresh articles, photos, videos and analytics. Irene was quickly becoming a kink in the plan.
But there was a potential solution. Green agreed to allow representatives from IBM, a technical support provider for USTA, to move the infrastructure and tools to the cloud until the storm passed, enabling employees to continue updating the site.
"There was no hesitation," he said. "Was I a little bit on edge? Of course."
Unlike Green and the USTA's leap of faith, many businesses have been reluctant to embrace cloud computing. To be fair, most organizations don't experience a do-or- potentially-die moment when faced with the decision. Nor do they have a vendor at the ready to provide the product and support. Still, new research suggests cloud computing is gaining ground, especially for business intelligence (BI) and analytics, with significant numbers planning deployments in the not-so-distant future.
The key is elasticity
According to the recent 2012 TechTarget Cloud Pulse survey, more than half of the respondents are using cloud IT services and applications today. Of the 643 respondents who detailed what tasks they were currently using cloud IT services for, 32% said BI and analytics.
While that number may seem high, the results don't surprise William McKnight, who said both BI and analytics are broader in meaning and more decentralized in practice these days.
"We are indoctrinated to think we're running the company on analytics," said McKnight, president of McKnight Consulting Group LLC, based in Plano, Texas. "But it could be shallow [rather than] predictive analytics."
More on cloud BI and analytics
Read up on Gartner's six elements of cloud analytics
Understanding the challenges and opportunities of SaaS and cloud BI
Is cloud BI in your future?
"One of the nice things about the cloud is that you don't have to predict ahead of time exactly how much you're going to need," he said. "The solution scales out. And, if you go with a public cloud provider, they will have the hardware at the ready when you've tapped out your current allocation. And you can continue to scale."
The cloud can also be an asset for organizations embarking on a "big data" analytics journey, according to Mark Theissen, CEO of Cirro Inc., which provides a product that connects data from Hadoop to BI tools. Big data tends to include multistructured data such as text, which multiplies quickly and can be difficult to load into traditional databases. Businesses faced with this challenge are left seeking a workaround, and some have turned to the cloud in the form of a technology called Hadoop. An open source software, Hadoop is known for its quick consumption and storage of large amounts of data -- regardless of format -- across a distributed computing environment. Still, Hadoop isn't a mainstream technology, and the use cases can be tough to find.
"It's really the leaders of your large corporations that I've seen doing the most when it comes to working with these challenges," said Theissen. "They have the money, the data, and they certainly are looking for these competitive differentiators."
Private or public cloud
Businesses may increasingly be embracing the cloud, but, Theissen said, they tend to gravitate toward a private rather than a public offering, citing security and availability concerns.
"We're in very early days for BI and analytics in the public cloud," Theissen. "Everyone's more focused on the private or hybrid implementation."
That was precisely the case for USTA, which continues to use IBM's private cloud offering. But, according to the Cloud Pulse survey results, even that may be changing. When asked what services respondents will be using in the public cloud or the public portion of the hybrid cloud in six months, 24.5% of 351 respondents said BI and data management.
That response proved to be the most popular when measuring where respondents are now compared with where they plan to be six months from now, gaining a 17.1% increase. The next closest -- hybrid cloud integration -- saw a 15.1% increase.
Part of the increase may be due to a maturing market. Razorfish Inc., a marking analytics service provider based in Seattle, Wash., moved to a hybrid cloud two years ago when the company was sold by Microsoft to Publicis Groupe. It was a transition the company had to explain to its customers, a conversation that tended to go pretty smoothly, according to Matt Comstock, vice president of BI at Razorfish.
"We sold it on the fact that it was going to be less expensive and more reliable than what we had," said Comstock, though he added his clients typically reside in the marketing rather than IT department.
Today, Razorfish is an Amazon Web Services customer, performing almost all data and analytics functions in the cloud. The company utilizes Hadoop to pull in unstructured data and Teradata-Aster to transform that data into structured information they can then analyze.
As for the common cloud concerns, such as security, Comstock said it's never been a problem.
"We've had an outage or two," he said, pointing to this week's Amazon outage as one that affected his company. "[But] you're getting a lot of good practices … that would take time and money to get done internally. That's definitely an additional benefit."
This was first published in October 2012