Cloudy IT forecast means change ahead

There’s a cloudy forecast ahead for corporate IT shops, and it’s getting cloudier fast. Analysts at Gartner predict that Infrastructure-as-a-Service, or cloud-based computing, will reach a compound annual growth rate of 41 percent through 2016. As enterprises adopt cloud computing models, they’ll need to make other changes, too, say researchers at the W. P. Carey School of Business.

There’s a cloudy forecast ahead for corporate IT shops, and it’s getting cloudier fast. Analysts at Gartner predict that Infrastructure-as-a-Service, or cloud-based computing, will reach a compound annual growth rate of 41 percent through 2016. As enterprises adopt cloud computing models, they’ll need to make other changes, too. Researchers at the W. P. Carey School of Business expect cloud adoption to bring organizations rapid business-process calibration, increased use of analytics, as well as possible issues around the tax implications of cloud-based IT spending.

Heads in the clouds

“I remember going to a conference at the end of 2011, and a senior-level researcher from a top university was saying most executives had no interest in cloud computing,” recalls Tim Olsen, assistant professor of information systems. “One year later, she was saying the exact opposite. Corporate executives are now sold on the paradigm of cloud computing.”

Olsen isn’t surprised, and he calls the shift to cloud computing a “no brainer.” Professor Uday Kulkarni agrees and sees underutilization of infrastructure as a key driver of this trend.

“It’s a matter of space and time,” he explains. “Space is storage. Space and time refers to processing time needed for computing jobs that are data heavy. These two things create huge costs for companies, and these costs are not sustainable.”

As Kulkarni points out, companies often house data that’s rarely used, and they must build out their processing systems for peak demand, much like electric utilities, which must build generation capability to serve extreme loads on the hottest days and coldest days of the year.

Compounding the underutilization of computing equipment is specialization. “Usually, companies have to customize infrastructure to meet specific needs,” Kulkarni explains. “If you’ve spent money on a large order processing or inventory management system, you might not be able to use that same system for payroll or data mining. The systems could require different types of warehousing and different architectures.” Or, there may simply be turf wars, in which one department insists on its own systems, resulting in duplication of infrastructure.

Wasted expense isn’t the only reason companies are turning to cloud computing. Speed to market also is a driver, Olsen says, because cloud infrastructure lets organizations jump right into the development end of things and skip infrastructure set-up. “If you’re going to develop an app, you probably can’t buy all the servers you need from the get-go” he explains. “The same is true for most companies developing software. Cloud computing lets them spend the least amount of money and still get what they need.”

Another factor making cloud computing more attractive is the maturity of web services themselves. “When you use an app on your phone, you’re consuming information from many different data centers and other services that are being integrated to provide one holistic service,” Olsen notes. This approach to development, which leverages the cloud itself, has become the norm in application development. “It’s mainstream. We all use these cloud-based services in our personal lives, and that familiarity with cloud is making corporations more prone to choose cloud solutions.”

Convergence of forces

What will we use those cloud solutions to do? Among other things, Kulkarni says we’ll eventually use them to deal with big data, or data sets so large and complex, traditional data processing technology can’t keep up.

“When I started in this field, we talked about data from ERP systems. Then customer relationship management came along and created even more data because you record information about every interaction,” he recalls. “That created pretty large data sets compared to ERP systems, but now we’re talking about web data – keeping track of things like click streams and how customers navigate through web pages.” Kulkarni also identifies RFID sensors, speech to text conversion, dynamically changing web pages, video, and, of course, the tweets and “likes” of social media as big data contributors.

According to research from International Data Corporation (IDC), 68 percent of data “is created and consumed by consumers — watching digital TV, interacting with social media, sending camera phone images and videos between devices and around the Internet, and so on.” But, the IDC team also says that enterprises are responsible for an estimated 80 percent of the digital universe, and the universe is growing dramatically.

At the end of 2012, IDC researchers estimated the entire size of the digital universe to be 2.8 zettabytes, or 2.8 times 10 to the 21st power. By 2020, the researchers estimate that the digital universe will reach 40 zettabytes, an amount they calculate to exceed the number of every grain of sand on every beach worldwide … multiplied 57 times.

“Just about three years ago, there was only one zettabyte of data representing the entire digital universe created by humankind,” Kulkarni notes. The volume and rapid growth of these data, combined with competitive pressures, are pushing companies into the cloud.

“Any business can use these data to compete more effectively,” Kulkarni adds. He also says they’ll need to do so, and he challenges his students to come up with ways to apply analytics.

One example he remembers is from a company that refills gas canisters. To ensure that the canisters were still in good enough shape for a refill, technicians x-rayed each unit when it came into the shop. Soon, the company had enough data to know how many dings and dents made a canister dangerous or leaky, and managers applied analytics to predict which canisters shouldn’t be refilled.

Initially, Kulkarni sees cloud computing’s contribution to analytics as support for data storage and rapid data processing needs. This is how many corporate leaders now use it, including such household names as eBay and Netflix. “Netflix needs huge computing power, but their core competencies aren’t in maintaining this power,” says Kulkarni. “Their strengths are in figuring out which movies to own and to recommend, what prices to set and so on. So they go to an Amazon or Google for public cloud computing services.”

Winds of change

Only about 3 percent of the data in the digital universe at the end of 2012 was "tagged, and even less was analyzed, according to IDC. Among the changes cloud computing will bring to corporate life, Kulkarni says looking for opportunities to mine data will be essential to maintain competitiveness. He also thinks people will need to rethink how they look at data storage because of streaming data. “We’re used to storing data, structuring them and then mining them later. Now we’re talking about analyzing data as it comes in. We’ll see the need for cloud computing to catch data on the fly and process it immediately,” he says.

Business processes companywide will likely change with cloud computing, too, notes Michael Goul, professor and IS department chairman. In fact, traditional approaches to procuring information technology capacity will probably be one of the first changes. “If you’re used to buying software and equipment and all your IT processes are focused on that, those processes will have to be revisited,” he says.

So will traditional ERP functions, because companies have moved well beyond experiments and into cloud delivery for mission critical and core systems. Goul says processes will need to become increasingly streamlined and precise to make the most of software-as-a-service offerings.

Service-level agreements will get more complicated, too, Goul says, particularly for big players. “If you’re a global organization, you’ll need a global cloud provider,” he explains. “The provider has to have service quality, up-time, reliability and the ability to respond to service-level agreements to or from Timbuktu.”

Regulation issues will probably crop up soon, as well, Goul continues. Security still is an issue, and, taxes are likely to come under scrutiny. “Leases are treated differently than capital equipment purchases in most countries, and different countries have different tax rates. How do you apply them when your whole company around the world uses software from a cloud service provider?” he asks.

Already, Goul sees such questions being asked and, while cloud computing may reduce workloads for IT departments, it could well add work for accountants. But, he says there’s a silver lining for cloud computing itself in that circumstance. “That’s how you know a technology has made it … when taxes become an issue.”

Bottom line

• Cloud computing has come of age and it’s headed for corporate systems, including core enterprise systems.
• Drivers for the shift include cost savings, speed to market, maturity of web services technology and the proliferation of big data.
• Ahead, companies will need to harness and leverage data through analytics to remain competitive, and cloud computing can support that effort by supporting storage and data processing needs.
• The proliferation of cloud computing in the enterprise will also prompt the need for business process changes throughout the organization, including those associated with the procurement of IT capacity itself.

Cloud Computing: The Evolution of Software-as-a-Service