Print this article | Return to Article | Return to CFO.com
Companies will increasingly use cloud computing for essential business functions as their fears about security and other issues prove groundless, a consultant and former CIO predicts.
David McCann, CFO.com | US
June 4, 2010
Cloud computing is fast gaining traction as an effective way to scale computing capacity according to shifting needs and to lower the operating cost of IT services. Indeed, most business functions are worthy candidates for outsourcing to the cloud, to the point where companies may soon begin to question the need to maintain large, expensive data centers.
That is the contention of Michael Hugos, a former chief information officer and a principal of the Center for Systems Innovation. Co-author of a forthcoming book on cloud computing, Hugos will speak on "Technologies that Will Change the Organization" at the upcoming CFO Core Concerns Conference, to be held June 27-29 in Baltimore.
To be sure, many companies are wary of moving systems with highly sensitive information to the cloud, largely because of perceptions that data is less secure there. But Hugos has little patience with such worries, pointing out that one of the most successful cloud vendors to date is Salesforce.com, which gained its footing among small companies but now has many large customers as well. "What could be more critical to a company than its lists of customers, prospects, and pricing proposals?" he asks. "A lot of the discussion about security rings hollow to me."
Few companies, he claims, can match the security capabilities of big providers of cloud-based infrastructure such as Amazon, Google, IBM, and Hewlett-Packard, whose profits are directly tied to providing secure data centers.
In fact, a key advantage of the cloud is that it lowers key risks, says Hugos. For example, for a company developing a new product with an uncertain long-term market, renting the associated computing capacity is far less risky than expanding the company's data center. "[The cloud] provides a lot of flexibility to probe into new markets, ride them as far as they go, and exit without a bunch of sunk cost," he says.
Another concern about the cloud is the lock-in effect. Cloud vendors use different operating systems and processes and have not yet provided clear specifications for portability of data and applications across clouds. Diane Bryant, the chief information officer of Intel Corp., recently told CFO that cloud providers "make it very easy for you to develop and run applications. But then that app is locked into their environment. It will be very expensive to port it into a different one, and they know that."
Hugos agrees that lock-in is a valid concern with some cloud vendors, but notes that companies also get locked in when they spend several years and millions of dollars installing an on-site enterprise resource planning system. "Lock-in is a problem, but not a new one, and it's not only related to clouds," he says.
A good portion of the anticloud noise comes from IT departments, but Hugos, a longtime CIO himself, suggests that their objectivity is questionable. "Cloud computing is not happy news to many people in the IT world," he says. "Most IT jobs are for system administration, and that's going to be outsourced. It will be the equivalent of what happened to blue-collar manufacturing jobs here in the United States in the 1980s."
The shift to the cloud is directly analogous to the development of electric power grids about 100 years ago, says Hugos. Operators of large commercial buildings debated whether to tap into the new power sources rather than continue to generate power on site. They weighed the obvious cost benefits against fears that the grids might not be reliable and would pose an unacceptable security risk, since it would be easy for someone to cut the power lines to the building. Those were valid concerns, says Hugos, but not compelling enough to carry the debate. Electric utilities steadily became more reliable and secure, and by the 1920s the debate was over.
"In the showdown between security and profitability, profitability will win," says Hugos. "I'm not suggesting that companies get rid of all their data centers in the next six months. But over the next four or five years, as they need to refresh their servers, I don't see why they would want to continue adding hardware that is becoming obsolete."