According to new research, demand for Internet usage will start to outpace the capacity of the Internet's access points. This potential crunch could spell trouble for CIOs.
The Nemertes Research Group examined revenues and expenditures of Internet infrastructure companies and the changing pace of demand for bandwidth to determine this. Nemertes president Johna Till Johnson said her firm isn't predicting a cataclysmic failure of the Internet. Instead, the quality of service will start to drag for businesses, stifling innovation and slowing the speed of transactions and communications.
Johnson explained that the pace of investment in the infrastructure that connects individuals and businesses to the Internet infrastructure is falling behind demand. Nemertes projects that by 2010, for instance, the average amount of time it takes to download a YouTube video will jump from 10 seconds to as much as two minutes.
"It's that last mile, the pipes at the edges. The access circuit," she said. "That would be cable for consumers, DSL. For businesses they are telephony circuits, and increasingly they're wireless. Wireless is one of the key limiting factors. The technologies for wireless broadband access exist, but they are not deployed widely."
Johnson likened the problem to a series of broad interstate highways. Traffic is flowing smoothly on the multilane highways, but the narrow access ramps that lead to those highways are clogged with cars waiting to get on.
Nemertes estimates that it will take as much as US$137 billion in global infrastructure investment in the next three to five years to prevent significant service declines, including US$42 billion to US$45 billion in North America alone. Nemertes research shows that service providers plan to spend just 60% to 70% of that total required global investment.
This could spell trouble for companies that conduct critical communication and business transactions and processes via the Internet. Slower or interrupted service means lost productivity.
Midmarket CIOs should pay attention
Johnson said large enterprises should do all right in this environment because they can afford their own direct access circuit to the Internet if the need arises. But smaller enterprises and midmarket companies will be vulnerable.
"If you're one of those lucky people who can get a big fat pipe into the Internet cloud, none of this matters," she said. "Where this gets fuzzy is the trend we've been seeing the past five years in our research, where enterprises are getting more and more distributed with branch offices. Some of these branches are small facilities and their staff is highly distributed. These are not places where you'll be getting fat pipes. And if you're moving toward telecommuting you're relying on the lowest common denominator of access to your employees. That's a bad idea. You've got a telecommuter waiting hours for an accounting dump of sales figures. They're not getting access to the data they need in a timely fashion."
Johnson said midmarket CIOs should pay close attention to their connectivity in the coming years. And they should advocate for infrastructure investment.
Johnson said government policymakers may have to make a choice to ensure that capacity doesn't become an issue. She said some believe the Internet is as primal a right as access to basic utilities, and the government should guarantee access for everyone. Others believe the free market should determine how investment is made in the infrastructure. Either way, Johnson said some sort of comprehensive policy is needed to encourage more investment in the infrastructure.
"We've been told we're fear-mongering," Johnson said. "No one wants to look at this problem. I don't know what the answer is. I don't know what the right policy is. They need to come up with that."
Bruce Mehlman, co-chairman of the Internet Innovation Alliance said he agrees with most of Nemertes' findings, "but we're a bit more optimistic."
Mehlman said he thinks investors and innovators will rise to the challenge that Nemertes has uncovered and make sure that bandwidth will meet demand.
"As I see what [Nemertes] is saying, the usage curve is steeper than what most people appreciate. To maximise the value of the Internet in 2010, the capacity curve has to be made steeper than is currently being made," Mehlman said. "That can be done through investment and innovation. This is akin to the Y2K scenario, where you could have an issue, but prudent planning and investment mean it goes off with a whimper. I see this as a call to investment, to help network providers and policymakers recognise the magnitude of what needs to get done."