As we head into the middle of the second decade of this new millennium, it seems all eyes in the IT world are cast upward, to the Cloud. There’s no doubt that cloud computing is set to create a paradigm shift in how we use (and pay for) computing services in the years ahead, but the success (i.e. user acceptance and satisfaction) of that big change is going to be dependent on much more than just the cloud providers.
A leap into the ubiquitous computing future envisioned by cloud proponents will require the right end-user devices, the right Internet provider service models, and the right user education programs to support the move. It will also require a change in mindset on the parts of consumers, IT professionals, and service providers alike. Let’s look at how the evolution of computing could be eased by the right business decisions, as well as the pitfalls to watch out for that could turn cloud computing into a disaster.
I’ve said before that two major recent trends in technology are on a potential collision course. The first is cloud computing, which aspires to put all of our data and applications on nebulous servers that exist somewhere “out there” (they can’t tell you exactly where for security reasons), where it will be taken care and protected by benevolent experts with resources for doing so that are far greater than our own. The idea is certainly attractive: less work, less worry, ubiquitous access.
The problem is that other, darker trend – which implements one of the essential characteristics of cloud computing (according to NIST): measured/metered service. It’s one that Internet Service Providers have been embracing eagerly over the last few years. The top two wireless providers, AT&T and Verizon, have done away with unlimited plans for their data services, right at a time when they’re also pushing consumers to use 4G as their primary Internet service instead of cable, DSL, FiOS and other wired options. The low caps (ranging from 1 to 10 GB per month) make that idea economically unfeasible for power users and has taken all the fun out of 4G for many of us.
The cable companies have made several attempts to cap bandwidth on their Internet plans, although the limits they’ve set have usually been much more generous (average 250 GB per month). But those numbers are coming down. After being pressured to get rid of their cap in 2009, last summer TWC introduced a 30 GB data cap in Rochester, N.Y., then yanked it after pressure from state politicians.
What does all of this have to do with business computing, you might ask? The problem is that due to yet another popular trend – mobile computing – many workers are using consumer data plans to connect to the corporate network and get their work done. Some are reimbursed by the company for their ISP costs and some aren’t, but either way, usage caps make it more expensive for them to engage in high-bandwidth activities such as video conferencing, which are so useful for telecommuters.
Once upon a time, there were dozens of “mom and pop” ISPs in most markets. These were small service providers that leased a T-1 line (or several) from the phone company to connect to the Internet, set up a bank of modems, and resold their Internet service to consumers and small businesses. That was before the phone companies decided to sell Internet services directly to consumers themselves, and quickly took over the market and drove the small fry out of business. Now, with their only competition the cable companies, the two can dominate the market and keep margins on their services extremely high.
If expensive metered data plans become the only option for Internet service, users will be reluctant to depend on cloud services that require using up some of those high-priced megabytes every time you want to run an application or open a document. In order for cloud computing to realize its full potential, we need cheaper, faster and more reliable Internet service as the foundation on which it can grow. But how do we get that, when the cable and wireless companies seem to have interests that conflict with those of the cloud providers?
Maybe the answer is for the cloud companies to become ISPs themselves. Google has led the way in this venture with its Google Fiber gigabit service, which not only has no data cap, but also provides users with speeds of 1000 Mbps, which is twice as fast as Verizon FiOS’s highest speed package (but which does apparently allow data usage of at least 77 TB per month under its “unlimited” plan). All this costs $70 per month. In comparison, Verizon Wireless’s 10 GB per month 4G data plan costs $100 per month and Verizon’s 500 Mbps unlimited FiOS plan is $299 per month.
Should Microsoft and other top technology companies consider getting into the ISP business as well? How about Amazon? They seem to have their fingers in a lot of pies. Perhaps the entry of several alternatives ISPs would shake up the cable and phone companies and the competition would drive cost down and bandwidth limits up across the board. One thing is certain: something disruptive needs to happen, in order for cloud computing to reach its full potential. Personally, I like the idea of more competition a lot better than the specter of more government regulation to “fix” the problem.